#help-with-radio | adafruit | Page 10

restive fjord · 2018-08-09T15:00:09.126Z

``` PORT->Group[0].OUTTGL.reg = (uint32_t)(1 Group[0].OUTTGL.reg = (uint32_t)(1 << 23); ```

restive fjord Aug 9, 2018, 3:00 PM

#

PORT->Group[0].OUTTGL.reg = (uint32_t)(1 << 23);
PORT->Group[0].OUTTGL.reg = (uint32_t)(1 << 23);

languid phoenix Aug 9, 2018, 3:00 PM

#

Yeah, that's a little bit beyond me..

restive fjord Aug 9, 2018, 3:00 PM

#

That means 'blink the LED as fast as possible' on a SAMD51 board. ;)

languid phoenix Aug 9, 2018, 3:01 PM

#

Alrighty

restive fjord Aug 9, 2018, 3:01 PM

#

OUTTGL is a toggle.

#

1 << 23 means to set the 2^23 bit in a 32-bit register (uint32_t).

#

0000 0000 1000 0000 0000 0000 0000 0000

#

that's what it means. ;)

#

I always get them mixed up so check if there's betting money involved. ;)

#

PORT ->Group[0] refers to PA23 (the LED) which is on PORT A.

#

(the 0 there for 'Group[0]is theAof PORTA`.

#

I decoded it, I didn't author it (at all).

#

Haven't looked at any documentation at all, either. ;)

languid phoenix Aug 9, 2018, 3:06 PM

#

I'm going to assume that you are a computer engineer then 😃

restive fjord Aug 9, 2018, 3:07 PM

#

No but I spent like way too much time figuring all this out.

#

;)

languid phoenix Aug 9, 2018, 3:07 PM

#

Ah

restive fjord Aug 9, 2018, 3:07 PM

#

It was so much easier to just try something and run it on the hardware, than to look it up anywhere in a structured way.

#

It if smokes, that's 30 bux. No big crisis (until it's the third one this week, that is. ;)

#

As it turns out, that strange code turns the port pin designated as D13 (in Arduino-speak) ON (only if it was presently OFF) and it also turns it OFF (if it was already ON).

#

I knew it was an OUTPUT so OUTTGL makes sense as 'output toggle'.

languid phoenix Aug 9, 2018, 3:10 PM

#

Ah, good catch

restive fjord Aug 9, 2018, 3:10 PM

#

I'm guessing the documentation (datasheet) would say 'OUTPUT TOGGLE REGISTER'.

#

My method is obtuse (to say the least) but by the time I have working (explained, or somewhat explained) code I've looked at so much unrelated stuff, that I have a good working knowledge of an entire subsystem (whether I wanted that or not, I now have it as a side-effect of my approach to 'learning').

languid phoenix Aug 9, 2018, 3:12 PM

#

Not a bad way to learn.

restive fjord Aug 9, 2018, 3:12 PM

#

I really like the Silabs datasheets for 8051 and really do not like our SAMD51 datasheet.

languid phoenix Aug 9, 2018, 3:14 PM

#

What about it do you not like?

restive fjord Aug 9, 2018, 3:14 PM

#

https://media.digikey.com/pdf/Data Sheets/Silicon Laboratories PDFs/C8051F330_1, 330D.pdf

#

It's visual, what I do and do not like about it.

#

Above is my favorite datasheet for a microcontroller.

#

On the 'splash' page it tells you all the good news about using C8051F330.

languid phoenix Aug 9, 2018, 3:15 PM

#

Indeed!

restive fjord Aug 9, 2018, 3:15 PM

#

You can do a lot in 768 bytes of RAM.

languid phoenix Aug 9, 2018, 3:16 PM

#

I love how they call it a datasheet, when it really is more of a full length novel 😛

restive fjord Aug 9, 2018, 3:16 PM

#

8 kb flashROM is roomy

#

right .. yeah chapter 31 . it's like going to the movies on a Saturday to get the next installment.

#

for the rest of your life. ;)

#

I have a working C2 (two wire) implementation for the C8051F330 .. it's like JTAG sort of.

#

Also a serial bootloader..

languid phoenix Aug 9, 2018, 3:18 PM

#

Nice!

restive fjord Aug 9, 2018, 3:18 PM

#

C2 is probably closer to the SWD interface for SAMD21.

#

Well the nice thing is I can swap the chips .. I use the F330 to talk to another F330 via the C2 interface.

#

(I didn't write any of the original or think any of it out; I ported existing work to F330 simply because it was the one chip Silabs had in that product family in a DIP20 package.)

#

I think the port came from either F300 or F310 (same product family).

#

So every chip I acquire has the bootloader on it and can be used to educate a new factory chip to get a bootloader onto it for the first time. ;)

#

If I lose the very last of them, I have no means to get started again.

#

(think about that, for a minute, NASA ;)

#

Like going to the Moon, and forgetting the cork screw -- how are you going to open the bottle of wine?)

#

(because what else would you on the Moon?)

languid phoenix Aug 9, 2018, 3:24 PM

#

You don't have the bootloader saved anywhere else?

restive fjord Aug 9, 2018, 3:24 PM

#

There's no means to get the bootloader into a chip without that program and without a chip that has that program already loaded on it.

#

I've had a board now for 15 years or so I can't do anything with because of this issue.

languid phoenix Aug 9, 2018, 3:25 PM

#

well, that is a problem

restive fjord Aug 9, 2018, 3:25 PM

#

I could go the canonical route and load Windows<tm> and Keil stuff and upload someone else's work.
I could have done so 15 years ago.

#

That wouldn't be the same thing at all.

#

The whole point is to run my code on it, not something I don't understand, on it.

#

It's a really nice board. I think they still have them!

languid phoenix Aug 9, 2018, 3:27 PM

#

Nice!

restive fjord Aug 9, 2018, 3:27 PM

#

C8051F120 developer's board.

languid phoenix Aug 9, 2018, 3:27 PM

#

I get wanting to do it all yourself.

restive fjord Aug 9, 2018, 3:27 PM

#

My goal is to have an entire system I understand. All of it.

#

I do not think I will make it, but it's my goal. ;)

languid phoenix Aug 9, 2018, 3:27 PM

#

That's an awesome goal to have!

restive fjord Aug 9, 2018, 3:28 PM

#

Thanks. You're very encouraging.

#

That's a valued trait around these parts. ;)

#

Might want to visit at a CircuitPython meeting (right here, on a Monday at .. oh.. 18:00 UTC).

languid phoenix Aug 9, 2018, 3:29 PM

#

I've listened to the recordings of a few of those.

restive fjord Aug 9, 2018, 3:29 PM

#

there's no requirement .. you are free to audit anytime in realtime.

#

they call on you unless you ask not to be called on. ;)

#

I made a Raggedy Anne that blinks red evil LED eyes is a fine report to make in that meeting.

#

;)

languid phoenix Aug 9, 2018, 3:31 PM

#

Lol. Yeah we'll see 1800 UTC is right in the middle of my workday, but I might have to jump into one before the school year starts up.

#

Been trying to get more into Circuit Python this summer, planning to push my students to try it out more.

restive fjord Aug 9, 2018, 3:31 PM

#

I think they should have a provision for pre-recorded contributions, personally. ;)

#

I think the learning curve on CircuitPython is pretty good. I find it helpful for hardware stuff too.

languid phoenix Aug 9, 2018, 3:32 PM

#

I think so too, and for the students who just don't get it, I can always nudge them to MakeCode

restive fjord Aug 9, 2018, 3:33 PM

#

I was supposed to be an electronics instructor in the Air Force, but I was only 21 and nobody spoke to me about it. I got orders to do it, but no meeting to ask me to do it. So I turned down the orders. ;)

#

MakeCode is great because it's visual.

#

(when you graduate technical school, everyone gets their orders at the same time in that class; and they allow you to 'swap orders` with a classmate)

languid phoenix Aug 9, 2018, 3:35 PM

#

They were going to make you an instructor straight out of tech school?

restive fjord Aug 9, 2018, 3:35 PM

#

Yeah I thought that was a bit much.

#

I was an honor grad and had 100 percentile average on every test, except a few, bringing it down to 97 percent. ;)

languid phoenix Aug 9, 2018, 3:35 PM

#

ah

restive fjord Aug 9, 2018, 3:35 PM

#

Plus I'd been a ham seven years at the time.

languid phoenix Aug 9, 2018, 3:36 PM

#

Well that would help 😛

restive fjord Aug 9, 2018, 3:36 PM

#

They knew I was a natural (most of my class hadn't heard of any of this stuff before they went to those classes).

#

I would have made a good instructor but a poor Airman, by their plans for me -- that's how I viewed it.

languid phoenix Aug 9, 2018, 3:37 PM

#

lol

#

I started teaching some stuff right after I graduated college. I now almost have the full confidence to teach college students just a few years younger than me.

restive fjord Aug 9, 2018, 3:37 PM

#

I would have laughable authority on matters 'air force'. I'd be a fool. ;)
(not that I've evaded that, haha)

#

I also taught high school electronics ..in high school. student teacher spot.
I was (rightfully) dismissed shortly after I'd begun.

#

It was a bad fit (I didn't have the social presence of mind to do a good job with them, emotionally)

#

I said you people and it all died right then and there, in that moment.

languid phoenix Aug 9, 2018, 3:40 PM

#

Ah, yeah. The social interaction aspect is probably the most important part of teaching I've found. I'm technically a lab director, but I end up teaching informally or giving workshops a lot. Attitude and demeanor make a huge difference.

restive fjord Aug 9, 2018, 3:42 PM

#

I didn't develop compassion (expressed compassion) on that level for decades afterward.

#

So it was never going to work -- timing was way wrong.

languid phoenix Aug 9, 2018, 3:44 PM

#

Something I'm learning as well. I tend to be a little harsher with students. I deal with mostly junior level engineering students, and expect them to know more than I guess they do. They don't like it when the only answer they can give me is "I don't know".

restive fjord Aug 9, 2018, 3:44 PM

#

I love being harsh ;)

#

I think I would be very good at leading a student out of I don't know because I know what it feels like to have to say only that much.

#

It's all stage fright.

#

They know a lot more than they know they know. ;)

languid phoenix Aug 9, 2018, 3:45 PM

#

I need to practice leading them out more...

restive fjord Aug 9, 2018, 3:46 PM

#

Yeah you're leading them with a lantern and it's dark outside.

#

You've been this way before and know there's a root you trip over, right there, so avoid it.

#

(then you trip on a stone next to it and we all laugh together)

languid phoenix Aug 9, 2018, 3:47 PM

#

Lol

restive fjord Aug 9, 2018, 3:48 PM

#

I got a really good one for you. (looks it up)

#

http://www.feynmanlectures.caltech.edu/I_91.html

#

This is Richard Feynman, talking about the approach he took with his lectures and where he thought he did well, and where he thought he needed to improve, and why. And about the students and what they are like to work with.

#

He's about as honest as Hemingway was, and that's saying quite a lot. ;)

languid phoenix Aug 9, 2018, 3:50 PM

#

Thanks! I'll need to listen to that.

restive fjord Aug 9, 2018, 3:51 PM

#

(it's to read; there ain't no audio ;)

languid phoenix Aug 9, 2018, 3:51 PM

#

My bad

restive fjord Aug 9, 2018, 3:51 PM

#

He's really good at this.

#

(but didn't know it)

#

In the olden days the author pretty much always opens a new book with their apology haha.

languid phoenix Aug 9, 2018, 3:52 PM

#

Wonder how much of that is just convention.

restive fjord Aug 9, 2018, 3:53 PM

#

Oh I think the urge to apologize is endemic and to ..somehow .. not give into it is the hard part.

languid phoenix Aug 9, 2018, 3:53 PM

#

Culture of false modesty perhaps?

#

Not wanting to seem too arrogant or prideful?

restive fjord Aug 9, 2018, 3:54 PM

#

Nothing false about it. The expressed language forms, those are a bit staid, but the deep of it is authentic. People are really very not sure they have the right to speak on things with any authority.

languid phoenix Aug 9, 2018, 3:58 PM

#

It seems that there's two ways to be considered an authority or someone who can speak with authority, self declared, or declared by your peers. I could see there being an aversion to self declaring because it can smack of arrogance. To be declared an authority by your peers takes a lot longer, but can give a more solid reputation to your authority

restive fjord Aug 9, 2018, 3:59 PM

#

I think everyone (pretty much or maybe 7 in 10 of everyone) should just take the chance and stand on what they know (But not on their ignorance).

#

The thing that floors me about what we see from Adafruit is the improbable characters who turn out to be .. pretty darned authorative. And very young, so.

#

I'd say 61 % of that is willingness to not be taken as authorative .. to take the risk of being undermined.

languid phoenix Aug 9, 2018, 4:01 PM

#

To take the risk of being wrong

restive fjord Aug 9, 2018, 4:01 PM

#

Or just not all that effective. ;)

#

The effect upon the mind .. upon the brain physiology; chemistry .. of going a bit too far .. and having it work out quite nicely .. is such an eye opener.

#

It's an experience people should have.

languid phoenix Aug 9, 2018, 4:04 PM

#

Agreed.

restive fjord Aug 9, 2018, 4:04 PM

#

oh these hakko wire strippers for wire wrap wire are really worth it.

#

I got my goodie bag via UPS a few minutes ago. Two scope probes in there, too. ;)

#

and the segger jlink EDU. ;)

languid phoenix Aug 9, 2018, 4:05 PM

#

Nice! It's always fun to get new toys.

restive fjord Aug 9, 2018, 4:06 PM

#

Well I'm like 58 years old now and it's my time. I don't know how much older I'm going to get, so I'm getting my toys.

languid phoenix Aug 9, 2018, 4:08 PM

#

I've never had the chance to do wire wrap. Seen a lot of it online, but solderless breadboards and perfboard are so common these days.

restive fjord Aug 9, 2018, 4:08 PM

#

too bad a 100 MHz scope probe does not make a 25 MHz scope into a 100 MHz scope ;)

#

I don't have any wire wrap 'posts' to wrap onto but the stuff has other uses similar to that use.

#

You can improvise in ways that breadboards frustrate. I use the 0.1" extra long headers extensively for this as well.

#

The test points work really well with 30 AWG wire. ;)
http://adafu.it/3824

languid phoenix Aug 9, 2018, 4:10 PM

#

So you use 0.1" Header pins as sort of wire wrap posts?

restive fjord Aug 9, 2018, 4:11 PM

#

Um yeah sort of.

languid phoenix Aug 9, 2018, 4:11 PM

#

That's really clever! I like it!

restive fjord Aug 9, 2018, 4:11 PM

#

What I do is allow adjacent pin room -- no shorts so I skip a pin.

#

I coil the wire by hand around the 0.1" header pin, still on its carrier with 4 others.

#

The turns are spaced about 5 wire diameters apart or more.

#

But they're tight.

#

Then I just shove the whole thing into the breadboard.

#

It just kind of squishes the loose spiral together.

languid phoenix Aug 9, 2018, 4:13 PM

#

Yeah, but it's a solid connection.

restive fjord Aug 9, 2018, 4:13 PM

#

I'm not running into problems yet and I am dealing with < 2mA current and up to 2 MHz signal 'intensities' (data rates).

#

Probably closer to 750 kHz.

#

I'm looking for low noise low current higher data rate connections for communications (SPI, i2c, UART style SERCOMs)

#

The test points are working well also.

#

I bolt the target board (Feather M4 Express) on standoffs to perfboard, for mechanical stability.

#

Then I just push the bare 30 AWG up through the bottom, and pin it down with the test point (sharing the same through-plated hole as the test point).

#

The rest is a form of 'sewing' and 'stitching' with the 30 AWG.

#

That fixes it mechanically so the tenuous connection at the test point isn't disturbed by tugging on it.

languid phoenix Aug 9, 2018, 4:17 PM

#

I like it!

restive fjord Aug 9, 2018, 4:18 PM

#

So I make sure when you pull on the bare wire (already poked through, from below) is already binding by the interweaving done to that wire in the perfboard holes.

#

No solder. ;)

languid phoenix Aug 9, 2018, 4:18 PM

#

None at all

restive fjord Aug 9, 2018, 4:19 PM

#

I can't solder anymore unless I had a lab which I do not have.

#

I do have photos but I can't find them.

languid phoenix Aug 9, 2018, 4:20 PM

#

Why can't you solder? Fumes?

restive fjord Aug 9, 2018, 4:20 PM

#

fumes. it gets into my breathing gear. sleep apnea machine.

#

cpap

languid phoenix Aug 9, 2018, 4:20 PM

#

Ah yes, definitely not what you want.

restive fjord Aug 9, 2018, 4:21 PM

#

it's too much trouble in a living situation with food safety and all that.

languid phoenix Aug 9, 2018, 4:21 PM

#

Agreed

restive fjord Aug 9, 2018, 4:21 PM

#

there's been talk lately about a lab but I think that's pie in the sky. ;)

languid phoenix Aug 9, 2018, 4:22 PM

#

Not necessarily

restive fjord Aug 9, 2018, 4:23 PM

#

I don't remember ever having a little grounding clip for a scope probe.

#

They all have them, now.

languid phoenix Aug 9, 2018, 4:23 PM

#

They do. I've never used one without a grounding clip.

restive fjord Aug 9, 2018, 4:24 PM

#

I think we used to be okay with a separate long test lead to connect the ground.

#

(or maybe just float hehe I don't remember)

#

I always use an isolation transformer to power the scope anyway.

#

But the little grounding clip is proving restrictive of motion.

#

I think I'll get a second pair of probes just so I can be rid of the grounding clip by disconnecting at the BNC bayonnette end of things ;)

#

wow no double consonants at all: bayonet

#

(I usually add all possibles in, then subtract the ones not required) ;)

languid phoenix Aug 9, 2018, 4:27 PM

#

Lol, that's one way to do it. I wonder why they all do ground clips these days.

restive fjord Aug 9, 2018, 4:28 PM

#

http://hparchive.com/Manuals/HP-122A-AR-Manual.pdf

#

I'm going to check myself on this one.

primal warren Aug 9, 2018, 4:29 PM

#

To measure higher frequency signals correctly, you need the ground clips. For low frequency (audio) stuff, it doesn't matter as much.

restive fjord Aug 9, 2018, 4:30 PM

#

200 kHz and we were darned glad to have it

#

Says it has differential amps on the inputs.

primal warren Aug 9, 2018, 4:32 PM

#

Nicer probes have a choice of removeable ground clips.

restive fjord Aug 9, 2018, 4:33 PM

#

CRT phosphors: specify by Phosphor Number (P-31 standard, P2, P7 with amber filter, P11 available, no charge).

#

!

#

@primal warren aha! we had the good probes (and lots of users handling them)

#

so I never even saw any clips. ;)

primal warren Aug 9, 2018, 4:35 PM

#

Removeable ground clip: for even higher frequencies, other methods are used.

📎 screenshot_218.png

restive fjord Aug 9, 2018, 4:37 PM

#

@silent jolt

📎 scope_of_manual.png

#

Mine are removeable but I don't want to remove then unremove them.

#

like it ruins the spring

#

So, two sets, one with the clips removed!

#

I have not yet found the chopped waveform setting on this (OWON PDS5022S 25 MHz) oscilloscope.

primal warren Aug 9, 2018, 4:42 PM

#

There are a mindbending variety of accessories available for really high frequency oscilloscoping.

📎 screenshot_219.png

restive fjord Aug 9, 2018, 4:42 PM

#

Until I saw the HP manual I'd forgotten how ubiquitous the chopped setting was.

primal warren Aug 9, 2018, 4:43 PM

#

I got spoiled by dual-beam scopes, so I didn't use the chopped or alternate settings much. Then again, Jim Williams would use a pair of 4-channel plugins on a dual-beam scope to get 8 traces at once.

restive fjord Aug 9, 2018, 4:43 PM

#

@primal warren is it hopeless to do anything with a 25 MHz oscilloscope in the way of verifying a 120 MHz signal is present/absent?

#

The scope sees it as 20 MHz.

#

Sometimes the frequency counter reads 48 MHz which is also possible.

#

(I may actually be generating 48 MHz not 120 MHz)

primal warren Aug 9, 2018, 4:44 PM

#

Depends on the scope. An analog scope rolls off at 3dB/octave above its frequency, so you can view higher frequencies, but they'll be attenuated commensurately, and their harmonics even more so (so a square wave may look like a sine wave).

#

However, it sounds like you're using a digital scope, so the situation is trickier.

restive fjord Aug 9, 2018, 4:45 PM

#

My shortwave portable radio (AM only; aircraft band) is quite sure of itself that there is a 120 MHz signal coming from theSAMD51.

primal warren Aug 9, 2018, 4:45 PM

#

It certainly sounds like something is present, but it's hard to tell what the frequency actually is.

restive fjord Aug 9, 2018, 4:46 PM

#

Yeah I agree.

#

I think it could be the sixth multiple of x to get to 120 (20 in this case)

#

20 x 6 = 120

#

I don't know if physics allows such 'harmonic' radiation.

primal warren Aug 9, 2018, 4:47 PM

#

Yeah, a waveform with steep rise/fall will generate a bunch of harmonics, so a 20MHz square wave will generally give a 120MHz signal too. Sounds like a job for a faster scope, a spectrum analyzer, or a frequency counter.

restive fjord Aug 9, 2018, 4:47 PM

#

The code's author claims it was verified at 120 MHz.

#

I'm unconvinced and can't think of a computer program that can do some convincin'.

primal warren Aug 9, 2018, 4:48 PM

#

Absent all of those, I'd bodge up a divider chain out of flip-flops and monitor the divided-down output.

restive fjord Aug 9, 2018, 4:48 PM

#

ah right. thanks. that's the right thing here.

#

I think it's a weak signal though.

#

May not trigger a flip flop as i think it's intentionally weak on that pin.

#

That's not a complete program (unless it is, by accident)

#

~~should run but I haven't tested itt.~~ I have tested itt.

restive fjord Aug 9, 2018, 6:51 PM

#

Correct program for the D.U.T.
Says 120 MHz but I don't have proof here (no fast scope or freq. counter/what-have-you).

📎 clock_init_120_demo_raw.c

#

Cooked version - has LED blinkie included.

📎 clock_init_120_demo_cooked.c

#

reference board is Feather M4 Express btw. ;)

#

D4 seemed the best choice for putting out the oscillator on an external pin.
D11 is an arbitrary GPIO (choose any). D13 is our usual LED friend.

#

I didn't see a lot of signal on D4.

primal warren Aug 9, 2018, 8:50 PM

#

Tempted to try it, even though my fast scope is currently out of action. Do I need Atmel Studio to compile this?

restive fjord Aug 9, 2018, 9:33 PM

#

*http://start.atmel.com/*

#

it's an online gui configuratororor. @primal warren

#

I have screenshots.

#

https://github.com/wa1tnr/ainsuForth-d51/tree/clockey-amrt-aa-8_AUG-d-
base directory and recent branch

#

it's probably a mess.

#

https://github.com/wa1tnr/ainsuForth-d51/blob/clockey-amrt-aa-8_AUG-d-/0-Distrib.d/clockey-amrt-aa-8AUG18.atzip

#

The build environment is the same as for Circuit Python in that you need the ARM gcc to compile this.

#

https://learn.adafruit.com/building-circuitpython explains that process in some detail.

#

This compiler may be the whole shootin' match, not sure:
https://developer.arm.com/open-source/gnu-toolchain/gnu-rm/downloads

#

It's part of it (the only part I care about, probably).

#

To install I use the arduino version of bossac and a Makefile target for a shell script that runs bossac with the correct parameters.

#

(for SAMD51 the older bossac didn't work, iirc)

#

We're talking about programs of less than 2kb stored in a .bin file, by the way. ;)

#

Same program in Arduino is about 11x that.

#

Maybe less. Haven't done an empty Arduino compile in a while.

#

Looks like Arduino doesn't run the clock as fast as possible.

restive fjord Aug 9, 2018, 11:54 PM

#

I've got the SAMD51 sending me five letters via the (bitbanged) UART pins (ignoring the UART itself for the moment).

#

I have a space for a separator, too, so that I can send small coded groups.

#

Basically 26 uSec for each element at 38400 bps

#

Every sequence begins and ends with a cleared bit (GPIO to ground).

#

Sandwiched in between are the bits for 2^0 through 2^6. 3.3 volts is a set bit (a 1 vs a 0 of the cleared bit).

#

keep the line at 3.3V when not sending anything. That's about it.

#

LSB goes out first.

#

MSB sent last.

#

(least and most significant bit, respectively).

#

So the start and stop bits are zero's.

#

(which differentiates them from the ongoing 'drone' of all 11111s which is just the line held indefinitely at 3.3 V level).

#

No event marks two adjacent bits with the same value.

#

Just keep the line held at that level (3.3V or 0 V).

#

And time it out -- use strict time intervals as the 'marker' of an event.

#

I would guess the listening end has to be forgiving, and make a best guess about the intent of the sender, if things are timed a bit sloppy.

#

It only has to pay attention for that brief moment; there's probably a minimum guard time between characters to make sure each data frame has its own reference point (with the start bit sent at any unexpected time).

#

This is what was typed by the SAMD51 as I wrote that, just now.

📎 allworkandnoplaymakesjackadullboy.png

#

I did not plan that. I just had it recite over and over what it knew. ;)

#

(eh they're not taking over, just yet!)

#

The oscilloscope helped a lot.

#

2096 bytes ;) That's the size of the binary uploaded to theSAMD51.

restive fjord Aug 11, 2018, 1:57 AM

#

After proving the technique on a dozen characters in the alphabet, I found an algorithmic method to take the ASCII value and pretty much send it right out the (ersatz) serial port to the PiUART.

restive fjord Aug 11, 2018, 2:19 AM

#

void blitwbits(unsigned int x) {
    // accepts ASCII value as input 'x'
    send_preamble();
    for(int i = 0; i < 7; i++) {
       (x & (1u << i)) ?  bit_SET() : bit_CLR();
    }   
    common_end();
}

#

Preamble and common_end are nothing more than the common parts to all characters.

#

bit_CLR() is the essence of send_preamble().
bit_CLR() bit_SET() is the intent of common_end().

#

The details involve making adjustments to timing (so that things don't happen so fast that the PiUART decides it doesn't understand what's going on anymore).

#

What was interesting to me was the complete lack of traffic on the RX pin of the TX/RX UART pair (here used dumbly as simple GPIO pins).

#

As the (immediate) aim was to get messages out of the SAMD51 I wasn't interested in hearing from the PiUART (as to accept keystrokes on a command line, for example).

#

I think I had a hidden assumption about 'handshaking' that just didn't apply here.

pliant garnet Aug 17, 2018, 12:55 PM

#

On an amateur radio, specifically a VHF/UHF radio, what filters are on-board?

restive fjord Aug 17, 2018, 4:46 PM

#

I don't know that they need them -- they're factory filters if they exist at all (besides a high 'Q' circuit, here and there).

#

OTOH if you're experiencing near field stuff that's unwanted (ever hear a pager come over a VHF repeater, unintended?) then it might be beneficial to reject such signals.

#

'adjacent channel' stuff (rejection I think) comes to mind. I haven't read any sales literature, lately.

#

If you're on a crowded band and there are strong signals nearby (not on your frequency, but close to it) then some filtering might be desirable.

#

There are all-mode VHF rigs, and probably UHF rigs as well. CW is CW; if it's crowded at all, then filters are a good idea.

pliant garnet Aug 19, 2018, 3:43 AM

#

In a power supply, why is there a filter followed by a regulator?

primal warren Aug 20, 2018, 5:03 PM

#

The classic lineup is transformer, rectifier, filter, regulator. The output from the rectifier is just half the AC waveform (so it's zero half the time) for half wave rectification, or the AC waveform with the negative peaks flipped to positive (so it's zero between peaks, same as AC) for full wave rectification.

#

Because of this, the regulator would have zero volts input part of the time, which won't work: the input voltage to the regulator generally has to be positive, and for a linear regulator, above the output voltage. The filter accomplishes this by storing voltage to ride through the zero periods from the rectifier. There will still be "ripple", but it won't go down to zero any more, and the regulator will vastly reduce the ripple, giving a clean output.

pliant garnet Aug 21, 2018, 12:12 AM

#

So the filter helps with maintaining that 'straight' DC output as demonstrated by a straight line on an oscilloscope?

#

And then the regulator reduces that voltage to a lower output voltage?

primal warren Aug 21, 2018, 2:37 AM

#

Right, and smooths it out further.

pliant garnet Aug 21, 2018, 3:25 AM

#

So without the filter, the voltage won't be a straight DC voltage and there will be some instability?

primal warren Aug 21, 2018, 3:26 AM

#

Yeah, you'll get a lot of buzz (if it's an audio circuit), and possibly even squegging.

pliant garnet Aug 23, 2018, 11:31 AM

#

A little while ago we were talking about modulation like SSB, FM and AM and a little bit of CW. What is PM?

cold viper Aug 23, 2018, 1:20 PM

#

I've been having an issue with the RFM 69 breakout board from adafruit for a long time now (link to it is here: https://learn.adafruit.com/adafruit-rfm69hcw-and-rfm96-rfm95-rfm98-lora-packet-padio-breakouts/overview). I already asked this question in several forums but didnt get an answer so Im coming here now.

Im following the guide series at the link, my wiring, code and library is equal to what youve got there. THe only difference is that I am using an arduino Nano instead of the board shown in this wiring diagram I am following: https://learn.adafruit.com/assets/40615. However the init function from the radio head library keeps failing no matter what i do. I exchanged the whole hardware rewired everything multiple times nothing helps. After messing a bit around with the library code i found out that it fails at the part where it is supposed to get the device version, namely:

    // Get the device type and check it
    // This also tests whether we are really connected to a device
    // My test devices return 0x24
    Serial.println("Obtaining device number");
    _deviceType = spiRead(RH_RF69_REG_10_VERSION);
    if (_deviceType == 00 || _deviceType == 0xff) {
        Serial.println("No valid device type found, is a device connected?");
        //return false;
    }
    Serial.println("Done");

(serial prints are by me)
Which as you can read in the comments means that there is either no device connected or the board is just not returning a valid one. If anybody knows a thing i could do or knows a person who might be able to help me, please tell me

primal warren Aug 23, 2018, 2:07 PM

#

I believe that's the initial transaction and it seems to be failing, so I'm guessing an SPI problem. That in turn can be caused by a few different things, such as clock frequency issues, pin contention, initialization, and (of course) wiring goofs.

#

Could also be a voltage issue (is the Nano a 3V device?) or a hardware issue. Figuring out what's really going on can be a bit of a slog, doing a bunch of tedious checks on things. So I'll lead off with some obvious-sounding questions: do you have a multimeter? An oscilloscope? Another SPI device? Another CPU? If you don't have any of these things, you can do some basic testing with just an LED, resistor, breadboard, and some jumper wires.

cold viper Aug 23, 2018, 2:54 PM

#

ok ill try that, thanks you are the first one answering after like a week

burnt kite Aug 23, 2018, 3:17 PM

#

@cold viper also, check your wire length and baudrate. During the CircuitPython library work, it was discovered that longer wires weren't handling higher baudrates.

cold viper Aug 23, 2018, 3:18 PM

#

im already using the shortest wires i got so i hope thats not the problem

burnt kite Aug 23, 2018, 3:20 PM

#

Just a mention for a possible "try". 1-10kHz worked at times that 100kHz didn't.

primal warren Aug 23, 2018, 3:35 PM

#

In my experience, there are a bunch of helpful people on this discord, with a variety of knowledge and experience. However, people are often busy, so replies may take hours or days, but generally you'll get good help here. I am all too familiar with the frustration of fighting the same issue for a long time (my time circuits bug had vexed me for over a year), and asking for help and getting no answers, so I try to at least check in here from time to time in case there's something I can offer.

cold viper Aug 23, 2018, 3:36 PM

#

i just tried with another SPI device (namely a SD Card adapter) and it worked perfectly no spi issues whatsoever

and thanks for that tip

primal warren Aug 23, 2018, 3:37 PM

#

Ah, that's useful information! That tells us that the issue (likely) isn't with the SPI port, pins, or support. That means we can skip all the testing for pin mode, I/O pin driver health, etc.

#

My usual response to debugging issues is "divide and conquer", which usually boils down to trying to eliminate possibilities and then concentrate on what's left.

cold viper Aug 23, 2018, 3:39 PM

#

what are the possibilities i should try to eliminate here in your opinion?

primal warren Aug 23, 2018, 3:42 PM

#

I don't think it's a voltage issue any more (I wasn't familiar with this breakout, so I read up on it, and it includes support for 3 and 5V logic, so it shouldn't matter which voltage your Nano uses.

#

You're using the Radiohead library, which is pretty solid.

cold viper Aug 23, 2018, 3:43 PM

#

yeah, even the fork by adafruit

primal warren Aug 23, 2018, 3:45 PM

#

At this point, I'm thinking either a low-level SPI issue, a quirk with your Nano, a wiring goof, or a problem with the LoRa breakout itself.

cold viper Aug 23, 2018, 3:46 PM

#

i actually got two nanos here and i tried with both

#

so i doubt its the nano

primal warren Aug 23, 2018, 3:47 PM

#

Not a hardware failure with the Nano, but perhaps a pin mapping issue, so the pins you're using for reset or chip select might not work quite the same way, or the voltage you're using for power isn't the same as the I/O voltage.

cold viper Aug 23, 2018, 3:48 PM

#

I guess the wiring goof issue is the easiest to check?

primal warren Aug 23, 2018, 3:49 PM

#

Your SD card test (a great idea, by the way) pretty much tells us that SS, SCK, MISO, and MOSI are okay, but that still leaves the reset pin or a voltage issue. I'd probably eyeball those first.

#

Doing a low-level test of SPI communication is what I'd try next, but it's a pain (basically emulate SPI by bit-banging the I/O pins and printing out what happens).

cold viper Aug 23, 2018, 3:50 PM

#

uuuuhm

#

ill get to that once i checked wiring and the reset ping voltage

primal warren Aug 23, 2018, 3:51 PM

#

Looks like you can just try disconnecting reset, it should be pulled high by the board

#

Also, if you're running a 3V Nano, you'll also want to run 3V to the Vin pin (not 5V like in the diagram)

cold viper Aug 23, 2018, 3:52 PM

#

nah its a 5V nano

primal warren Aug 23, 2018, 3:53 PM

#

Okay, not an I/O voltage issue.

cold viper Aug 23, 2018, 3:58 PM

#

yeah so just wired it again and flashed the code, still init failed

#

so not i wiriing problem i hope

primal warren Aug 23, 2018, 4:02 PM

#

Hmm, probably time to look at the software. What does your init line look like? Mine looks like this: ```#define RFM95_CS 8
#define RFM95_INT 3

static RH_RF95 rf95(RFM95_CS, RFM95_INT);

cold viper Aug 23, 2018, 4:03 PM

#

#define RF69_FREQ 433.0
// Singleton instance of the radio driver
#define RFM69_INT     3  
#define RFM69_CS      4  
#define RFM69_RST     2 
#define LED           LED_BUILTIN

RH_RF69 rf69(RFM69_CS, RFM69_INT);

int16_t packetnum = 0;  // packet counter, we increment per xmission

void setup() 
{
  Serial.begin(115200);
  //while (!Serial) { delay(1); } // wait until serial console is open, remove if not tethered to computer

  pinMode(LED, OUTPUT);     
  pinMode(RFM69_RST, OUTPUT);
  digitalWrite(RFM69_RST, LOW);

  Serial.println("Feather RFM69 TX Test!");
  Serial.println();

  // manual reset
  digitalWrite(RFM69_RST, HIGH);
  delay(10);
  digitalWrite(RFM69_RST, LOW);
  delay(10);
  
  if (!rf69.init()) {
    Serial.println("RFM69 radio init failed");
    while (1);
  }

#

just the code from them with other macros

primal warren Aug 23, 2018, 4:06 PM

#

Hmm, that looks wrong to me. I think the reset pin is active low, so that could would try to initialize it while it's being reset.

#

Try swapping LOW and HIGH in the lines where it's driving the RFM69_RST pin.

cold viper Aug 23, 2018, 4:07 PM

#

alright

#

uuuuh

#

nope

#

didnt help

#

same error

primal warren Aug 23, 2018, 4:08 PM

#

I figured it was worth a try.

cold viper Aug 23, 2018, 4:13 PM

#

so

#

the SPI low level stuff is next?

primal warren Aug 23, 2018, 4:16 PM

#

Yeah, I think that is next. I don't want to waste your time with untested code, and I don't have an implementation handy, so I figured I'd try to bash something together here to make sure I understand it and have working code, then share it with you for you to try.

#

I don't have a nano or that breakout board, so I figured I'd try it with an ordinary LoRa shield and Arduino, it's just some of the pin numbers might be different.

cold viper Aug 23, 2018, 4:19 PM

#

i assume I'll need some meassuring equipment for that SPI thing which i dont have here but i know somebody irl who i can ask for help with that. But thats not going to happen that soon so leave yourself time for testing your code and ping me when you got everything done you need ok?

primal warren Aug 23, 2018, 4:21 PM

#

Nope, this will be a software-only test. I'm going to put together a simple program to do an SPI transfer a bit at a time, and print out what it's sending and receiving.

cold viper Aug 23, 2018, 4:21 PM

#

oh thats great

primal warren Aug 23, 2018, 4:22 PM

#

This initialization step is pretty simple, it sends a single byte (RH_RF69_REG_10_VERSION) and receives a single byte that should be the device version.

cold viper Aug 23, 2018, 4:22 PM

#

so you'll just send that yourself and see what happens?

primal warren Aug 23, 2018, 4:23 PM

#

So I'm going to emulate just that transaction and see what happens. SPI is a funny protocol, it sends and receives a single bit with each clock. So what it's doing under the hood is sending 8 bits (and receiving 8 bits, which it ignores), then sending 8 more bits (which are ignored by the target) and receiving the 8 bits that are the reply.

#

So I'm going to do exactly that and yes, try it myself on this LoRa shield to verify that what I'm attempting is actually working.

primal warren Aug 23, 2018, 4:50 PM

#

Okay, got the code written and compiled, time to plug in the board and give it a go.

#

Sure enough, I get 0xff. Now to figure out why.

cold viper Aug 23, 2018, 5:06 PM

#

at least that means Im not alone with my issue

primal warren Aug 23, 2018, 5:13 PM

#

Realized I left out the select logic. Added that, getting slightly different results, but not the correct results. Now perusing the RFM69 data sheet looking for clues.

#

I may break out the oscilloscope to see what's really happening (I'm making a few assumptions on pinout, since I haven't used this board before).

primal warren Aug 23, 2018, 5:30 PM

#

I may have figured out what's going on. You may be using the wrong driver for your chip. The LoRa chip is the RFM96, the RFM69 is a different chip (an ISM transceiver).

#

It just so happens that the RFM96 uses register 10 as the RSSI threshold control register, which contains a default value of (yup!) 0xff. Whereas the RFM69 uses that register as to hold a chip version number, which is what the driver is looking for.

cold viper Aug 23, 2018, 5:32 PM

#

wait so the chip here infront of me is in fact an rfm96

#

not 69

primal warren Aug 23, 2018, 5:33 PM

#

But the code you pasted read RH_RF69 rf69(RFM69_CS, RFM69_INT);

cold viper Aug 23, 2018, 5:33 PM

#

yes

#

because i assumed im using a 69

#

all the time

#

and the subpage i got that code from says

📎 unknown.png

#

and the chips have RFM69HCW printed on them

primal warren Aug 23, 2018, 5:37 PM

#

Oh, how confusing. The chips have a common pinout, but the software required to run them is different, and I'm guessing the software you're using isn't the right software for the chip you have.

cold viper Aug 23, 2018, 5:38 PM

#

so the adafruit docs on their own chips are wrong?

#

i mean i can try the rfm96 very quick and see if that fixes it

primal warren Aug 23, 2018, 5:40 PM

#

Sort - of. It appears that there are different breakout boards. If you go to the "RFM9X Test" section, it shows how to use the 9x series chips.

#

That said, if you want to play with the code I wrote, here it is:

#

Software SPI test code

📎 lora-spi-test.ino

#

I'd suggest trying the RFM96, since that matches the markings on your chip (and you said it's a LoRa board: the 69 chips are a different kind of radio).

cold viper Aug 23, 2018, 5:44 PM

#

but for this project i must use the 69

#

its one of the requirements

primal warren Aug 23, 2018, 5:44 PM

#

In that case, you may have the wrong board 😦

cold viper Aug 23, 2018, 5:44 PM

#

why print rfm 69 on a not rfm 69 board

primal warren Aug 23, 2018, 5:46 PM

#

It could be because the two chips have a common pinout, so the same board serves both of them. You're saying the board itself is marked RFM69 and the chip is RF96?

cold viper Aug 23, 2018, 5:46 PM

#

no

#

the chip is supposed to be 69

#

thats why i am confused by your conclusion

#

like i ordered a 69 and now you are telling me it might be a 96

primal warren Aug 23, 2018, 5:47 PM

#

When you said "wait so the chip here infront of me is in fact an rfm96", that's what makes me think you've got the other chip.

cold viper Aug 23, 2018, 5:47 PM

#

oh

#

no its an rfm

#

at least it is supposed to be one

#

rfm 69

primal warren Aug 23, 2018, 5:48 PM

#

I'll admit to being somewhat confused.

cold viper Aug 23, 2018, 5:48 PM

#

yes same here

primal warren Aug 23, 2018, 5:48 PM

#

I wonder if I have an RFM69 lying around.

cold viper Aug 23, 2018, 5:49 PM

#

It just so happens that the RFM96 uses register 10 as the RSSI threshold control register, which contains a default value of (yup!) 0xff. Whereas the RFM69 uses that register as to hold a chip version number, which is what the driver is looking for. this made me think its an rfm96 because i read in the lib that a return value of 0xff would cause it do fail the init

primal warren Aug 23, 2018, 5:50 PM

#

Sorry about that.

cold viper Aug 23, 2018, 5:50 PM

#

so

#

its an rfm69 for sure

#

quite sure

#

📎 20180823_195914.jpg

primal warren Aug 23, 2018, 6:09 PM

#

After some rummaging around, I dug up a Moteino and an RFM69 Feather. I'll do some basic testing on that.

#

Okay, I can read the version register on the RFM69 with my software SPI code and get the expected value.

cold viper Aug 23, 2018, 6:20 PM

#

okay

#

in that case ill try that myself too. Im assuming the SS ping is for the chip select one on the rfm?

primal warren Aug 23, 2018, 6:25 PM

#

Yes, SS "slave select" is the SPI term for "chip select".

cold viper Aug 23, 2018, 6:25 PM

#

sending request for version
sending request 0x10
sending byte 0x10
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 1
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
received byte 0x0
getting response
sending byte 0x0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
received byte 0x0
got response 0x0
sending request 0x42
sending byte 0x42
sending bit 0
received bit 0
sending bit 1
received bit 1
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 1
received bit 1
sending bit 0
received bit 0
received byte 0x42
getting response
sending byte 0x0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
received byte 0x0
got response 0x0

im getting this with the code you sent

#

is that good?

primal warren Aug 23, 2018, 6:26 PM

#

That tells us some things. Sure enough, reading register 10 (the version register) does not return the expected result.

#

Interestingly, the read from register 0x42 gets 0x42 back, which sounds like a weird data coupling issue (a clue!)

cold viper Aug 23, 2018, 6:28 PM

#

\o/

primal warren Aug 23, 2018, 6:44 PM

#

Trying to figure out the best way to determine if MISO is being driven correctly. My thoughts so far are try grounding the CS input, and hooking an LED and resistor to MISO to see if it's getting driven.

cold viper Aug 23, 2018, 6:49 PM

#

so CS from arduino to ground and an LED plus resistor into the MISO line?

primal warren Aug 23, 2018, 6:50 PM

#

Disconnect CS from the Arduino and connect it to ground (don't ground the Arduino CS pin), then run the sketch again and see what happens.

cold viper Aug 23, 2018, 6:55 PM

#

alrighty

#

stuff has changed

#

software bitwise SPI test
sending request for version
sending request 0x10
sending byte 0x10
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 1
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
received byte 0x0
getting response
sending byte 0x0
sending bit 0
received bit 1
sending bit 0
received bit 1
sending bit 0
received bit 1
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 1
sending bit 0
received bit 1
sending bit 0
received bit 0
received byte 0xE6
got response 0xE6
sending request 0x42
sending byte 0x42
sending bit 0
received bit 0
sending bit 1
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 1
received bit 0
sending bit 0
received bit 0
received byte 0x0
getting response
sending byte 0x0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
sending bit 0
received bit 0
received byte 0x0
got response 0x0

primal warren Aug 23, 2018, 6:56 PM

#

Oho, the plot thickens! That looks like a response, but not really the expected one.

#

Maybe try uncommenting the #define CLK_ACTIVE_LOW line and running it again?

cold viper Aug 23, 2018, 6:58 PM

#

k

primal warren Aug 23, 2018, 6:59 PM

#

Maybe comment out the #define VERBOSE line if you don't want to see all the individual bits.

cold viper Aug 23, 2018, 6:59 PM

#

software bitwise SPI test
sending request for version
sending request 0x10
sending byte 0x10
received byte 0x0
getting response
sending byte 0x0
received byte 0x0
got response 0x0
sending request 0x42
sending byte 0x42
received byte 0x3
getting response
sending byte 0x0
received byte 0xC0
got response 0xC0

primal warren Aug 23, 2018, 7:00 PM

#

Try running it again and see if you get the same numbers.

cold viper Aug 23, 2018, 7:00 PM

#

sending request for version
sending request 0x10
sending byte 0x10
received byte 0x0
getting response
sending byte 0x0
received byte 0x0
got response 0x0
sending request 0x42
sending byte 0x42
received byte 0x0
getting response
sending byte 0x0
received byte 0x0
got response 0x0

nope

primal warren Aug 23, 2018, 7:02 PM

#

It seems to me like the MISO line isn't being driven, so it's just picking up random noise. Try disconnecting the reset wire entirely and running it again.

cold viper Aug 23, 2018, 7:03 PM

#

software bitwise SPI test
sending request for version
sending request 0x10
sending byte 0x10
received byte 0x0
getting response
sending byte 0x0
received byte 0x0
got response 0x0
sending request 0x42
sending byte 0x42
received byte 0x0
getting response
sending byte 0x0
received byte 0x0
got response 0x0

just 0s

primal warren Aug 23, 2018, 7:04 PM

#

It's really starting to look like you have a broken module 😦

cold viper Aug 23, 2018, 7:05 PM

#

that is shocking

#

in fact my friend ordered these two (we had others before which didnt work either) just to verify that we dont have broken modules

#

they are barely a week old

primal warren Aug 23, 2018, 7:08 PM

#

Hm. Seems unlikely then. Well, let's keep going, shall we?

cold viper Aug 23, 2018, 7:08 PM

#

sure

primal warren Aug 23, 2018, 7:08 PM

#

Try it again with the reset lead connected directly to +5V (run it a couple of times to see if the numbers change), then again with it connected to ground (again, not connected to the Arduino).

cold viper Aug 23, 2018, 7:11 PM

#

1st ground all 0 2nd same 1st 5V 0 2nd same

#

or

#

should i rewire all that miso and CS stuff first?

primal warren Aug 23, 2018, 7:13 PM

#

Perhaps time to try the LED. I'd like to try the LED two ways, once to +5V (to see if MISO is pulling down at all) and once to ground (to see if MISO is pulling up at all).

cold viper Aug 23, 2018, 7:16 PM

#

so ill just leave the wiring as is now and put an LED in the 5V and the GND line (just want to be sure im not doing things wrong as im not particulary into electronics)

primal warren Aug 23, 2018, 7:17 PM

#

Right. The wiring changes are to eliminate other possibilities.

#

So one lashup is +5 to a resistor to the LED anode, and connect the LED cathode to the MISO pin.

#

The other way has the MISO pin to the resistor, to the LED anode, and the LED cathode to ground.

cold viper Aug 23, 2018, 7:18 PM

#

ok

#

and what do i expect to see then

primal warren Aug 23, 2018, 7:19 PM

#

Then run the sketch a few times, peering at the LED. It may stay on or just flicker briefly when the data is flowing (you may need to dim the lights to see it, or slow down the code).

cold viper Aug 23, 2018, 7:28 PM

#

ok so

#

its permanently on for +5 to resistor to led to miso

primal warren Aug 23, 2018, 7:29 PM

#

Interesting, so it is being driven. Does that change any if you change the reset line?

cold viper Aug 23, 2018, 7:30 PM

#

and for the gnd thing short flicker

#

do you mean if i take the 5v away from rst?

primal warren Aug 23, 2018, 7:31 PM

#

Yeah, but the short flicker indicates it's sending data (which is what we want).

cold viper Aug 23, 2018, 7:32 PM

#

ok and

#

if i remove rst it still stays on for the +5v

#

so its sending data

#

does that mean we can be sure its not broken?

primal warren Aug 23, 2018, 7:36 PM

#

Not for certain, but signs are good. The trick now is to figure out why the data it's sending isn't getting picked up at the Arduino end.

cold viper Aug 23, 2018, 7:40 PM

#

alright

#

sounds good, what do i have to do for that

primal warren Aug 23, 2018, 7:42 PM

#

Remove the LED and resistor, and disconnect the MISO wire from the RFM board to the Arduino, and connect that Arduino pin to +5V and run the sketch. The expected result is to get 0xff for all the replies.

cold viper Aug 23, 2018, 7:44 PM

#

yup

#

getting that

primal warren Aug 23, 2018, 7:44 PM

#

So the board can send and the Arduino can receive, why doesn't this work?

#

We've eliminated most of the possible variables by now.

#

Go ahead and disconnect the wire from 5V and reconnect it to the MISO output of the RFM board, then go back into the code, comment out the #define CLK_ACTIVE_LOW line again, and re-run it.

cold viper Aug 23, 2018, 7:47 PM

#

nothing

#

0x0

primal warren Aug 23, 2018, 7:49 PM

#

I'll admit it: I'm stumped here. Everything I can think of checks out. But it's not working.

#

Any chance you're in the Baltimore-Washington area? It's about time to break out an oscilloscope.

cold viper Aug 23, 2018, 7:51 PM

#

germany munich

#

😐

#

but

#

i know somebody with an oscilloscope

primal warren Aug 23, 2018, 7:52 PM

#

Oh, maybe we can get them to bring it over (or you could bring your setup to their place).

cold viper Aug 23, 2018, 7:53 PM

#

yeah but not today i think

#

its 10 pm here

primal warren Aug 23, 2018, 7:53 PM

#

Right. We've been banging on this for a while.

cold viper Aug 23, 2018, 7:53 PM

#

yup

#

just to be sure i did everything correct
your macros

// LoRa low level SPI test

//#define DRAGINO
#define RST_ACTIVE_LOW
//#define CLK_ACTIVE_LOW
//#define VERBOSE

#define RFM69_MOSI  11
#define RFM69_MISO  12
#define RFM69_CLK   13
#define RFM69_SS     4
#define RFM69_RST    2

#ifdef RST_ACTIVE_LOW
#define RST_ASSERT    LOW
#define RST_DEASSERT  HIGH
#else
#define RST_ASSERT    HIGH
#define RST_DEASSERT  LOW
#endif

#ifdef CLK_ACTIVE_LOW
#define CLK_ASSERT    LOW
#define CLK_DEASSERT  HIGH
#else
#define CLK_ASSERT    HIGH
#define CLK_DEASSERT  LOW
#endif

#define RH_RF69_REG_10_VERSION 0x10

and my pins from arduino to rfm

5V -> VIN
GND -> GND
D3 -> G0
D13 -> SCK
D12 -> MISO
D11 -> MOSI
GND -> RST

primal warren Aug 23, 2018, 8:00 PM

#

Looks right to me. And the fact that you can talk to an SD card tells me that SPI basically works.

cold viper Aug 23, 2018, 8:01 PM

#

sigh

#

@primal warren earliest i can get access to an osci is next week so we'll have to wait until that

primal warren Aug 23, 2018, 8:14 PM

#

Sorry about that. Usually these problems aren't so fraught with mysterious complications.

#

I may take an oscilloscope to mine and see if I can glean any more details about how it's supposed to work.

cold viper Aug 24, 2018, 4:07 PM

#

@primal warren unfortunately the guy with the oscilloscopes has only got time during our morning which means night for you so I guess you couldn't really tell us what to do then. The only solution I see for this (tell me if you've got a better idea) is that you write up what we should do and I tell you about the results later the day

primal warren Aug 24, 2018, 9:00 PM

#

Perhaps I'll make some measurement of mine and send you the pictures, so you can compare what yours looks like.

silver marten Aug 26, 2018, 3:37 PM

#

@pliant garnet that is phase modulatution

pliant garnet Aug 26, 2018, 11:46 PM

#

Is is just another modulation for amateur radio? Is it even used commonly?

silver marten Aug 28, 2018, 12:27 AM

#

@pliant garnet I don't know if it is used much except maybe in digital modes

pliant garnet Aug 28, 2018, 1:52 AM

#

Ohh, ok, thank you for your help.

restive fjord Aug 30, 2018, 2:39 AM

#

@pliant garnet the wikipedia article on the UART says simplex is one-way communication without expectation of a reply.
It says sharing a channel and taking turns transmitting and receiving is half-duplex; talking at the same time (both talking, both listening) is duplex (or full duplex).

pliant garnet Aug 30, 2018, 2:39 AM

#

Ohh, isn't simplex radio to radio only?

#

And not through repeaters?

restive fjord Aug 30, 2018, 2:41 AM

#

In common use, simplex on 2 meter FM is direct communication on a single frequency -- requiring push to talk in order to hear the other person (can't just key down forever).

pliant garnet Aug 30, 2018, 2:42 AM

#

Right.

restive fjord Aug 30, 2018, 2:42 AM

#

push to talk could be considered half-duplex according to that USART article.

#

Also to muddy the waters further, there are antenna systems that allow unexpected things to occur.

#

duplexer is a term you will hear.

#

I've put snubbers (diodes) back to back across a receiver's antenna, so that I could listen on that frequency while transmitting on it. ;)

#

That requires a separate antenna to do that -- either that or (on some higher bands) the duplexer I just mentioned can be used (or so I've heard).

#

I'm pretty sure that's how repeaters work -- the duplexer allows them to hear at the same instant that they transmit (thus allowing the repeater to amplify and retransmit the incoming signal at the same instant, rather than by some amount of delay).

#

Pretty sure they can use the same antenna, with the duplexer device. Which I don't quite understand, though I suppose I know the general ideas involved.

#

Repeaters on 2M FM usually are on 600 kHz splits (.19/.79 or .28/.88 are two examples).

pliant garnet Aug 30, 2018, 2:52 AM

#

Yeah.

restive fjord Aug 30, 2018, 2:52 AM

#

That 600 kHz separation from the repeater's input and output is probably enough relief to allow a good notch filter to operate correctly (to reject the very strong 'nearby' signal of that very repeater, by its receiver).

rapid hearth Aug 30, 2018, 3:43 AM

#

@restive fjord Hey, welcome back. I think people were asking about you and you may have been rigged for silent running for some time.

restive fjord Aug 31, 2018, 5:20 PM

#

@rapid hearth Thanks!

I'm never really gone. I have been coding constantly since my last presence here.

https://github.com/wa1tnr/ainsuMtxd51-exp

rapid hearth Aug 31, 2018, 6:04 PM

#

ah, genius at work. do not disturb.

silver marten Aug 31, 2018, 9:32 PM

#

@restive fjord I was the one asking about you. was wondering if you were ok

restive fjord Sep 4, 2018, 9:58 PM

#

computers are a specific variety of radio. ;)

blissful zodiac Sep 8, 2018, 1:11 PM

#

Does anyone have a favorite RF + Servo combination they recommend?

tiny dust Sep 8, 2018, 1:23 PM

#

anyone have good design techniques/tips on creating RF filters at HF ham radio bands (160,80,40,20, and possibly WARC) that require minimal if possible no manual tweaking at all

#

like, high tolerance SMD components, or at least slugged inductors and no crappy spinny air gap capacitors

restive fjord Sep 8, 2018, 9:32 PM

#

@blissful zodiac pretty sure that the Ham 50 MHz band (6 meters) was the frequency of choice for radio control of model airplanes with homebrew transmitters, in the early 1970's or just a bit before then.

#

With the advent of microelectronics (moving away from discrete components such as 3-leg transistors in a metal case) things have probably changed enough to make that recommendation outdated. ;)

#

Also fire departments were, I think, centered near 30 MHz (Ham 10 meter band would be comparable) for good propagation in the ranges they operated on. They moved up to 2 meters (145 MHz spectrum) and many are still there.

#

But that's with repeaters -- I don't know if they had 30 MHz repeaters!

silver marten Sep 10, 2018, 2:40 PM

#

@tiny dust i doubt you're going to find one with those requirements. At least I have never heard of any

tiny dust Sep 10, 2018, 2:41 PM

#

yeah

#

i decided to design a homebrew DDC radio

silver marten Sep 10, 2018, 2:43 PM

#

I am not familiar with DDC. Direct digital conversion radio?

tiny dust Sep 10, 2018, 2:43 PM

#

digital downconversion yeah

#

basically a really fast ADC

#

that's one way to not deal with fiddling with handmade inductors

silver marten Sep 10, 2018, 2:44 PM

#

Ah

subtle helm Sep 10, 2018, 4:13 PM

#

@restive fjord Around my area some of the bigger departments are changing over to 800MHz digital trunking systems now. The rurals are still running on 2 meter repeater. Good to see you around again! 73s, KI7LNA

restive fjord Sep 10, 2018, 5:01 PM

#

@subtle helm Yeah we have some apco25 (is that the name?) around here, but 145 MHz is usual in (fairly flat, low building) cities in Connecticut. 460 MHz (iirc) is highly utilized for business.
I hear 1.2 GHz traffic (iirc, been quite a while now) including some ambulance.

#

(wa1tnr)

#

I was also wn1tnr in 1974, and ran CW only at that time. ;)

#

I used to go to Civil Defense meetings; I think we worked a MARS net there. Tiny tiny shack by the town dump. When I wasn't in attendance, he ran the 'meeting' by himself -- nobody ever came by except other hams who never entered the building.
The one guy I remember who visited that way ran RTTY and I think, later, a packet node.

#

Civil Defense in the early 1970's was on its last legs in that area. I do not know what became of them; i almost never heard about it from anyone at all. I don't even remember how I became involved. Probably a simple word of mouth suggestion "Yanno, there's a CD meeting at the dump on Wednesdays".
Everything happens on a Wednesday, like that. Weird preference.

subtle helm Sep 10, 2018, 6:39 PM

#

Yeah, they also just call it P25, short for Project 25. Out here in Montana they tend to be closer to the 156 MHz range here. They've also got a internet linked P25 system statewide so you can talk back to your local dispatch talkgroup from anywhere in the state, pretty slick...until the internet goes out, then it's back to the Ham HF nets 😛

silver marten Sep 10, 2018, 7:49 PM

#

Has anyone tried FT8call yet?

subtle helm Sep 10, 2018, 10:03 PM

#

I haven't.

limpid citrus Sep 14, 2018, 8:44 PM

#

Does anyone know if instead of buying a LoRa gateway with multiple channels I can simply use another node as the receiver and give each sender an ID so it can differentiate nodes

normal drift Sep 14, 2018, 8:54 PM

#

@limpid citrus What radio's are you using and what Library? The Radiohead library protocol prepends a "header" that allows you to specify an "address" https://github.com/adafruit/RadioHead/blob/master/RH_RF95.cpp#L202

GitHub

adafruit/RadioHead

A github'ified version of http://www.airspayce.com/mikem/arduino/RadioHead/ - adafruit/RadioHead

limpid citrus Sep 14, 2018, 9:09 PM

#

I’m using the Feather LoRa modules made by adafruit with the RFM95, and the Radiohead library

#

Thanks that helps, just trying to save money because those gateways are expensive

rugged perch Sep 17, 2018, 6:04 PM

#

I am reviewing Tony DiCola's rfm_simpletest.py and notice on rfm69.send('xxx\r\n') that the text is delimited with a CR and Newline. I have read the docs and do not find a terminating delimiter. Is the /r/n required or just a nice to have?

normal drift Sep 17, 2018, 7:25 PM

#

@rugged perch The message is just string of bytes - the "/r/n" are just for the convenience of the receiving side.

#

I often jsut send "data" without any delimeters.

rugged perch Sep 17, 2018, 7:54 PM

#

@normal drift Thanks, I understand.

vocal veldt Sep 18, 2018, 9:08 PM

#

OK SO SNEAK PEEK

#

#help-with-radio things

#

gif too large 😦

#

📎 RFOPTIMIZED.gif

fallen matrix Sep 20, 2018, 3:15 PM

#

Is there anyone here who has experience with Nordic NRF24L01 modules? 🤔

fallen matrix Sep 21, 2018, 5:17 PM

#

I'll go ahead and start describing my plight.

#

I've invested in a few different versions of the NRF24L01 modules. And these are the three types I currently own.

#

📎 Screenshot_20180921-101620_Google.jpg 📎 Screenshot_20180921-101610_Google.jpg 📎 Screenshot_20180921-101543_Google.jpg

#

I've gotten a solid connection between two of the NRF24L01+ with the PCB antennas using them with a couple of Arduino nanno clones and the TMRH20 libraries, and the same result with the NRF24L01 PA LNA modules ( bottom picture)

#

But I've been having trouble getting the shielded module to work at all.

#

Has anyone ever messed with these before?

primal warren Sep 21, 2018, 5:57 PM

#

I haven't but I'm wondering if you might have RP-SMA antennae.

fallen matrix Sep 21, 2018, 7:52 PM

#

The antennae that I have are all male. Could that affect anything? @primal warren

#

They all seem to fit snug.

scenic cliff Sep 21, 2018, 8:10 PM

#

the issue could be RP-SMA vs SMA. Look into both connectors, if they both have a little hole and no pin then there is your problem.

#

@fallen matrix

fallen matrix Sep 21, 2018, 8:21 PM

#

Nah, the antennae are good to go.

primal warren Sep 22, 2018, 1:04 AM

#

Then I'd suspect a cracked solder joint.

craggy leaf Sep 25, 2018, 11:44 PM

#

recently got a plutoSDR to make a SDR transceiver for a future cubesat I'm helping build. i can ssh into it, but does anyone have any idea about where the power consumption of the ad9363 or arm processor on it? (Xilinx Zynq Z-7010)
The ad9363 is "/sys/bus/iio/devices/iio:device1" but there's nothing for power or current under there, only voltage. any ideas?

wheat pulsar Sep 26, 2018, 3:39 AM

#

Page 11: http://www.analog.com/media/en/technical-documentation/data-sheets/AD9363.pdf it's variable depending on the frequency/power of course, but looks like maxing out around ~660 for the chip alone. (not sure what external components may be drawing)

#

~660 mA*

vocal veldt Sep 26, 2018, 3:32 PM

#

📎 unknown.png

#

ok definitely some progress on the Pi LoRa stuff

#

that's a node form the learn guide I launched, talkin to a pi

primal warren Sep 26, 2018, 3:34 PM

#

Sweet!

vocal veldt Sep 26, 2018, 3:39 PM

#

thats a big deal for me, ngl

#

all I had to do was watch some videos about LoRaWAN. TLDR: The the protocol makes the frequency jump when joining, to get a better join req. The LMIC library also starts it on a subband, so it jumps between channels.

dapper valve Oct 2, 2018, 1:06 AM

#

so i remember someone saying that you can use a LoRa for about 20km away from each other, and they still talk to each other, which LoRa is that ? 😛

restive fjord Oct 2, 2018, 2:31 AM

#

That 20km figure would be pretty hard to achieve. ;)

burnt kite Oct 2, 2018, 3:14 AM

#

@dapper valve it definitely requires some power and a friendly environment...but it seems to hold up. [anecdotal] evidence: https://gridlocate.com/blog/lorawan-20km-range-test/

dapper valve Oct 2, 2018, 8:37 AM

#

Yeess thanks

normal drift Oct 2, 2018, 12:12 PM

#

“Your mileage may vary” 😉

vocal veldt Oct 2, 2018, 8:50 PM

#

@dapper valve It's more like 2->10Km. Certain antennas have claimed 20km

#

One guy (unconfirmed) pinged the adafruit factory router from governors island

#

should host a contest lol

dapper valve Oct 2, 2018, 8:58 PM

#

haha yeah 😛

wheat pulsar Oct 4, 2018, 12:45 AM

#

Picked up a HackRF a few weeks ago, anyone have any good recommendations for an antenna analyzer and/or a power meter? Wanting to learn more about RF stuff so wanted to set up a dummy load and test some RF circuits, but didn't want to go "too crazy" cost wise for this hobby. (got a million others ;D) Also want to see what these eBay antennas are tuned to, and if they're anything close to their "specs". 😃

#

Found this and looks tempting, especially for the cost: https://www.ebay.com/itm/New-83-5-4300MHz-Spectrum-Analyzer-W-RF-Signal-Source-For-WiFi-BLE-LT-GSM-GPRS/382426504414?_trkparms=aid%3D111001%26algo%3DREC.SEED%26ao%3D1%26asc%3D20160908131621%26meid%3Dbdc19fa34f4f49f4900fe59ff783a5aa%26pid%3D100678%26rk%3D1%26rkt%3D15%26sd%3D382426504414%26itm%3D382426504414&_trksid=p2481888.c100678.m3607&_trkparms=pageci%3Ac433f25b-c736-11e8-8734-74dbd1804d89|parentrq%3A3b1a3b461660aa16e054cfa5fffb0412|iid%3A1

eBay

New 83.5-4300MHz Spectrum Analyzer W/ RF Signal Source For WiFi BL...

3, Input: connect the RF input signal through SMA1 interface, SMA2 is a sweep local oscillator output signal, which can be used as signal source. RF signal source performance indicators. 1, the host net weight: 90g (using CNC aluminum alloy CNC shell);. | eBay!

arctic cloak Oct 4, 2018, 3:19 AM

#

you want a VNA like the AAI N1201SA or something from rigexpert

slender current Oct 4, 2018, 6:04 AM

#

@young cove @torpid echo looks my grandpa was W7HAV

pliant venture Oct 4, 2018, 1:53 PM

#

@slender current Oh neat, you found it.

slender current Oct 4, 2018, 3:50 PM

#

yup! this site has good resources http://w0is.com/oldcallsigns/oldcalls.html

restive fjord Oct 5, 2018, 12:16 AM

#

http://morsecode.world/m/eJxjYuLicg5UgCAXV4Vwcw_HMBTSGwB81QfQ

silver marten Oct 5, 2018, 12:18 AM

#

@slender current Go to the FCC ULS and check to see if that call is available. You might be able to get it as a vanity

restive fjord Oct 5, 2018, 12:23 AM

#

It's a club callsign, since before 2010. /deliberately_vague

#

This is a nice database interface:
http://callsign.ualr.edu/callsign.shtml

#

One time we were out for a drive in the family's car, and I had my mobile mag-mount antenna on the roof, working 2 meter repeaters in the area. I put my brother's wife on the air; she enjoyed it very much (this was before cell phones were in common use).

#

Years later she referred to it as Hand Radio

slender current Oct 5, 2018, 12:39 AM

#

@silver marten yup looks like its taken. I should probably get licensed and then try to convince them they should give it up

restive fjord Oct 5, 2018, 12:40 AM

#

That's unlikely -- look at the city name of the present holder.

slender current Oct 5, 2018, 12:40 AM

#

ya, havre

#

its a small town though 😃

restive fjord Oct 5, 2018, 12:40 AM

#

HAV havre get it

silver marten Oct 5, 2018, 12:40 AM

#

Lol

slender current Oct 5, 2018, 12:40 AM

#

plus, me getting licensed is a big if. I've got too many hobbies

restive fjord Oct 5, 2018, 12:41 AM

#

aw come on

#

:)

slender current Oct 5, 2018, 12:41 AM

#

if it was free I'd have reason to do it before it got taken

restive fjord Oct 5, 2018, 12:41 AM

#

you're a very nice person and ham radio can use nice people as on the air role models. ;)

silver marten Oct 5, 2018, 12:42 AM

#

Laurel vec tests are free.

Emcomm is a good thing. what are you going to do if the cell towers go out

restive fjord Oct 5, 2018, 12:44 AM

#

N7HAV has been available since 1995.

#

After the W's and K's were fairly well maxed out, the N calls became very prominent.

#

I think that's the only other prefix I know, come to think of it.

slender current Oct 5, 2018, 12:47 AM

#

he got it in the 30s I think so its kinda neat to have the special one

restive fjord Oct 5, 2018, 12:48 AM

#

I looked -- they assigned 1x2 calls like W7HB or whatever, and it took a while for them to circle around and add in 1x3 W7HAV type calls. Definitely before 1940. ;)

slender current Oct 5, 2018, 12:49 AM

#

😃

primal warren Oct 5, 2018, 2:22 AM

#

I didn't spend a lot of time getting licensed (I learned Morse while commuting). I eventually upgraded to Extra and celebrated with a vanity 2x1 call.

restive fjord Oct 5, 2018, 2:47 AM

#

OU812

vocal veldt Oct 11, 2018, 4:07 PM

#

small radio, long range :leek:

📎 DpPRUAJUYAAk4bf.png

normal drift Oct 14, 2018, 7:18 PM

#

@vocal veldt Thanks for the TTN guide! -- I just got it up and running - with the OLED! Works great!

vocal veldt Oct 16, 2018, 2:25 AM

#

@normal drift no problem - there's more to come!

lusty coral Oct 16, 2018, 6:55 AM

#

Can you have multiple receiver radiofruit feathers per one transmitter radiofruit feather? I’m making a prop that glows brighter and makes noise when it gets closer to certain objects. Worked fine with one transmitter (prop) and one receiver (set object), but I’d like to have 2-3 receiver objects if possible. Each receiver would be in a different room of a house and the prop would be held by an actor moving around the whole house.

normal drift Oct 16, 2018, 11:47 AM

#

By radiofruit, are you using the rfm69 or rfm95 radios.? If so, there can be any number of receivers, especially if they just want to receive a message from the base. It gets a bit more complicated if they have to respond to the base with an acknowledgment since all will respond and all will "hear" all responses. That can be handled via the Radiohead library protocol since it includes a packet header that can contain a "source" and "destination" address id so a receiving system can tell if the message was intended fo it and where it came from. You may not need to bother with that for your application

primal warren Oct 16, 2018, 3:15 PM

#

I think the RadioHead library (and most of the hardware) includes the ability to report the received signal strength (RSSI?). If you choose a protocol that allows multicast, it should be practical.

lusty coral Oct 16, 2018, 8:54 PM

#

Cool, yeah I’m using rfm69. Good to know, putting in an order for a few more and will check back in if I have any questions. Thanks guys

vocal veldt Oct 18, 2018, 3:43 PM

#

📎 frecuencias_LORA.txt

#

@normal drift Many LoRa Freqs and channel relationships

#

courtesy of wero1414

normal drift Oct 19, 2018, 7:31 PM

#

@vocal veldt Got the sniffer going - nice! ordered a few more LORA boards == fun stuff

normal drift Oct 20, 2018, 8:03 PM

#

@vocal veldt Cool! your TinyLora example is working now as well -- I had to make a few minor tweaks to get it to compile but it seems to be working. Let me know if you want the details

vocal veldt Oct 20, 2018, 8:19 PM

#

@normal drift Were you able to get it working on first-run via ABP to your gateway?

normal drift Oct 20, 2018, 8:20 PM

#

yes

vocal veldt Oct 20, 2018, 8:20 PM

#

yey

#

There's a weird tweak, you need to disable the frame counter on the Device itself from TTN

normal drift Oct 20, 2018, 8:20 PM

#

I did not have a problem -- set to 16 bit counter

vocal veldt Oct 20, 2018, 8:21 PM

#

it'll cause some repeats/history issues if left on, I encountered it after a 12hr inital soak test

#

let me know if yours suddenly stops transmitting, though

normal drift Oct 20, 2018, 8:23 PM

#

hmmm - it might have

#

I had to remove the "static" from the contant definitions and remove the include of util/delay --- did you have t do that?

#

disable frame counter checks -- came back

#

do you have a "decoder" -- not making much sense of the received payload

#

changed from "hello Lora" to {0,1,2,3,4,5,6,7,8,9,10} - first time came through as expected -- then not recognizable....

#

I think the way it is written - the data package is overwritten by it's encrypted self so only the first transmission is correct. -- checking

#

@vocal veldt yes, that works -- since you only initialize LoraData at the start, it is only sent correctly once then it sends its re-encripted copy. If you recreat LoraData before each send, it goes through the same each time.

vocal veldt Oct 20, 2018, 8:50 PM

#

Oh, I have a decoder. hold on

#

If you'r esending plaintext

#

  // Decode plain text; for testing only 
  return {
      myTestValue: String.fromCharCode.apply(null, bytes)
  };
}```

#

sensor data: https://github.com/TheThingsNetwork/workshops/tree/master/The Things Network

GitHub

TheThingsNetwork/workshops

Workshops for The Things Network. Contribute to TheThingsNetwork/workshops development by creating an account on GitHub.

#

https://github.com/adafruit/TinyLoRa

GitHub

adafruit/TinyLoRa

Tiny LoRa Node for TTN. Contribute to adafruit/TinyLoRa development by creating an account on GitHub.

normal drift Oct 20, 2018, 8:52 PM

#

I tried taht , but it gave garbage after the first packet.

vocal veldt Oct 20, 2018, 8:53 PM

#

which?

normal drift Oct 20, 2018, 8:53 PM

#

I was using your repo -- should I go to the adafruit repo

vocal veldt Oct 20, 2018, 8:54 PM

#

The adafruit one is in active development

#

I have a few PRs to finish on tuesday (doing school things this weekend) but if you'd like to test on that, it's what we'll release with the guide and for cpy

normal drift Oct 20, 2018, 8:55 PM

#

ah -- ok ---I'll take a look

#

sorry for the confusion

vocal veldt Oct 20, 2018, 8:55 PM

#

no, no confusion. I'm glad you're testing 😄

#

if you hit any radio weirdness with that, i'll take a look

normal drift Oct 20, 2018, 8:56 PM

#

sure will -- thanks for getting this going!

normal drift Oct 20, 2018, 9:17 PM

#

@vocal veldt just a heads up - I can get it to work with a clone of your repo -- but the adafruit repo does not work fro me - no errors, bt no packets received -- I have to make the same changes in each t get it to compile. Maybe I am not getting the right version of things. In any case, I'll keep playing with it -- It can all wait until next week.

vocal veldt Oct 20, 2018, 10:26 PM

#

@normal drift hm, strange. What changes are you making? The delay?

normal drift Oct 20, 2018, 10:28 PM

#

Remove delay from .h remove static from the const definitions in . cop

#

Are you building for M0?

vocal veldt Oct 20, 2018, 10:29 PM

#

Building for 32u4

normal drift Oct 20, 2018, 10:29 PM

#

Ahh. That may be the difference. Different compilers . I’m using feather m0. Minor issue

#

You’re .cpp has lots of changes from adafruit repo. I have not dug into it.

#

No need to spend time on it this weekend. Glad to clear up the compiler difference

normal drift Oct 21, 2018, 12:06 PM

#

@vocal veldt It turns out, the Adafruit repo only needed one change to work just like your repo ( the data rate in TinyLora.h was different - change SF7BW250 to SF7BW125 and it now works) so now I am using the Adafruit Repo and it is working well-- I posted some notes to an issue. Thanks again for this.

vocal veldt Oct 23, 2018, 2:51 PM

#

@normal drift did you encounter an issue where the initial device -> gateway connect took a while (I guess to find the correct freq?)

normal drift Oct 23, 2018, 2:53 PM

#

@vocal veldt There does seem to be a lot of variability - but I did not notice a real problem - I'll pay more attention. Its been working very well for the past day.

vocal veldt Oct 23, 2018, 2:54 PM

#

Using the adafruit/ repo?

normal drift Oct 23, 2018, 2:54 PM

#

just spent a few hours getting one of my applications to send data via ttn to AIO -- it was a bit of a struggle but works now.

vocal veldt Oct 23, 2018, 2:55 PM

#

I just reconnected my test 32u4 to the gateway and it doesnt seem to be showing up in the TTN console..again

#

(ABP isn't great for a few reasons)

normal drift Oct 23, 2018, 2:55 PM

#

I'm using the adafruit repo with your PR#4 applied

#

you did fix the FULLCH to MULTICH , corrcect, its in your PR, but not in master

vocal veldt Oct 23, 2018, 2:57 PM

#

yeah, i'm actually using the brentru/ branch

normal drift Oct 23, 2018, 2:57 PM

#

ah

vocal veldt Oct 23, 2018, 2:59 PM

#

the PR #4 you're referring to is is the one which uses SPI transactions, correct?

normal drift Oct 23, 2018, 3:00 PM

#

this one https://github.com/adafruit/TinyLoRa/pull/4

GitHub

Move printf->#debug by brentru · Pull Request #4 · adafruit/TinyLoRa

Addressing issue #2. Moved prints to #ifdef DEBUG and disabled #debug default in TinyLoRa.h

#

don't think it has the SPI transactions -- but I may be confused.

vocal veldt Oct 23, 2018, 3:04 PM

#

it does but not the correct way 😆

#

ok, updating master to reflect some of your comments

normal drift Oct 23, 2018, 3:05 PM

#

oh -- now I see it. seems to work!

vocal veldt Oct 23, 2018, 3:06 PM

#

oooh there it is!

#

alright, pushed the version with the fixed SF/MULTICH to master

#

Yeah, you can do TTN->AIO via Paho MQTT

#

https://www.thethingsnetwork.org/docs/applications/mqtt/

The Things Network

Data API (MQTT)

MQTT

#

(We already use Paho MQTT Python for the Adafruit IO Python Client)

normal drift Oct 23, 2018, 3:12 PM

#

ah --- I did it via IFTTT but I'll look at the MQTT

vocal veldt Oct 23, 2018, 3:14 PM

#

way faster with TTN MQTT API -> Adafruit IO Python client

#

(we'll do something from IO soon)

normal drift Oct 23, 2018, 3:15 PM

#

too many toys 😉

vocal veldt Oct 23, 2018, 3:16 PM

#

Never enough!

#

but it's all linking together in an IOT-marriage nicely 😄

normal drift Oct 23, 2018, 3:19 PM

#

https://github.com/adafruit/TinyLoRa/pull/4 I think you have to take out the arguments in this call

GitHub

Move printf->#debug by brentru · Pull Request #4 · adafruit/TinyLoRa

Addressing issue #2. Moved prints to #ifdef DEBUG and disabled #debug default in TinyLoRa.h

#

oh nevermind -- you did not make that change -- I thought it was in the PR....

#

I like being able to set the pins in the call. ...

#

I'll pull the current master and give it a try.

normal drift Oct 23, 2018, 3:42 PM

#

@vocal veldt pulled master -- made my usual compiler changes -- works fine!

vocal veldt Oct 23, 2018, 4:41 PM

#

nice! I'm adding in code for channel selection (added sglch, multi/single selector bit next)

normal drift Oct 23, 2018, 8:44 PM

#

@vocal veldt trying latest pr (#10) for Mulitchanel I had to set both channel and datarate for it to work, Is there a default?

vocal veldt Oct 23, 2018, 8:44 PM

#

dr being SF7BW125 and ch being MULTI

normal drift Oct 23, 2018, 8:45 PM

#

hmm - did not work for me. where is the default set?

vocal veldt Oct 23, 2018, 8:46 PM

#

I added a new example sketch

#

in examples/

#

they're set in setup() before lora.begin() as lora.setChannel(MULTI), lora.setDatarate()

normal drift Oct 23, 2018, 8:47 PM

#

for both or only for single? Single channel is not working for me -- digging...

vocal veldt Oct 23, 2018, 8:47 PM

#

you need to set both the channel and the datarate before the call to .begin()

normal drift Oct 23, 2018, 8:49 PM

#

ah -- there is no default -- could then be set to SF7BW125 and MULTI if not set ?

vocal veldt Oct 23, 2018, 8:49 PM

#

oh yep i didnt default it

#

you're right

normal drift Oct 23, 2018, 8:49 PM

#

and did you update the multichannel example ?

#

does not look like it. It works if I set them -- when I try single chan -- I don't anything tranmitting on my SDR.

#

bu the single cha example does not set datarate.

vocal veldt Oct 23, 2018, 8:53 PM

#

updating this..

normal drift Oct 23, 2018, 8:56 PM

#

also should it call SPI.begin()??

vocal veldt Oct 23, 2018, 8:58 PM

#

within rfm_write it calls begintransaction

#

~rfm_write

normal drift Oct 23, 2018, 9:00 PM

#

will my TTN gateway see the single channel mode?

vocal veldt Oct 23, 2018, 9:01 PM

#

only if it's listening to that channel

normal drift Oct 23, 2018, 9:06 PM

#

I must have done something -- THanks for the updates -- I'll keep looking at it.

vocal veldt Oct 23, 2018, 9:06 PM

#

np.

#

I'm going to go "dark" for the next week until midsemester projects and midterms are over, I'll pop up when I'm done.

#

Thanks for taking the time to comment and test :). I'm curious if we could/should include the m0 support and how much space it'd take up

#

currently looking at ~7.53kb

#

The real problematic area is the frequency/region plans for TTN

normal drift Oct 23, 2018, 9:09 PM

#

@vocal veldt I hope I have not been distracting you. I think support for M0 would be good -- Good luck with exams -- I'll post any comment to the repo.

vocal veldt Oct 23, 2018, 9:09 PM

#

Definitely not distracting me

#

If we can keep it under 8k and support the M0, I'm all for it. That's the size I'd like to use for the lib.

normal drift Oct 23, 2018, 9:10 PM

#

I have a 32U4 feather -- I can try hooking up a radio to it so I can test both versions and a combined one if we get to that.

vocal veldt Oct 23, 2018, 9:10 PM

#

I've been thinking of ways to generate the freq plans (the MSB of each freq on each ch reg. for the same frequency (i.e: USA9x) is the same but the MID, LSB are different)...

normal drift Oct 23, 2018, 9:11 PM

#

thats convenient! may be workable.

#

Whlie you are taking exams hopefully I can make some progess on a Python version.

vocal veldt Oct 23, 2018, 9:11 PM

#

see: 0xE5 in AU915 freq. plans

normal drift Oct 23, 2018, 9:12 PM

#

you have bigger problems this week 😉

ruby roost Oct 27, 2018, 7:40 PM

#

Hello radio community. I want to use a BME680 with a Feather RFM69 (433MHz for France). For the BME680 gas sensor, it must be kept on. I would like it to run on battery for several weeks, with 4 measurements per hour. Does that seem possible to you?

trail acorn Oct 27, 2018, 8:20 PM

#

@ruby roost depending on the size of your battery and how clever you want to get with your programming, it's certainly possible

ruby roost Oct 27, 2018, 8:22 PM

#

@trail acorn i look at 2000mAh. But it seems to last less than 10 days. And recharging times are more than an hour.

#

With a basic temperature sensor, i can do deep sleep between measure. But i'm looking for idea with a gas sensor like this BME680

trail acorn Oct 27, 2018, 8:24 PM

#

have you looked at the data sheet to see if has a sleep mode?

ruby roost Oct 27, 2018, 8:25 PM

#

I look for the ESP32 (Wifi) , but i will look for the M0 RFM69.

#

Unless I wire the sensor directly on the battery, all the time, and I do a deep sleep of the feather. this should work.

trail acorn Oct 27, 2018, 8:26 PM

#

The m0 absolutely does but I was talking about the bme

ruby roost Oct 27, 2018, 8:26 PM

#

The BME is very Low power consumer

trail acorn Oct 27, 2018, 8:28 PM

#

someone here probably knows better than me but I think you should be able to turn off the radio and sleep the m0 between reads. Also do you need the reads in realtime or can you record a bunch of reads and then send a blast of data once every 15 minutes or something?

#

you could also just hook the feather up to one of these:

#

https://www.adafruit.com/product/3435

Adafruit TPL5110 Low Power Timer Breakout

With some development boards, low power usage is an afterthought. Especially when price and usability is the main selling point. So what should you do when it's time to turn around and ...

ruby roost Oct 27, 2018, 8:31 PM

#

Yes good idea. No need to send each time i do a measure. Now I'm switching from Wifi to radio because of the energy consumption. But if i can deep sleep the Wifi feather and keep the sensor powered, with a 2000mAh Lipo it should do the job

#

Good little board if the feather doesn't manage deep sleep. Nice

#

I still can't decide if I should switch to 433MHz radio, and make a gateway to process the data on my PC, or if everything in wifi goes on battery.

trail acorn Oct 27, 2018, 8:33 PM

#

the m0 in the feather absolutely can do a deep sleep, it's just a matter of figuring out how to do it. The TPL5110 would just be easier, but then again it would add to your BOM

#

🤷

ruby roost Oct 27, 2018, 8:34 PM

#

5$ is OK 😃

trail acorn Oct 27, 2018, 8:35 PM

#

As long as you don't plan on making more than 10 or so.

#

Your time is probably worth more than $5

ruby roost Oct 27, 2018, 8:37 PM

#

I plan to have like 5 or 6 module with sensor (t°, P, gaz ..... but for some only t°)

#

5$ is cheaper than big battery

#

Exchange data in Wifi directly with my PC on the network, or exchange in radio 433MHz and make a gateway 433<>Wifi. Is it the same to code? I mean doing MQTT in Wifi or radio is the same thing, right?

#

HUZZAH32 with ESP32, datasheet says 150uA in deep sleep. looks great.

trail acorn Oct 27, 2018, 8:43 PM

#

I don't actually know much about Lora; I'm just lurking. Someone else should be able to tell you

ruby roost Oct 27, 2018, 8:48 PM

#

OK. Thnaks anyway for your feedback and proposal

#

Thanks

ocean locust Oct 29, 2018, 7:21 AM

#

@ruby roost consider adjusting the charge rate of your feather if you want faster charge times too

https://cdn-learn.adafruit.com/assets/assets/000/032/914/original/feather_schem.png?1465421956

Top right shows you the resistor value vs. charge current. By default it is 100mA but you can probably make it a lil higher than that

trail acorn Oct 31, 2018, 8:02 AM

#

@slender current , you around?

slender current Oct 31, 2018, 8:05 AM

#

@trail acorn for a little bit longer

trail acorn Oct 31, 2018, 8:07 AM

#

I'm testing a cp32-m4 and getting cdc but no mass storage; it shows up in system info but no CIRCUITPY and I have a bunch of

CP32-M4: family specific matching fails

from dmesg

slender current Oct 31, 2018, 8:07 AM

#

using the internal flash for the fs?

trail acorn Oct 31, 2018, 8:07 AM

#

No, a qspi flash

slender current Oct 31, 2018, 8:09 AM

#

can you get a serial connection?

trail acorn Oct 31, 2018, 8:10 AM

#

ya

#

The flash is a W25Q32JV which I added a config for in devices.h; Could a bad config there be the issue?

restive fjord Oct 31, 2018, 8:54 PM

#

The W25Q series chip used for the SAMD21 Xplain'd Pro board worked out to the same thing Adafruit did with their chip as it had the same geometry and interfaces.

#

I was surprised when the code ran, designed for that Xplained board, on Adafruit hardware. ;)

trail acorn Oct 31, 2018, 10:49 PM

#

(it was a bad config, wrong memory_type)

errant pilot Nov 3, 2018, 1:14 AM

#

hello

#

can anyone help me with how to use a nRF24L01+ to make a wireless SNES controller wireless on a gamecube

#

i already have the wired verson working

tiny dust Nov 4, 2018, 1:14 PM

#

[22:10] <hatsunearu> i am trying to built a canonical DDC SDR radio
[22:10] <hatsunearu> the problem is my clock is some round number like 75MHz or whatever
[22:11] <hatsunearu> but i believe the frequency resolution goes up according to 1/2^N where N is the width of the numerically controlled oscillator phase width
[22:12] <hatsunearu> so i can't seem to get a nice round frequency from my NCO
[22:12] <hatsunearu> like, if i wanna tune exactly to 10MHz or something
[22:12] <hatsunearu> it's kinda impossible
[22:12] <hatsunearu> im wondering what the canonical solution to this is
[22:12] <hatsunearu> do i just throw more bits at the phase register
[22:13] <hatsunearu> or do i add another frequency shifting stage after decimation to get more fine grained frequency shifting

#

any ideas?

primal warren Nov 4, 2018, 1:37 PM

#

The canonical solution used to be to divide down to an even submultiple and then use either a multiplier or a PLL to multiply it back up to the desired frequency.

#

In some cases, another local oscillator would be used to drive a mixer that would then generate the desired frequency.

#

Ed Nisely's blog has some useful articles on the subject https://softsolder.com/?s=frequency+synthesis

tiny dust Nov 4, 2018, 1:56 PM

#

anyone know what happens if i multiply two signals that both have a certain spectral noise density?

#

@primal warren so you're saying i should have two mixers?

#

i already have one mixer that mixes the raw rf input with a fast DDS/NCO LO

#

and it will go through AA filter + decimation

#

you're suggesting i do another shift after decimation?

primal warren Nov 4, 2018, 1:57 PM

#

As for the noise density question, it depends on the noise distribution and if the noise is uncorrelated.

tiny dust Nov 4, 2018, 1:58 PM

#

let's say uncorrelated

#

GWN

#

both

primal warren Nov 4, 2018, 2:00 PM

#

If I remember correctly, I think the noise does increase, but not linearly, so you S/N does improve.

tiny dust Nov 4, 2018, 2:00 PM

#

also i saw some solutions where they just have a huge ass phase accumulator

#

like 32 bits

#

which gives a frequency offset that's too small to matter

#

is it square root of the sums of squares again

primal warren Nov 4, 2018, 2:01 PM

#

Yeah, I think that's the approach Nisely ended up going with: just throw more bits at it to get "close enough".

tiny dust Nov 4, 2018, 2:01 PM

#

yeah

#

i probably have enough juice left over in the fpga

#

not much going on in there anyways

#

thanks

primal warren Nov 4, 2018, 2:03 PM

#

The nice thing about FPGAs is it's not particularly tricky to build odd-size registers and the like.

tiny dust Nov 4, 2018, 2:03 PM

#

yeah

#

i'm currently trying to figure out an optimum combination of my PAC LUT input width and output width

#

currently i have 2x RAM LUTs with 9 bit inputs that also exploits symmetry to get 1 extra bit

#

so it's 11 bits total

#

and output width is 18 bits

#

formulas say my phase spur is at maximum -66dB

#

and my input ADC is 10 bits

#

so i'm /guessing/ it's mostly ok

#

but then again idk, that's a -66dBc tone so who knows

#

the AM spurs due to limited output quantization is too low to matter i think

#

18 bits

#

the other option is 10 bit input width so 12 bits, and 9 bit output width

#

i dont quite like that the NCO output is lower resolution than the ADC itself

primal warren Nov 4, 2018, 2:07 PM

#

That is awkward. It seems to me that you're into "try it and see" territory.

tiny dust Nov 4, 2018, 2:07 PM

#

yeah

#

sim is probably the best idea right?

primal warren Nov 4, 2018, 2:08 PM

#

I'd breadboard it, but that's how I roll.

tiny dust Nov 4, 2018, 2:08 PM

#

don't have the hardware yet, lol

#

not to mention i can't breadboard the low noise RF input stage

primal warren Nov 4, 2018, 2:08 PM

#

Then, yes, sim is the next step!

tiny dust Nov 4, 2018, 2:08 PM

#

this is running at 75MHz and this is for ham radio

primal warren Nov 4, 2018, 2:09 PM

#

I was guessing this might be for ham radio.

tiny dust Nov 4, 2018, 2:09 PM

#

yeah lol

primal warren Nov 4, 2018, 2:09 PM

#

You write like someone who reads QEX

tiny dust Nov 4, 2018, 2:09 PM

#

most homebrew SDRs seem to use NE612 gilbert cell mixers or QSE switching mixers

#

what's QEX lol

primal warren Nov 4, 2018, 2:09 PM

#

QEX is a magazine for ham radio experimenters

tiny dust Nov 4, 2018, 2:10 PM

#

wanted to try and make a cheap and dirty DDC SDR

#

oh lol

#

i'm just a young fart who doesn't know what he's doing

#

also (open) note to self this is a good resource for DDS/NCO https://www.analog.com/media/cn/training-seminars/tutorials/450968421DDS_Tutorial_rev12-2-99.pdf

primal warren Nov 4, 2018, 2:11 PM

#

Ah, useful!

#

I'm just a tinkerer with years of experience, not a lot of formal training.

tiny dust Nov 4, 2018, 2:11 PM

#

i think i'll do my next post on sim results

#

ah

#

i don't even have a degree yet 😄

#

and this is my first RF project

primal warren Nov 4, 2018, 2:12 PM

#

I never quite managed a degree, but I've built various RF gadgets over the years. As you pointed out, low noise RF stages are an animal unto themselves.

tiny dust Nov 4, 2018, 2:13 PM

#

yeah

#

so my schtick is

#

unlike homebrew ham

#

i might use some rf op amps

#

and see what happens

#

noise performance will be /okay/ i think

#

decompensated rf op amps

#

pretty good for HF with gainz

primal warren Nov 4, 2018, 2:14 PM

#

Us homebrewers use everything, nothing is off-limits! I used RF op-amps a while back as gain stages for a high bandwidth fiber optic link. Worked pretty well.

tiny dust Nov 4, 2018, 2:14 PM

#

oh yeah those are totally good for those

#

but it seems like rf op amps are a taboo among hambrew

#

is that a word

#

it is now

#

i saw this SDR design

primal warren Nov 4, 2018, 2:16 PM

#

Pfft, nothing is taboo for me, I'll give anything a try. FPGA direct driving vacuum tubes? Sure!

tiny dust Nov 4, 2018, 2:16 PM

#

quadrature zero-IF

#

NE612

#

spur perf is kinda ass from my analysis i think

#

like they drive it with a square wave

#

so you get 3rd harmonic mix products at -10dBc

primal warren Nov 4, 2018, 2:17 PM

#

The mathematics look interesting, but I'm dubious of the real-world performance.

tiny dust Nov 4, 2018, 2:17 PM

#

anyways

#

the part i didn't understand was

#

it had a preamp but it was weird

#

inductor loaded BJT common emitter

#

but like the input impedance is like 2 ohms

#

so not sure what's up with that

#

gotta bounce

#

good talking to you

primal warren Nov 4, 2018, 2:18 PM

#

73!

tiny dust Nov 4, 2018, 2:19 PM

#

73s!!

vocal veldt Nov 6, 2018, 6:39 PM

#

@normal drift quick q if you're around

normal drift Nov 6, 2018, 6:40 PM

#

yup

vocal veldt Nov 6, 2018, 6:40 PM

#

Were you getting back the "frame counter: #" serial print when runnign on the m0

#

I removed the static in front of AU915, EU836, US902, AS920 and the S_Table but I don't see it sending on my M0 feather...

normal drift Nov 6, 2018, 6:42 PM

#

yes - runs just like on the 32U4

vocal veldt Nov 6, 2018, 6:43 PM

#

weird, mine gets stuck at Starting LoRa... Sending LoRa Data...

normal drift Nov 6, 2018, 6:45 PM

#

are you using a Feather M0 and breakout? what pins fpr RST, CS

vocal veldt Nov 6, 2018, 6:45 PM

#

oo, might be pinmapping

normal drift Nov 6, 2018, 6:46 PM

#

I just built my same version for 32u4 and it also runs -- tested multichan

vocal veldt Nov 6, 2018, 6:46 PM

#

yeah forgot the DIO0 got remapped on the M0 breakout

#

to 3 from 7

normal drift Nov 6, 2018, 6:46 PM

#

ah -- yup!

#

DIO0 is what I meant , not RST . I use 11 on my M0 -- 12 for CS

vocal veldt Nov 6, 2018, 6:50 PM

#

Yep, the DIO0 gets changed. I might add two mappings into the example and comment out the M0 one

normal drift Nov 6, 2018, 6:51 PM

#

are you using the M0 LoRa board -- just got one, but have not set it up yet.

#

too many options 😉

#

rebuilt both from your latest master -- all good -- thanks!

vocal veldt Nov 6, 2018, 7:06 PM

#

yay, thank you for the testing and m0 advice

vocal veldt Nov 6, 2018, 9:09 PM

#

https://github.com/adafruit/TinyLoRa/releases v1 of TinyLoRa is out 📻

GitHub

adafruit/TinyLoRa

LoRaWAN Library. Contribute to adafruit/TinyLoRa development by creating an account on GitHub.

normal drift Nov 6, 2018, 9:18 PM

#

@vocal veldt you around for a question?

vocal veldt Nov 6, 2018, 9:20 PM

#

yep

#

fire away

normal drift Nov 6, 2018, 9:25 PM

#

awhile ago I reported taht when I used your example the "Hello LoRa" string was corrupted by the encryption after the first tranmission. Do you see this?

#

I modified my examples to repopulate the message for every transmission.

vocal veldt Nov 6, 2018, 9:28 PM

#

You repopulated loraData?

normal drift Nov 6, 2018, 9:28 PM

#

yes

#

If you don't - te message is not readable after the first transmission.

vocal veldt Nov 6, 2018, 9:29 PM

#

I was wondering it it was having issue with sending characters vs. numbers since I was testing it with #'s earlier. and it was reading post-first-transmission

normal drift Nov 6, 2018, 9:29 PM

#

te encryption process overwrite the Data buffer with the encrypted data

#

so the sencond transmission is re- encrypting the encryped message -- it all works, but its garbage after the first pass.

#

just a se - I'll post my change

#

here is waht I did ```/************************** Example Begins Here ***********************************/
// Data Packet to Send to TTN
unsigned char loraData[11] = {0,1,2,3,4,5,6,7,8,9,10};

// How many times data transfer should occur, in seconds
const unsigned int sendInterval = 30;

// LoRa Object - (RFM DIO0, RFM NSS)
TinyLoRa lora = TinyLoRa(11,12);

void setup()
{
delay(2000);
// while (! Serial);
Serial.begin(9600);

// Initialize pin LED_BUILTIN as an output
pinMode(LED_BUILTIN, OUTPUT);

// Initialize LoRa
Serial.println("Starting LoRa...");
// define multi-channel sending
lora.setChannel(MULTI);
// set datarate
lora.setDatarate(SF7BW125);
lora.begin();
}

void loop()
{

uint16_t i;

for(i=0;i<11;i++)
{
loraData[i]=64+i;
}
Serial.println("Sending LoRa Data...");
lora.sendData(loraData, sizeof(loraData), lora.frameCounter);
Serial.print("Frame Counter: ");Serial.println(lora.frameCounter);
lora.frameCounter++;

// blink LED to indicate packet sent
digitalWrite(LED_BUILTIN, HIGH);
delay(1000);
digitalWrite(LED_BUILTIN, LOW);

Serial.println("delaying...");
delay(sendInterval * 1000);
}

#

the inital setting are actually never sent - just placeholders -- the for loop in loop set the actual values.

#

I used 64 + i -- so they would be ASCII characters

vocal veldt Nov 6, 2018, 9:35 PM

#

hm, populating that way could be confusing for people. Could you re-declare the array at the top of loop()

normal drift Nov 6, 2018, 9:37 PM

#

strings are a pain in C++ -- I don't think its that simple. You can probably regenerate it with sprintf (if ardunio has that) or strcpy.

#

I took the easy path --

#

It's something users have to be aware of -- or change the lib to make an internal copy and leave the original unchanged (probably not a good idea)

vocal veldt Nov 6, 2018, 9:42 PM

#

yeah, I'm trying to think of an alternative to the for loop char population

normal drift Nov 6, 2018, 9:42 PM

#

basically, need to generate the whole message each time.

#

sorry -- I mant to bring this up before it got released...

vocal veldt Nov 6, 2018, 9:43 PM

#

its ok, i've been away for a bit and forgot to address this issue

normal drift Nov 6, 2018, 9:44 PM

#

we'll see if anyone notices 😉

vocal veldt Nov 6, 2018, 9:44 PM

#

eh

#

maybe add a call to a library helper function to re-generate it from the loop

#

what's confusing me is within sendData(), the encrypt_payload function is called but what's built is the RFM_Data package

normal drift Nov 6, 2018, 9:47 PM

#

I have not done enought wit harduino -- must e a simple way to create a string, "on the fly". in C I just use sprintf.

#

just a sec

#

https://github.com/adafruit/TinyLoRa/blob/master/TinyLoRa.cpp#L460

GitHub

adafruit/TinyLoRa

LoRaWAN Library. Contribute to adafruit/TinyLoRa development by creating an account on GitHub.

#

it encrypts Data

#

then copies it to RFM_Data https://github.com/adafruit/TinyLoRa/blob/master/TinyLoRa.cpp#L480

GitHub

adafruit/TinyLoRa

LoRaWAN Library. Contribute to adafruit/TinyLoRa development by creating an account on GitHub.

#

it destroys Data when it encrypts it

vocal veldt Nov 6, 2018, 9:53 PM

#

yeah

#

Sending LoRa Data...
Frame Counter: 0
189
delaying...
Sending LoRa Data...
Frame Counter: 1
2
delaying...
Sending LoRa Data...
Frame Counter: 2
137
delaying...
Sending LoRa Data...
Frame Counter: 3
181
delaying```

#

looking at char[0]

normal drift Nov 6, 2018, 9:55 PM

#

I have to run -- I'll think about it too -- It's great to have this working.

vocal veldt Nov 6, 2018, 10:40 PM

#

Yet when I place it within the send_data() function, it's the original data (unmodified)...hrm

normal drift Nov 6, 2018, 10:44 PM

#

Where are you checking it in send_data?

vocal veldt Nov 6, 2018, 11:12 PM

#

prior to the encryption call

normal drift Nov 7, 2018, 2:49 AM

#

@vocal veldt I am not seeing the same thing. If I write the data in sendData jsut before teh encryption, I see it changing every time and agreeing with what Is transmitted.

vocal veldt Nov 7, 2018, 2:50 AM

#

Might be a good idea to do the first option and create a temporary data value, i'll look into that tomorrow

#

probably do the copying over within senddata, use framecounte = 0x00 to check against the first run (original value)

normal drift Nov 7, 2018, 2:51 AM

#

I'm not convinced taht is the way to go.

#

Etiher is should copy it every time or not at all.

#

them mesage may be different every frame

#

especially if it contains data

vocal veldt Nov 7, 2018, 2:53 AM

#

right, sendData(tempData), for example

normal drift Nov 7, 2018, 2:55 AM

#

I think either sendata should make a local copy (wastes up to 252 bytes) or jsut document the behavior and let the user beware

vocal veldt Nov 7, 2018, 2:56 AM

#

I really dont like the idea of declaring the data within the loop()

normal drift Nov 7, 2018, 2:56 AM

#

taht is normal if you are captuing sensor data

#

I think the message should jsut be a byte array that si populated in the loop - decalred outside.

vocal veldt Nov 7, 2018, 2:57 AM

#

yeah, maybe I should write a DHT example when I'm at the office tomorrow to better document sensor sends. And redo the hello lora example for re-population inside the loop

normal drift Nov 7, 2018, 2:59 AM

#

I'll try a few ways to generate strings "on the fly" and let oyu know if I come up with anything.

vocal veldt Nov 7, 2018, 2:59 AM

#

idk if it has to be generated on the fly, maybe can do a memcpy/strcpy

normal drift Nov 7, 2018, 2:59 AM

#

strcpy would be fine -- or strcat

#

  lora.sendData((unsigned char *)loraData, sizeof(loraData), lora.frameCounter);```

#

@vocal veldt that worked -- just declare char loraData[11}; in the setup

#

I'll post an issue to the repo.

vocal veldt Nov 7, 2018, 3:29 PM

#

@normal drift what if we declare the character array as a const

normal drift Nov 7, 2018, 3:31 PM

#

@vocal veldt not sure I understand. sendData still needs t change it -- I think taht will cause an error -- besides - normally it will be changing.

#

which character array do you mean - loraData is not const

#

BTW - finally got it running on my feather_m0_LoRA board -- nice!

vocal veldt Nov 7, 2018, 3:45 PM

#

Yeah, the only solution aside from setting inside the example I can think of is creating an argument for static or dynamic data sends

#

but that's really inelegant...

normal drift Nov 7, 2018, 3:48 PM

#

I don't really think it is a problem to set it in the loop -- I would expectd most uses to send sensor data or some changing information, not a fixed string.

vocal veldt Nov 7, 2018, 4:10 PM

#

@normal drift opened a new branch with the fix, I'm going to set up the HQ router again https://github.com/adafruit/TinyLoRa/tree/fix-encrypt

GitHub

adafruit/TinyLoRa

LoRaWAN Library. Contribute to adafruit/TinyLoRa development by creating an account on GitHub.

normal drift Nov 7, 2018, 4:32 PM

#

Great! I’m off for a few hours. I’ll take a look when I get back.

vocal veldt Nov 7, 2018, 5:38 PM

#

looks like it's good on my end

normal drift Nov 7, 2018, 6:10 PM

#

@vocal veldt what if len(loraData) > 10 ??

vocal veldt Nov 7, 2018, 6:12 PM

#

I could make it dynamically allocate to the length of data...

normal drift Nov 7, 2018, 6:14 PM

#

Guess I still am not sure why it needs to be changed in sendData at all - do this in the user code where you have complete control and knowledeg of the packet size.

vocal veldt Nov 7, 2018, 6:23 PM

#

I want to make the user not think about their array in the main code and just worry about the send function and whatever else they're doing, similarly to how IO sends data. I just redid the allocation to resize with DataLength as a commit

normal drift Nov 7, 2018, 6:25 PM

#

OK -- should work fine.

#

"hello Lora" is 11 bytes a C string has a 0x0 terminator appended to it by the use of double quotes

vocal veldt Nov 7, 2018, 6:31 PM

#

I'm using a modified sketch right now for the DHT22, but I noted to change it

normal drift Nov 7, 2018, 6:31 PM

#

OK -- storing it in a 10 byte array could cause memory issues...

vocal veldt Nov 7, 2018, 6:33 PM

#

are you using the latest commit?

#

Yeah, reducing it to 10byte array would cause a mem issue

normal drift Nov 7, 2018, 6:38 PM

#

trying it now

#

from fix_encrypt branch

#

shouldn't the copy go to i < DataLength not Datalength-1? if the array is 11 bytes then go from i-0; i< 11; i++ that will send 11 char data[0] through data[10]

#

I saw a funny packet with the 11th byte incorrect...

#

We can't assume the array contains a character string -- it may contain DataLength bytes of information

vocal veldt Nov 7, 2018, 6:48 PM

#

Right..

normal drift Nov 7, 2018, 6:51 PM

#

other that that, the new version is working fine with a fixed array like the original example.

vocal veldt Nov 7, 2018, 6:51 PM

#

Yeah, it works with a dyanmic one too (dht22 sensor reads)

normal drift Nov 7, 2018, 6:51 PM

#

good!

vocal veldt Nov 7, 2018, 6:51 PM

#

since you're selecting bytes in the decoder on ttn

normal drift Nov 7, 2018, 6:55 PM

#

Do you want me to caomment on the i< DataLenght in the issue or have you already fixed it?

vocal veldt Nov 7, 2018, 6:55 PM

#

just fixed

normal drift Nov 7, 2018, 6:55 PM

#

great - thanks

vocal veldt Nov 7, 2018, 6:56 PM

#

I think we're good to pull into master and release a v1.0.1 with the DHT example

normal drift Nov 7, 2018, 6:56 PM

#

yes -- do you have a decoder for the dht22?

vocal veldt Nov 7, 2018, 6:56 PM

#

mhm sec

#

  // Decode an uplink message from a buffer
  // (array) of bytes to an object of fields.
  var decoded = {};

  // Decode bytes to int
  var celciusInt = (bytes[0] << 8) | bytes[1];
  var humidInt = (bytes[2] << 8) | bytes[3];
  
  // Decode int to float
  decoded.celcius = celciusInt / 100;
  decoded.humid = humidInt / 100;
  

  return decoded;
}```

normal drift Nov 7, 2018, 6:56 PM

#

Thanks

vocal veldt Nov 7, 2018, 6:59 PM

#

np!

#

https://github.com/adafruit/TinyLoRa/pull/13

GitHub

Fix corrupt message sends due to encryption by brentru · Pull Req...

In the examples, a fixed string is created to send the message "Hello LoRa"
but when the message to be transmitted is processed, the contents of the data array passed to senData are destr...

#

@normal drift I can't tag you as a reviewer, but you would mind reviewing/okaying if possible?

normal drift Nov 7, 2018, 7:03 PM

#

sure - glad to

#

I'll pull the PR and run it

#

do you think 2 sec is a good example -- kind of fast ...

vocal veldt Nov 7, 2018, 7:07 PM

#

Forgot to change that back from local testing to 30

#

running out for lunch, be back in 15-20

normal drift Nov 7, 2018, 9:08 PM

#

@vocal veldt Nice Gude!! I had not looked at it in awhile.

vocal veldt Nov 7, 2018, 9:10 PM

#

Thank you! New pages aren't up yet for the 32u4, I'm doing a final proofread in a bit

normal drift Nov 7, 2018, 9:28 PM

#

I think I am seeing them(in red) -- look good -- the M0 should not be much different

vocal veldt Nov 8, 2018, 12:22 AM

#

@normal drift started on the cpy port of tinylora, didnt get much done but I'm just setting it up for now

#

https://github.com/brentru/Adafruit_CircuitPython_TinyLoRa

GitHub

brentru/Adafruit_CircuitPython_TinyLoRa

LoRa and The Things Network - with CircuitPython! Contribute to brentru/Adafruit_CircuitPython_TinyLoRa development by creating an account on GitHub.

#

I'd like to address the counter (framecount) as an external value called in code.py rather than internal to the library - lora.FrameCounter (in TinyLoRa) and pass it into the sending method. The user is manipulating it from the file rather than the library, anyways.

#

thats the main design choice I've made so far -dont know if you agree

normal drift Nov 8, 2018, 12:34 AM

#

If the node (tinylora) is not going to look at the response framecounter. I’m not sure it is worth adding the complication to the user to manage it. It could be an optional keyword that the user can set at init if desired.

#

I’ll hold off working on my python port. You are much faster at getting things out. There are some good examples for doing AES in python. Let me know if you want a link.

vocal veldt Nov 8, 2018, 2:53 PM

#

yeah, I'll make it a kwarg

alpine niche Nov 9, 2018, 1:20 AM

#

Starting to look at a new project is LoRa a 2 way radio?

#

If so is someone able to recommend a module? it'll be roughly 600 meters (2000feet) almost line of site

vocal veldt Nov 9, 2018, 1:40 AM

#

@alpine niche Yep, LoRa is bi-directional

alpine niche Nov 9, 2018, 1:45 AM

#

Cool! is 600 meters unreasonable to expect out of it?

urban oyster Nov 9, 2018, 2:00 AM

#

Are you having fun with LoRa yet @normal drift?

normal drift Nov 9, 2018, 2:00 AM

#

@alpine niche Of course, a lot will depend on your environment and antenna but 600Meters should be fine.

#

@urban oyster sure am.

urban oyster Nov 9, 2018, 2:01 AM

#

Lol, I might get a LoRa radio at some point, I don't know yet. I have my eyes on some other goodies in the Adafruit Shop.

#

Well, back to fixing OctoPrint, have a great night @normal drift.

normal drift Nov 9, 2018, 2:03 AM

#

@alpine niche I have not done a lot of range testing but I did a simple walkaround my neighborhood using a The Tihngs Network LoraWan Gateway at home -- My transmitter just has a simple wire and there were lots of buildings an trees. I still was able to get at least 500 M

#

@urban oyster Good luck with Octoprint!

urban oyster Nov 9, 2018, 2:04 AM

#

Lol, it was working fine, but I think I messed it up, I'll know after a reboot. Cya. Have fun with LoRa @alpine niche.

alpine niche Nov 9, 2018, 2:06 AM

#

I'm thinking the feather 32u4, with LoRa built-in. Ideally temperature, flow sensor and potentially a servo for flicking a switch

#

would like a power monitor as well but can't seem to find a decent one for AC

normal drift Nov 9, 2018, 2:08 AM

#

I was using a RFM95 featherwing on a feather32u4 adalogger for that test. I think all the RFM9x's will perform similarly.

#

Are you using the 433MHz or 915MHz version -- I have 915 -- (in the US)

alpine niche Nov 9, 2018, 2:10 AM

#

I was looking at the 915, if I'm not mistaken Canada's spectrum is the same as the states

normal drift Nov 9, 2018, 2:12 AM

#

I have not tried a real range test with 2 LoRa boards -- at least not more than about 50M yet. I hope to do a test this weekend but with the reciever inside I'm not sure what to expect for just wire antennas.