#general-chat
1 messages Β· Page 186 of 1
?
How in the HECK could anyone refuse rabbits??
Allergies
Also 2 rabbits == 10 rabbits, 10 rabbits == 50 rabbits, 50 rabbits == 250 rabbits... You can see how it gets out of hand
That's personal responsibility.
Get them "fixed", like we did.
:O
Worst thing with rabbits is a lack of care and attention.
I didn't know buns were broken by default
lol
Sadly, it's in the BIOS... So, not a simple fix. π€£
What, you can't just unplug the hardware like everyone else?
Lol
I mean, nothing you can't manage with a cut trace.
Oof.. why is grad school so expensive
$58,400 for an online masters program through NC State
π§
My wife's medical degree was like.. $90k (so she tells me).
... she doesn't work in the medical field.
π
Lol..
I'd say online degree is not worth that.
In person - maybe...
good evening
Itβs NC State.. itβs supposed to be a top 25 engineering masters program.. :/
$1882 per credit for out of state
Plus one of the few online computer engineering masters programs that would allow me to do VLSI online and without the GRE
out of state tuitions at a lot of state schools are 2-3x, it's crazy
luckily I got my undergrad in-state, and employers paid for advanced degrees
My work will pay for my masters but their limit is $20k a year so 3 years or so to do it
Thing that sucks is even if you did your undergrad in state, if you move out of state, your grad program will have to be out of state
I did my first masters over five years I think, but more driven by course load on top of other life things as opposed to out of pocket
I only have one IT cert.. A CompTIA A+ from 2010. π
that's not too bad
and if you work full-time, spending 3 years for masters may be a good idea anyway
If my one grad class is accepted (from a UNC system university) that will reduce the credits I need from 31 to 28. I can do 10 credits my first year. (They have a weird 1cr class) and 9 credits the remaining two years.
Okay so I still haven't figured out how to get the on-chip boost conversion to work so instead I'll make my own boost converter. Use a 555 to drive a 30KHz signal into the MOSFET I pulled from the original board, then send that through a custom wound transformer for the 3.3:5 ratio I'm looking for. Then what I'll do is I'll run that boosted amount through the self contained buck converter to bring it to 5v
Or just voltage regulate it
Either way I have a solution
And I have plenty of fine shellacced wire to work with
That way at minimum battery voltage it can still drive the boost at or above 5v and then down regulate
Are we ... transformer buddies now?
I mean I'm not going anywhere near as complex just a simple 3.3:5 ratio with 2 coils...
But yes we can be transformer buddies
Transforming the world
One winding at a time
I can't believe I just went there...
I mean I could but I'm far more comfortable with the mechanics of a dual coil with a fixed ratio... If you'd be willing to teach me how this actually works then sure but I'm just going to get something working using what I know and have experience with
Don't really want to be implementing things in my test bed laptop that I don't understand
I mean the FPGA board is kind of already pushing that
I would be careful hooking a homemade supply to a laptop: magnetics can do interesting things, and inductors can make voltage spikes.
I'll be shielding all inductors completely
Not risking that EMI
In fact I will probably shield all of the way out to the filtered voltage out
That way the only thing coming in and out is a steady signal
In particular, when you're pulling current through an inductor and switch off that current, the voltage on the inductor can soar to surprising amounts (hundreds of volts)
Yes, flyback diodes or a snubber network. Verify that it's behaving before hooking to something you care about.
Yep, doing all verification and shielding before I actually plug anything but test boards to it
Will do some of the testing on El Pico Yolo
But for now I'm just trying to get it to power the status decoder PLD I put in while on battery power
Because as it stands, that doesn't seem to power up properly off battery
But works fine with USB plugged in with a diode to the output voltage
Some chips are picky about power
Yep so I'm going to make a quick low power boost and see if that fixes it
I think I actually remember reading in the datasheet that it's a 5-12v device
So it makes sense
Just need to get the voltage up a bit
Although... How well would a 555 to a diode / capacitor voltage doubler work?
Less RF to worry about?
Also no arcing concerns because only bringing up to ~8v max
A little charge pump? That should work for low currents and loads that don't vary too much
Hmmm... So it would work to just get the display decoder chip operating from battery but probably best to do proper boost for actual system power?
An ordinary 555 can source/sink 200mA or so, which is often plenty for such uses
For a diode/capacitor doubler, you'd need push/pull drive, so more than the single MOSFET you'd need to buzz a transformer.
Okay I'll just do the transformer as I only know where one MOSFET specced for this is at the moment
And I know where plenty of transformer wire is
Just getting the 555 signal generator mapped and then I'm going map out the requirements for the actual boost section
Hmmm... Is it possible to do PWM on the voltage going into the MOSFET to modulate output voltage via software?
Like maybe with a low pass filter tuned to just below the PWM frequency
Possible, but tricky: it's essentially a sampled data system with a time delay/phase shift. Such things have a tendency to be unstable and oscillate.
(to allow a feedback loop to maintain a constant output voltage using a variable input voltage is the idea)
Alright then I'll just use the custom boost into the self contained buck module I have like I was saying then
That's likely to be less frustrating
And it gives analog outputs tuned to a 5v range for amperage draw and input voltage too that I can tap into later
Alright design of the circuit has commenced. Wish me skill.
(because relying on luck is a bad idea here)
Oh
One last thing
If I'm intending to draw a decent current through it should I have fairly beefy wires on my coils and what kind of numbers of turns (i.e. Impedence requirements) should I shoot for?
Depends on your frequency, core material, etc.
I'm looking to do 30khz with a ferrite core
150Β΅H or so is likely a practical minimum then
Alright alright noted
I'm using a TDK N97 ferrite core with an A-sub-L of 2250nH
I might actually have some labeled 200Β΅H single coil inductors I could start with but then I need to count the windings...
Or add a second winding on top
Right but to get the right ratio
I need to know how many windings in the coil that exists
Which basically means completely rewinding it
Add N turns, drive the primary, measure the secondary voltage, then adjust the turns by the needed ratio
There are several approaches, but "empirical" avoids needing to unwind your existing inductors.
Yeah fair if I have some means of approximate but simple to enact emperical measurement it cuts out the need for a more tedium laden one
(which would give an exact result)
But I don't need exact I just need good enough
In fact, I have a margin of at least +6v overage I can do safely on the buck unit
I saw one blog where someone had bought some transformers from a surplus outfit, and how he derived their parameters from basic principles. I was impressed. I realized I had bought some of the same transformers, but took the approach of writing the manufacturer to ask about the parameters. Happily, the two results lined up nicely.
(above the required 5v)
It was an 11.4(?)v battery that I bought it for and made sure to check ratings, so I can do anything between 5v (tested with USB input) and at least that and if I remember correctly it's actually rated for ~18v input
So I have plenty of overshoot margin
So a few too many turns is way better than too few
For example, suppose you add a 10 turn secondary, and measured a 0.4:1 voltage ratio. You may decide you want a 2:1 voltage ratio, so just divide 2 by 0.4, multiply by 10, and therefore derive that you'd need a 50 turn secondary.
Soooo
I'm silly
I somehow didn't have the switch connected
It shows the battery test when I have it on battery power now
And I have the flashlight working now too
Because it also requires the button
Still going to make the converter to power the 6507
But don't need it just for charge cycle testing of the color codes I am decoding the bar graph to
Color codes are wrong but it is decoding them to unique combinations of colors
Just have to swap a few things in the PLD
This brought to you by checking all your connections
The right way to tinker
β’οΈ
O nose
It cut off some of vidya
Point is it works now
And I am very very happy
Because all my single component charge indicator dreams are about to become reality! Muahahahahaha
It only has 4 bits of color depth arranged as RGGB but that should allow a bunch of colors as it turns out red and blue with 3 shades of green can go a long way
And since the PLD has extra logic and inputs that aren't being used, I can actually use another pin to tell it whether to display the information as though it's a bar graph or just a straight RGGB representation
That way because the system doesn't show battery information except with a button press during a discharge cycle, I can entirely control that key input through software commands since all it does is output battery level on a single press and turn on flashlight output on 2
Which I can have a keyboard button for each of if I want.
Or just have one for the flashlight and using the same pins the Pico uses to display its own color code, do a software controlled latching of that value so the Pico can keep displaying it after periodic checks
Basically multiplex the color code inputs to the PLD with 2 different decode modes and when the system is charging, default to charge color code but have software controlled overrides for the system and while discharging have the color codes entirely system buffered and controlled.
That way I can have the code light always show something if things aren't horribly broken
Anyway point is I was silly and I should feel silly
Turns out buttons don't do anything if they're disconnected
well w/o power they still make a nice clicky sound some of em
Indeed
okay, found a 555, found a ferrite mandrel to just roll my own coil, found the resistors I need, cap should be easy, I think I just need to put it together...
Decided to take the balls to the wall approach and just make it using this electromagnet wire I have with 50 turns on primary and 125 on secondary for a 2.5x boost
(should bring peak battery output voltage level up to about 12v for the buck converter input which I'm pretty sure is what it's voltage sense is tuned for too)
Oh wait no only 10v for peak battery output boosted, but 12.5v when getting supply from USB during charge
Should also allow software charging detection with a fairly accurate readout of system current draw and rail voltage
All this to say
Purty
Even purtier knowing I made it myself π
Ahahahahahahahahaha
Got power flowing across my homemade transformer
LED lit just by induced current :D
Just have it going through a BJT for now
Will add MOSFET soon
Funny story I think my transformer is backwards
swaps and sees if the output voltage is actually higher this time
You'll have to name it Humblebee. π
.. so long as the finished version is yellow.
... that's a good question
Ruffle collar?
Ascot?
Floofscot?
Colonial ruffled ascot according to Google
PowerBoost boards should be fine sitting in a PLA drawer without bags, right?
Like these
Thought so... Most things aren't very sensitive to electrostatic discharge
I keep looking at the $100 spools of ESD safe filament.
Lol
Look up Korean Jesus on Google images and laugh with me
Be like Korean Jesus, jacked on the Holy Spirit
Not only did Korean Jesus raise from the grave, he set a PR for biggest bench
And Korean Jesus said: βI am the Whey, the Truth, and the Light. No one gets as big as the Father, except through me.β
Lol
They sell rolls of sheets of the stuff they make ESD bags from, right? Lining drawers with that is probably cheaper
Probably off Amazon
I got mine from Digi-Key. Don't know what you consider "cheap". Came in a big sheet.
Or maybe Adafruit sells it
Or Digi-Key
50 Pack Mighty Gadget (R) 12" X 12" X 1/8" Anti-Static Foam Wrap Sheets, Safely Wrap Electronics, Dishes, China Furniture Foam Wraps Cushioning for Moving Storage Packing and Shipping Supplies (Pink) https://www.amazon.com/dp/B0844Z2LGY/ref=cm_sw_r_cp_api_glt_fabc_VBFDXEJBXEVPBFGP4BS8
Not sure you need this much
Lol
But cutting down ESD foam bags might work just as well
I consider cheap to be just as inexpensive as possible XD but lower cost isn't always better, so I'll spend a little more for a better product
It'll last a while XD
Okay so after some experimentation and thinking, I'm pretty sure I accidentally over counted the primary windings and under counted secondary and somehow ended up making an isolation transformer instead of a boost...
Tired brain do silly
Also somehow I'm having trouble pinning down bit order and polarity for the LED decoder
But once I have those solved the battery module will be complete enough to set onto a board
Anyway it might just be a lay back day, give my brain a bit of rest
Turns out red lining an engine for too long quickly results in requiring maintenance
And I'm not trying to end up back in the looney bin
Oh and a note: too few primary turns and your drive transistor burns out if you're using a BJT
I now have 3 dead transistors.
You're drawing a distinction where there isn't one: "isolated" and "boost" are orthogonal, you can have either, both, or neither.
Your transistor may be getting overcurrent due to core saturation, or overvoltage due to inductive kickback.
Luckily I have a few more to work with
My point is there was no meaningful voltage increase across it
It was changing a bit but somehow it ended up very close to the same voltage no matter which coil I used as primary
You could try swapping the windings, you should get step-up in one direction and step-down in the other. If you don't, then you have other issues (too much inductance, too little inductance, coupling issues, core saturation, wrong drive waveform/frequency, there are a lot of gremlins to choose from when dealing with magnetics)
Yeah it was showing a step up and down but it was miniscule
Ah, "very close to the same voltage" points to those "other issues"
Hmmmm
I really don't have the tools to debug this...
(in the sense of physical tools and knowledge)
I'm kind of guessing core saturation, it looked to me like you had a lot of turns, and you mentioned highish current (like an amp) and lowish frequency (like 30kHz)
An oscilloscope is a big help, but there are other approaches.
Yeah I tried with fewer windings at the same ratio and it just burned out my driver
Luckily I had the forethought to buffer through a BJT for testing
So it wouldn't burn out the 1 MOSFET I know the location of or the 1 555 I managed to find...
There wasn't a lot of scale clues in your pictures, but it appeared to be a teeny little core, like the size of a pencil eraser. If so, you'd have to get up to half a megahertz or so to avoid saturating it (which may or may not work anyway, if it's the wrong kind of ferrite for that duty and frequency)
Sad but true. I avoided custom magnetics for years until I started seeing articles with lots of information in DIY magazines
Well...guesss it's time to finish my oscilloscope
Also the really good WΓΌrth Elektronik books on magnetics
Even a cheap 70 year old analog oscilloscope is useful here.
In fact, I prefer the old vacuum tube scopes for working with inductors as they won't be damaged by random voltage spikes of a few hundred volts.
I love this passage from the instruction book for one of mine that actually comes right out and says "overloading cannot damage the instrument." Bold claim, that.
oh actually
if that can still be helpful
I'll ask my dad to borrow his old tektronix scope
could give a lot more insight than just a simple multimeter
even just the shape of the wave would be helpful to diagnose what I need to add / change
since it'll show me saturation of core and whatnot by distortions in the waveform
also I'm thinking I may try and just stick a crystal oscillator in the megahertz range onto it and see what happens
thing is I kind of need a small core
this is going in a laptop
and really I don't need to draw that much current from it...I'm not running anything like motors or actuators
I think the highest load will either be the 6507 or the LED backlight for the screen probably
oh maybe the FPGA module
@late fulcrum hey do you think I might be able to take an old backlight inverter coil and rework the secondary to the voltage I need?
oh never mind
this is epoxied together
wouldn't be able to unwind / rewind easily
Though...
I did draw some inspiration from it for doing a bundle of 3 wires in parallel on primary, then the same length of wire but in series on secondary
What is "cathrode"? π
A misspelling of "cathrodeo". An event where electrons see just how quickly they can all jam into one space
Lel
In a ray-type cathrodeo they all try to get across the void and hit their target with precision
I thought it would like.. an electron jumping on a positron and seeing how long it can hold on.
that actually....makes more sense
π€£
Positronium (Ps) is a system consisting of an electron and its anti-particle, a positron, bound together into an exotic atom, specifically an onium. The system is unstable: the two particles annihilate each other to predominantly produce two or three gamma-rays, depending on the relative spin states. The energy levels of the two particles are ...
Yay! I got a sciency thing!
key point here being it's unstable
so the cathrodeo would be trying to see how long they can orbit
before inevitably annihilating
kind of like marriage
Hey!
I've been married nearly 7 years... Lol
yes, and my parents have been together nearly 30
doesn't mean that the increase in likelihood isn't one of annihilation as time goes on
My parents didn't even manage 10 years, so... Ehh
My mum once quipped; "this cooker has lasted longer than my first marriage"... In front of my dad. π€£
"Almost 50 percent of all marriages in the United States will end in divorce or separation. 7. Researchers estimate that 41 percent of all first marriages end in divorce." https://www.wf-lawyers.com/divorce-statistics-and-facts/
I mean....let's sign a legally binding contract that we do and always will love each other, what could go wrong?
not married but I've seen the kind of traps it can present. Have plenty of siblings who have fallen for it.
if you love someone, that should be enough. you shouldn't need to get it legally arbitrated
and as far as marriage-based tax breaks on the assumption of cohabitation? yeah, how about couples that live apart and still get those breaks, while groups of individuals who live under the same roof aren't given them.
My wife has been living with me in the UK since 2015.
We've had some troubles, but... Who doesn't?
I don't try to even date anymore due to past trauma...
Fair.
too much instability for my life thank you very much
already have enough of my own crazy, don't need to take on someone else's too.
OMG
I THINK I FOUND MY BAG OF ASSORTED MOSFETS!!!
My wife's in the US right now.. she took all the crazy with her. π
Lol
My resting heart rate is waaaay down... Amazing what some exercise can do
Minimum heart rate too
ah so they're not all MOSFETs
but they do all seem to be power transistors
these could absolutely come in handy for beam deflection on my oscope
Okay I've got permission on the tektronix
Looks like I'll be able to actually learn some stuff rather than just bumble through and get something that kind of works but I don't know fully why
Bill and Ted's Bogus Journey. π
I could sit and watch all three. π
Lol
I'm going to see Free Guy tomorrow, so looking forward to that.
Okay so I decided that I'm not going to try and push any innovation today...
Brain too muddled
But what I can do is get all the things I need set up for when I'm feeling it
ULTIMATE POWER DRAWER
Nice
Coffee time
I may rebuild the transformer I was doing last night and run it at ~1.8MHz (found a self contained crystal oscillator unit I can use) and see if the voltage does indeed increase
But I'm not doing any trace analysis until tomorrow so I have to clear a space today in my room for the oscilloscope since my lab is almost but not quite ready
Waiting on a bit more of the roof framing and then roof can be paneled, floor sealed, furniture moved in, door framed, and walls paneled and it's ready to start using
(electrical can be run after panels are on and before I get insulation and the pressure seal system in place but as soon as it has a roof, furniture, a door, and walls I will be able to start doing something in there...
My current "lab" space that isn't my room is running off of an extension cord...
I would have a lab all set up with power and everything that I personally connected to the breaker systems but my brother decided to squat in my lab and then start living there.
So I have to build a new one.
And it wasn't the "I have nowhere to go" kind of squatting. It was the "I don't like the accommodations I have been afforded so I'll take yours."
The former I understand and am happy and willing to accommodate
The latter... Well... If it were my property he would be saying "I have nowhere to go"
He has burned me too many times on too many things.
Now I'm in a funk. Bad FBS no think of that
Think of the fact that you get a brand new lab to spec
Well
"Brand new" in the sense that it is a whole new building
Oh did I show the lab progress?
Just needs an overhang on the low side and the roof can be paneled
And I get the area behind / around it as sort of "my yard" π
And the short side is tall enough for me to stand in, contrasted with my current space where I have to duck on ingress/egress
Only the section on the close side of the dividing wall will be my lab, the other side is for my parents to have a space to store things during organizing the other spaces
But still a bigger more high tech lab than I ever thought I would have here...
Ahh, another Valheim player! π
XD
Like
I got the "tri" part right away
How did it take me this long to equate "dent" to teeth
lol
hehe
TriLars
CCFL transformers are designed for use in Royer oscillators so they may not be happy with asymmetrical drive. Watch out for the original winding, it can kick out 1350V
I have a few dozen CCFL bulbs, what can I do with them? XD
Yeah I would remove the original winding and just do the perturbation tuning method for rewinding it
But that would require me to very carefully chip away the adhesive and find a suitable replacement
Because it's in a horseshoe core with an adhered bar to complete a loop
So probably not the easiest to rework
Will try later
one eternity later "oh yeah I was gonna try that thing..."
But uh
I think I'm going to try a higher frequency driven by a crystal into a remake of the original coil that actually (marginally) worked but with twice the primary windings and about the same to possibly a little higher ratio... Lost count on the secondary near my number I needed and just decided to finish filling the mandrel
Also @late fulcrum how's this for scale?
That ought to give a pretty good representation of the coil size
Yeah, that's not as smol as I thought but still not big
Can't get much bigger in the final design
Going in a laptop for boosting battery to system voltage
If anyone has a moment, I could use some help in #help-with-arduino
Sorry for asking here.
Hey I mean, it is off topic and an active channel, I'll go take a look see and try and help
That was original and off the cuff believe it or not
Funny? No.
But I feel pride in having crafted it. Like an over-bodged prototype that somehow just works
π
Immortalizing that so my friends can be scarred for life amused
This is even funnier than the original joke
Lol
Not a high bar so we're clear.
XD
READ MORE BELOWβ¬
First recording of Debussy's Arabesque No.2 with the new 4th harmonic pedal invented by Debussy's fellow countryman Denis de La Rochefordière in Paris. The harmonic pedal allows for the first time controllable sympathetic reverberation in staccato sections.
Follow with score: https://www.youtube.com/watch?v=RMjfUIKQCQM
My ...
2:30
That transition is the only reason I want to play this
Man this proto board is great for figuring out circuits on the fly and having them set onto a board after
Especially since if you only tack end points on things with more than 3 pins, it is pretty easy to rework
Not so easy to rework strip board... Lots more bodging required
But if you have a solid design already strip board is the way to go
At least for one offs. If you're making anything more than a single board of that same design you're better off just getting them fabbed, quality and price-wise
And you can have smaller boards
I mean to be honest if you're good with layout and trace routing and have the money for it, getting custom boards done for everything is the way to go these days
But when you can get 20 9cmx7cm proto boards for $10+ shipping there's an incentive to have a larger number of potential projects per unit of money by far
For about that same price I'm only getting 5 custom boards, but I can cram all of the things from 4 of the proto boards onto a custom board easily so still worth if I had enough cash flow to actually get boards done for each project. Getting there but for now I use what I can.
And yes, getting 5 boards for one project all identical is often a boon for me. Especially if I make them configurable and leave a prototyping area with all extra connections broken out
Which come to think of it that's absolutely what I should have done with the extra space on the last board I had fabbed. Just put breadboard-like layout on the rest of it...
Oh well noted for next time
Oh so I've been being prodded to learn kicad lately by a bunch of different people and I think it is time. Today may be a bit more lax on design work and be more of a tool learning day
Because I just remembered I have the money to get at least one run of my console's board fabbed and put together at least the hardware side of things.
And this time I will design my boards entirely around through hole components because that's mostly what I have like what was I even thinking...
Especially on a prototype
Lol
Last board I had the smart idea of "it will pack better with SMD components" which a) turned out not to be true and b) I ended up surface mounting through hole resistors all over the board. Not pretty. But it worked lol
Turns out if I had designed around the through hole resistor packs I have aplenty, it would have lined up really well with the MCU
Because the resistor packs have the same pitch
Honestly I should have just done a prototyping array with various port connections and profiles around the edges and actually I just designed the prototype for my game console board.
Basically just a board that's mostly strips of traces broken up periodically in some way and with standard pitch hole spacings, maybe even with SMD prototyping space too...
And then all the connectors I want around the edges broken out to a set of these strips as well as the space for a Pico socket
Could sell that as its own thing...
Especially if I can figure out how to make it safe to cut traces on the board but I don't think jlcpcb does phenolic do they
Anyone know of a place I can get cheap phenolic single side boards custom fabbed?
Or is this another thing I'm going to have to set up for in my lab?
ngl piano players are low key witches and wizards
Yeah... I plunk and when people compliment it I'm like "uhh you should watch/listen to someone who actually knows how to play..."
I don't have the focus to learn it that thoroughly
lol
I get bored after like 20 minutes and I'm not very good at reading sheet music
I can't read sheet music at all
But I'm alright at feeling out a song by ear and I throw a few things together. Nothing impressive but it's a good de-stress
I just sing XD
I play whatever instrument is available to me until I get bored and it starts gathering dust... My voice on the other hand I keep coming back to
And whistling
Iβm getting better at reading sheet music
I donβt know what I was thinking, but I learned all of Clair de Lune from its sheet. Memorized it. All
From the sheet
I basically have to read it note by note... Like someone reading a sentence letter by letter...
Well I mean, unless I'm singing... Somehow I got a feel for that, I guess concert choir in hs had an influence there
Sheet music is just squiggles to me, I can't relate it to sounds
I just copy things
At its core it's just a series of harmonic and duration notations.
Oh and amplitude
Heh
Yeah, I can't match those things up with the squiggles XD The actual scales I don't understand either... I just have an understanding of what sounds good and what doesn't, what's up or down a note, and can copy things
If you're addicted to scales do you have to go through a 12 step program?
lol
The scales themselves are easy. Every jump between two adjacent lines or two adjacent spaces is a whole step, every jump between a line adjacent to a space is a half step.
And the squiggles on the far left tell you which ones you're actually referencing in the scale
I mean, I understand the scale system, but I can't relate the paper to actual sound
I guess singing in church from hymnals at a young age probably also helped...
I did that too... but I also had trouble reading
I've got a bit of a disconnect between sound and sight... I still have trouble reading sometimes
Dysmusica?
Something like that idk
I always had trouble not reading things even when they weren't words...
Miiiight have a link to my schizoaffective bouts lol
Like you know the floaters in your eyes?
Ugh, they drive me INSANE
I used to be convinced there was something written on those as a kid and it was words floating around in my eyes and if I could just focus on them I could read them...
Oh boy, yeah, I can see that being connected to schizophrenic thinking
It got to the point they became people in my dreams and I could read them
Turns out looking back on it, it was mostly numbers and symbols
(in the dreams)
Wait, you can read in your dreams?! I'm not alone?
Oh yes
I THOUGHT I WAS THE ONLY ONE
One of the reasons it's hard for me to become lucid
I mean I'm always aware something is off but I scrutinize the environment's features and in the moment it all checks out so I continue as though it's just my life
Wooooo federal worker/contractor vaccine mandate
Sometimes I don't even exist. Nothing exists but a biiiig mathematical grid that I have control over and can throw equations at to plot.
Saw the news about it this morning
And I can only imagine the number of people the company I work for will lose lol
time to demand a raise (unless the whole company is folding)?
That's how i study for tests
I go to sleep so frustrated I become a conscious graphing calculator
Not even joking...
I let robots graph for me
Ok, I need to not look at words, they're starting to fight
Well yeah I don't really have that ability while I'm awake, and it's so much easier to get intuition for a problem when you can just think of a smooth transformation between two functions and it just happens...
So I use robots while awake to explicitly state what transformations and functions I want and it takes much longer...
Plus it uses up time I could be building something
I see physics stuff on the fly, but I can't translate that to numbers and figures you could use
homotopy-invariant robots
Rest your enseffelinguistics
Sorry don't actually try to parse that if your brain can't words...
Okay I have to find that bridge rectifier so I can put filtering on output of the setup
With a beefy polarized cap
I doubt they would give me a raise because they donβt even give good raises for regular annual merit raises
Hmmm very nice.... Starting to look like an actual circuit now
"look mom I almost know what I'm doing" - me
If it doesn't utterly fail I'll tack down the transformer with glue
So it isn't flopping the copper
Wow that sounds dirty...
Inducing work hardening in the copper*
That might be worse...
Anyway test time!
I've double checked connections and this just needs voltage and a clock that way I can test different clock sources
Hence the header
That's switch signal input
As well as voltage and ground
Probably should put a header on the output rail.
Ooh hang on...
What if I drive this with a 200KHz sine wave?
@late fulcrum would a sine wave be easier on the system? I have a 200KHz crystal that I could build a little sine wave oscillator with
I would drive it off a microcontroller but I'd rather not uh... Have hv feedback issues
Wait! A perfect task for El Pico Yolo!
I could do a sine wave approximation off a 4 resistor DAC using PIO probably and feed that into the transistor input
Keep in mind, that header and transistor will be replaced with an oscillator and MOSFET in the final design
In fact I may do kind of a darlington array but with a BJT feeding into a MOSFET
Wait a second! @late fulcrum sorry to keep pinging you, but only doing so when it's important enough to not want it getting lost and when I'm pretty sure you can answer. Would it be possible to use a single solid state relay to control a charge pump?
Instead of 2 MOSFETs
Since it can switch AC it should be able to pass DC both ways for each part of the cycle
And use diodes for controlling direction of flow
Or wait what am I saying I found a bag of power transistors and some are MOSFETs I'm pretty sure
May as well add one now since I have plenty in case of a burnout
Haha some are voltage regulators even looks like just assorted beefy 3 pin ICs
This is lucky as a ducky
Oh man 10A 60v rated power transistor!
Power transistor in, just need to get a heat sink
Have some in one of my drawers inside so probably break time and I'll try hooking the 555 I know works into this
Then if not getting good output, try a higher frequency, then if still no dice then I'll clear a space in my room for the oscilloscope
But uh
This should burn out the coil before the transistor hehe
Especially with my flyback diode across the coil
So shouldn't be seeing huge reverse spikes
And if this design works well enough I may just build the battery circuitry off of this
Since at the moment it's mostly on a breadboard
And once this works it's just a bit of thinking about how the logic states for the status LEDs translate into the PLD inputs...
And the whole circuit will be complete
And I will be able to power it without being tied to power
Well it all makes sense now.
I didn't burn out 3 transistors
Just one and a 555
;_;
Good thing I made the oscillator input socketed on the new board
Welp time to build a crystal based oscillator
What you want is an "H bridge"
I don't think a crystal oscillator is necessary or even useful here.
*dies
C:\Users\ckyiu>ping google.com
Pinging google.com [142.250.65.174] with 32 bytes of data:
Reply from 142.250.65.174: bytes=32 time=138ms TTL=119
Reply from 142.250.65.174: bytes=32 time=1066ms TTL=119
Reply from 142.250.65.174: bytes=32 time=1500ms TTL=119
Reply from 142.250.65.174: bytes=32 time=2120ms TTL=119
Ping statistics for 142.250.65.174:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 138ms, Maximum = 2120ms, Average = 1206ms
C:\Users\ckyiu>ping google.com
Pinging google.com [142.250.65.174] with 32 bytes of data:
Reply from 142.250.65.174: bytes=32 time=14ms TTL=119
Reply from 142.250.65.174: bytes=32 time=190ms TTL=119
Reply from 142.250.65.174: bytes=32 time=1284ms TTL=119
Reply from 142.250.65.174: bytes=32 time=1860ms TTL=119
Ping statistics for 142.250.65.174:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 14ms, Maximum = 1860ms, Average = 837ms
C:\Users\ckyiu>
It's a simple source of megahertz range input to test it at higher frequency... Turns out uh... It got really hot when I tried to do that... I'm just going to throw a timer controlled signal at it from an arduino...
And by "it" I mean the power transistor, the coil, and the bridge rectifier
But the indicator LED didn't light though it may be dead...
Yummy melting shellac
Okay I'm going to do a more proper learning course on this stuff over the next week or so as it appears I suck at it
Gonna see if it still gets melty sans clock input
(see if for some reason it's sending a dead short across the coil)
Oh... I wonder what the switching characteristics of the power transistor is...
Could be it's switching too fast and is looking like a constant high input
Should be within spec, it is a 20MHz device and I'm only giving it ~1.8
Usually the frequency given is the "transition" frequency, which is gain times frequency. So if you're getting 100X current gain, it's only good for 200kHz.
And yes, magnetics are somewhat of a dark art. https://www.we-online.com/catalog/en/TRILOGY_OF_MAGNETICS_EN
EMI Filter Design, SMPS & RF Circuits - 5th Edition | Order Samples Free of Charge & Request Quotes Online | High Quality | Fast Delivery
Ah okay so it may just be saturating due to the current draw
And unable to make a full cycle
Thus shorting the coil
Yes, it looks like you have a lot of turns there, which concentrates the magnetic field.
Yeah I went a little hog wild... I'll try at a lower frequency
Probably just throw a signal at it from an arduino
Look at this one I made a while back, with a princely total of TWO turns on the primary (looks like more, as there are two primary windings, and each of them is three wires in parallel for greater current carrying capacity).
That was based on this article: https://www.elektormagazine.com/labs/12v-200v-dc-dc-converter-for-valve-amplifiers
Fairly easy to build DC-DC converter for valve amplifiers using through hole components (just one SO-8 IC). Galvanically isolated high voltage output. The transformer must be created by oneself and its turns ratio sets output voltage. Input voltage is a 5A/ 12VDC AC adapter, output power is 50 W max.
Hmmm... Seems like uh... Seems like I'm going to need to devote a bit more space to power...
At least I know the output stage is built correctly haha
It's just a simple rectifier / filter stage
So if I were doing something like an inductor that charges up its magnetic field internally to then blast it out at higher voltage, a large number of turns might be more useful
But for something that your goal is inducing transitions in a nearby coil you want to be able to switch internal state faster, right?
... it depends ...
I guess I can expect that answer a lot with inductance related questions huh?
That transformer with the 2-turn primaries has more turns on the secondary, since it steps up 12V to 200V
Basically more turns is for lower frequency. You can think of the core as able to store a certain amount of magnetic field. So if you want to transfer a lot of power with a small core, you need a high frequency, as it will "fill up" quickly.
Right I just meant on the one you are driving a longer wire is going to require more time to settle into state so the resulting field should take longer to settle into that state
Thus more turns == lower frequency because higher settling time
Additionally more turns means more field, and more current means more field.
Like waiting for a longer or shorter hose to fill with water after turning the faucet on
It'll fill each unit of hose at the same speed but more units of hose means longer time to fully pass through
It's weirder and more subtle than that, basically putting the current through the wire induces a magnetic field, and that same field opposes the current through the wire.
This is known as "back EMF"
So it's like if the hose were made of rubber...
It's not the electrons themselves, it's the motion of the electrons
A rubber hose is actually more like a capacitor than an inductor.
An inductor is more like an electronic flywheel. When you push it, it resists turning, until you keep pushing it and it spins up. Then it tends to keep spinning.
I just want 5v from 3.6v-4v input why is this so hard ;_;
You see boards like a Mintyboost or the converter in a power bank and it's just a handful of parts, it doesn't look complicated, but the amount of engineering in them is amazing, dealing with the complexity of switching power supplies, fancy magnetics, layout, power handling, etc.
Yeah I'm like how did they make a boost with just a single inductor and a zener?
(I'm guessing it actually has something to do with tuning the capacitor / resistor array that doubles as feedback input)
The magic is in the inductor. The circuit stores energy in the inductor's magnetic field, then releases it through a diode. That inductive kick is harnessed to get a voltage increase.
Okay so if I have 5.1v zeners how hard is that to build?
Does it has schematic? checks
Because while this would be perfect I'm trying to save as much budget as I can.
Yes, there's a link to the learn guide at the bottom of the product listing, and the last page of the learn guide includes the schematic. β€οΈ 
The smarts (and the switching transistor) are in the chip.
That one's good for half an amp, this one is good for an amp: https://www.adafruit.com/product/2030
Alright will look deeper into these when I have my laptop back online out here
It would be tough to build it into a PLD as you need analog voltage sensing
Ah... Can I do thresholding with Schmitt triggers and dividers maybe?
If I'm just looking for specific trigger levels?
I don't know: it sounds hard to me
This one is even smaller and less expensive: https://www.pololu.com/product/2564
This tiny (0.35β³Γ0.45β³) U1V10F5 switching step-up (or boost) voltage regulator efficiently generates 5Β V from input voltages as low as 0.5Β V. Unlike most boost regulators, the U1V10F5 automatically switches to a linear down-regulation mode when the input voltage exceeds the output. The pins have a 0.1β³ spacing, making this board compatible wit...
Holy crap it's so cute it would fit right on the battery board!
And switching to regulator mode when at or above 5v in? Yes. I need...
checks budget
Or if you need several amps: https://www.pololu.com/product/2891
This powerful synchronous switching boost regulator efficiently generates an output voltage of 5Β V from an input voltage as low as 2.9Β V while using an input current as high as 8Β A. This regulator features reverse voltage protection, and unlike most boost regulators, it offers a true shutdown option that turns off power to the load.
Nah I'm just running laptop stuff off battery
Should not need several amps and if I do ill have enough money to do that later
The tiny one barely fit in my budget as is
But it is on the way
Now I can just make a converter for the fun of it.
Thank you so much for links
Will build the rest of the battery circuit onto a board and get charge indicators figured out since that's just logical analysis
Then once it arrives I will just have pins on the board waiting for it.
And the awesome thing is I can actually just take the setup as it currently is and hook its output straight into that module
And then tap off that for internal system voltage
Since it does automatic mode switch to downregulation mode when at or above the output voltage
Which is super exactly what I was already going to do
πΊ
Seriously thank you so much for pointing me to this. I had $10 of play and used it happily.
(came out to like $9.17 after shipping and tax)
There's some great stuff out there, and I like the folks at Pololu.
Well my wallet likes that it could be used without getting savaged π
And if I like this I may design battery packs as an addon for the console too using these to keep system voltage and use a purely parallel battery array for simpler charge and balancing
Then just run the battery rail to every card slot as part of the bus as well as system voltage and ground, that way you can stick one of these on each board that requires a bit more juice than the system rails could safely apply
And so long as it works as advertised I will be very happy with the purchase
Anyway I guess that means I can move on from inductors until I have a less chonky scope accessible
Time to figure out exactly why the charge indicator wasn't working lol
Opinion poll: what colors should be associated with what states. Primary colors are 1 bit of red, 2 bits of green, 1 bit of blue and there are 5 indicator statuses. Off, low, half, mostly full, and full and I'm thinking I want it to be "use same 4 signals as a direct color mode for the 4 output bits when not charging, then when charging (specifically detect USB 5v before an input diode and pull low through high resistance on that same pre-diode point) it simply displays the charge indicator
Get all the latest Google for Education news on this episode of EDU in 90 -- follow along for updates across training, coding, and more.
Donβt miss an episode! See the full EDU in 90 series playlist at g.co/EDUin90
Are you an Education Leader interested in learning more about Googleβs K12 solutions? https://bit.ly/2GBQCPt
Want to hear abou...
I love a chonky scope
Anyway not sure what colors to use for what so I'll probably make up a map of the colors to see what my palette is.
Seems adding an extra green bit actually gives more color options than a single bit of either red or blue could.
(or at least more significantly distinct ones)
I guess if you were trying to do smooth shades on a sunset green wouldn't help much...
Or wait it would
Because orange
And yellow
Always very nice these Google for Education videos on YouTube.
Oh and @late fulcrum what was it you were saying about H bridges?
(the monolithic units on the ends are self contained H bridges)
(for motor driving)
Yes it has paint splattered on it, somehow this got in the path of backsplatter from gessoing canvases
In high school I used to watch it mostly everyday.
I still watch educational YouTube every day
The hangouts with teachers... Good old days....
for a second I thought you were talking about US states and was really confused
I need at least red green and blue for that.
π
Something something charging states in court
Idk too much effort to be funny rn
No, I was talking about Google for Education videos itself.
Oh! @late fulcrum i figured out why the original circuit calls for such high voltage sense resistors. It's so the passive current to ground through the inductor is low!
That's the reason the original inductor was getting hot when I put it in with resistors of lower value but at the same ratio
I bet if I stuck a high value resistor between it wouldn't get nearly as hot but I still doubt boost will work...
Worth a shot if I can get onboard boost working then I can use this tiny module in my handheld version of the console I have in the works
(the one I just bought)
And yes handheld version of the console will have a single expansion slot primarily for cartridges (but also an SD card slot if you want to add peripherals) and each cartridge for it will have a USB port integrated to break out the USB from the peripheral port
Or just plop a peripheral breakout on if you just want to play around
And it will use the SMD variant of the 65C02
Alongside a bare RP2040 chip
Actually hang on with a few peripheral components the RP2040 could do battery charge control, monitoring, and dynamic boost control via PIO
And since handheld won't likely have onboard VGA capabilities that opens up more PIO options
Unless I drive the tft through PIO too for reeeeeaaaaally smooth screen buffering with sync control
That could be cool
And since the TFT module I'm designing around has onboard RAM, the Pico doesn't have to buffer it
I mean it does but only piece by piece if done right
Oh, here is were you went.
Lego lord of the rings is fun π
Collectibles in an open world kinda setting. and the lego humor stays funny to me
Love the ferrari
Ayyyy, they released the OLED as it's own thing!
tessellation math is killing my brain.
I am enjoying the freedom the esp32 provides.
And the fpga is nice but too espensive, i did some math and using a pico as gpu. might even work better as i'm getting better timings with the pico and lcd than the fpga for some reason.
i got some some work done on running a os under the ulp so now i have battery, signal and system events including deep sleep in place.
And being a fan of the psp's quick ability to reset memory and system states i'm making sure you can quick save the system as it is at any time.
This will dump everything in system to a file on the SD.
and now i'm thinking of using the left over speed and space on the SD controller and i decided it needs a fast Z zipping algorithm to package and read data.
my code is a mess and requires comments XD
this guy has also been super helpful.
it is indeed a very beautiful area of math
Going to the cinema now. π
Wind is the enemy.
Eat less onions.
Not my wind, mother earth's wind.
π
Can't wait until my lab can be used to block Satan's sneezes
Ahh, our Lord and Saviour.
Lost a pile of medicinal herbs because I forgot to weigh down the paper they were sitting on and I don't have freaking walls because my brother took my lab space
Oh dear
And my new one isn't done yet
I reiterate: wind is the enemy.
And it always picks up at the worst times.
You're furiously typing an explanation and can't quickly shift focus? Sure lemme start blowing important stuff everywhere
In my ceres base wind will be a constant gentle breeze
Anyway I need some coffee or my brain is going to implode
I figured out my color code system for the laptop though
I'm going to have default charge indicator colors that can be overridden by the system to display any color via PWM
But the defaults don't need the system running to operate
Today I had a strange paradigm shift mentally
My hobby in my mind became my job.
But that's not bad. I don't mind doing something as a job, it's just the mindset that I will be compensated for work done. What I really don't like is bad bosses.
Now it isn't actually a job until it's making money but my mind shifted to thinking of it as one
But one where my boss is understanding because he is me.
It's a really nice feeling
And thinking about how good it feels to have a boss willing to compromise and work things out made me realize something
@delicate stream i do eventually want people on payroll to be able to pass that feeling along.
So as for your question about getting hired, it will have to wait until the company is profitable but I'd love to have you on board
Also anyone who has contributed significantly to the game console project will be offered a position when things are profitable
I want to make a company capable of doing what I wanted to find in a company to work for.
(and never did)
Which means if I can manage it (though I'm not sure how feasible it is) I want to be able to hire a few skilled people who are payed well enough to compensate the work hour difference, expected to push hard for 30 hours, then get the rest of the week to themselves.
Broken up however they want this 30 hours spaced out.
Self-scheduling negotiated between team members on a given project
And if you get a project done before the optimistic deadline I have set for a given task then any time between the optimistic deadline and the hard deadline you get payed for as though you were working on that project during that period but you are also free to pick up another project during that effectively payed vacation time to earn more. This is the model I would like but I'm not sure how feasible it is
That way if you're meeting quota you're payed well, if you're exceeding quota you're payed better.
No reason to not make both incentivised positively rather than a false negative / positive dichotomy between just meeting quota and excelling.
Because meeting quota is good. Shortening optimistic timelines is even better.
Especially for things that require extensive testing before releasing them
Because if you can short an optimistic timeline then bug fixing can be more thorough
But if something is delivered with bugs and gets kicked back by QA then you are still active on the project
Until it doesn't get kicked back. QA is there as a check not a fix.
The team who made the buggy system still has the responsibility to fix those bugs, QA is just there to catch any that slip through the team's view
And there should be some kind of limit on the number of times you can submit something to QA maybe like once a day
That way you aren't using QA as your personal bug finding team.
Heh, after one company laid me off, they hired me back as a contractor when they realized they couldn't handle the stuff I used to do. I quoted them $5000 firm fixed price with a two week period of performance. They said they didn't do business that style. I said take it or leave it. They took it. I finished the work in an afternoon then sat on it for an additional week, then delivered early.
Noice
Yeah I want to be able to enable go getters while paying those who perform to standards well too.
But this won't be for a while
And I want it baked into payroll not something that an individual has to have the guts to stand up to me and say "this is how it's going to be through a contract"
I want to allow those who are like me to flourish and not feel like their talents are being taken advantage of
That is my biggest gripe about working for someone else. They just saw me as another set of muscles with programmable control code. I'm much more than that.
And I know there are many others like me who would like to be rewarded for innovation and outside the box thinking not shunned and reprimanded for it.
I want to be
The modern bell labs
But distributed across the globe using the internet
Which means I'll probably be entirely crypto based as far as holdings so that paying people for work is easier to do, which if I can get big enough might help stabilize crypto into something people view as an actual currency without question
Since buying with crypto will be cheaper since you'll be able to have it pay straight into a wallet, whereas to buy with fiat you have to cover the exchange fees
Which will incentivise people who like the products to have at least some crypto holdings to make for just paying exchange fees once, or accept crypto themselves so they have a cheaper option without having to cover exchange fees because whoever sent it to them for work they did presumably payed that or mined the crypto themselves or were payed it for work they did.
So rather than being fiat backed and accepting crypto I'm going to flip it on its head
Crypto backed but we accept fiat currency with a small fee to allow people to pay who don't have any crypto
It's the only way to stabilize it is to tie a product good or service to it as the primary means of purchasing that product good or service
And if I can get my whole supply chain on crypto then even better for that stability
Especially since I'm going to work towards selling plastic I've recycled into 3D printer filament using no fiat based input to tie it to that I can give a constant, set price for in some crypto currency and never have to worry about its relation to fiat
Yes I will have a recycling plant in my lab.
It will probably take up part of the high end ceiling space
And have the extrusion run along the interior divider wall down the slope to a spooling station in the corner. Can use an archimedes screw to lift shredded plastic up into the extruder hopper from table height on the high end and put a little maintenance ladder into the wall in case something jams
Then I can have a shredder on the high wall/divider wall corner that feeds straight into the miniature plastic silo lift
So I can just drop cleaned plastic in, turn on the system, and have a spool of plastic some time later
Will be very nifty for printing things in PETE as it is very common
And I will be able to sell rolls of it at a stable price only accounting for the crypto to work value ratio not the crypto to fiat ratio
Or I guess in this case crypto <-> work <-> fiat exchange rate that a lot of even "crypto-backed" economies tend to use
And the ones that do tie it to constant crypto prices tend to be illicit economies and black markets. It's happening more with things like bespoke software development which only requires a one time fiat purchase and direct work input but we need physical goods tied to crypto so that isn't so arbitrarily defined on both sides as far as "what is the currency worth to me" and "how much do I value my skills"
If we at least tie it a bit more concretely to "what is this currency worth to me", "what is this physical good worth to me", "what is this currency worth to others", and "what is this physical good worth to others" then we can have an actual economy based on it. The value of something like bespoke software development is too ethereal and abstract to anchor a currency
Seems like the person deciding these deadlines has financial incentive to make them such that they allow him to pay what he actually wants to pay
Also I'm not sure whether paying less for the same work on different timelines is legal
I think you're misreading. I want to pay above market value for the hard deadline and then bonuses above that for the optimistic deadlines
Incentivise both
Hm ok, if it's an employee rather than a contractor then that sounds kosher
That's the idea
And allow those who have finished work early to pick up tasks in the time they have freed up to earn more if they want.
While getting payed whatever they already would have for the time between when they turn it in and the hard deadline
Basically you get payed for turning in before the hard deadline and however early you turn it in doesn't matter as long as it is complete, you are payed for how long it should take a normal person
And with the time that you have between turning it in and a hard deadline you have 2 choices. Take paid leave, or start another project that is separately payed and adds to your income
The model should accommodate workers from average to stellar
And award for increasing productivity
While paying everyone fairly as a baseline
But again I need to be profitable first so initially stock options then contract work until then
Yeah steady-state capitalism basically doesn't allow this. You might be able to do it for a while if you significantly outpace the competition
What specifically is the issue?
Because I'm good at bug fixes and workarounds to get basically the results I'm looking for
One issue is that a company that is identical in all respects but doesn't pay the bonuses (or pays lower bonuses) could out-compete you
Another issue is that the person who sets the deadlines has personal financial incentive to set them dishonestly
This is a good point... Also with the fact that I have no real baseline on what a normal person should be able to do, even if I try my best I may be viewed as setting it dishonestly even if I'm just poorly referenced on what one should be capable of in a period of time...
There is no normal. One person can be excellent but slow at thing 1, fast and average-quality at thing 2. Another person can be the same but reversed. A third person can suck at everything when working alone, but excellent when working on a team of people at least one of whom is wearing a green shirt.
Right I just mean as far as setting a reasonable deadline
You can take industry averages for individual tasks, but you still need to be careful about both quality standards and team culture
I have no reference on reasonable because I have a habit of doing things unreasonably quickly and other things take unreasonably long
I could sort all of my electronics components in a week with a concerted effort but I haven't been able to get myself to do it for almost a decade now as an example
There's trauma associated with the reason they're currently unsorted so it's hard to do, but someone untrained but without my trauma blocks could get it done in a month tops with a reference sheet...
Well more like a reference book...
I only need a reference for the IC's and even still I can sort them by code if I don't know what they do.
Just can't seem to bring myself to do it.
I get started, something triggers the trauma, and it ends up getting left around, tripped over and ends up even worse than it started
Heh, I've been doing this for a local makerspace. They bought an old radio factory and had to sort through decades worth of parts and materials.
I could absolutely do it for someone else's components
Just these particular ones are uh... Tainted with bad memories
But it's all I've got so I use it
I just bumble through the unsorted mess every time I need a particular value
Once my lab is up and I can sprawl it out on the floor and leave it without it being in my bedroom or the common space where nobody can keep their hands to their own it will be a lot easier
Trading sorting might work. You sort someone else's emotionally tainted hoard, and they sort yours.
Could do... Might be able to trade my mom to sort. Give her a sheet that just says the important things to look for and a magnifier and offer to sort her storage space in exchange
She has already offered to help sort several times... Probably should take her up on it...
And offer work in exchange
Things like resistors I want to do myself
But anything with a number/letter code I feel comfortable passing on to someone else.
Or things that are obviously the same like switches and such
I used to have drawers and drawers of sorted and labeled components until a fit of rage ended it. Don't do anger kids, takes years off your tinkering...
Also parents out there if your child is going through crisis and asks to be alone for a while and you just keep badgering them, it may not end well.
Mistakes were made all around and regret is felt by all involved.
And afterwards, perhaps trade garages wholesale
Anyway sorry didn't mean to make things so heavy...
Just ya know
PSA
Mental illness can cause irrational responses to what an external observer sees as standard fare behavior. Know the symptoms of "I'm being too much for this person" and don't keep pushing when you see them.
Turns out some people (myself included) have some data paths shorted in sections of memory and control logic and when those areas are activated it wreaks havoc and is very hard to debug...
And since that logic has control paths to actuators, crazy external results can happen too.
Which are then hard to sort out afterwards
Heh
I was like a malfunctioning 3D printer crashing the head with no reset button
I mean, I make jokes in hindsight but in the moment it was not funny.
Turns out technicians had to be called and I was hauled away to the repair center. They tried their best...
Turns out much like me they are better at the low level hardware repairs than high level software patches and just use crude shutoff mechanisms that disable necessary system components and claim it's fixed because the head isn't crashing anymore.
And look it's still capable of printing... A single layer.
So the process for me has been to figure out how to avoid broken code areas, fix jumps to it as they come up so nothing references the broken code, and at the same time write in new fixed code to jump to that does the same useful things. All using code that already exists and that I gather through my journey and sift. It's a hard road but I'm almost ready to go in and repair my own print bed because as I've been attempting fixes it has been getting more and more marred up with each failed attempt so I'm waiting for a stable enough version of the code to start making repairs so it isn't in vain
And yes I know this analogy went way off the rails but it all relates
And if you get it you get it. If not, well then it's a nice story about a self repairing printer
I was at a bad meeting at a job, getting more and more angry. The other people there could see the boss was being unreasonable and ignoring the signs that I had Had Enough, but he kept going and yelling at me, until he finally accused me of lying. I quit on the spot and stomped out of the building, and he followed me, continuing to yell at me. I got in my car and started the engine, but he was still yelling at me and blocking my car. I finally drove over the curb to get away from that maniac.
"Violence isn't the answer....
It's the question. The answer is 'Yes'."
Justified anger is a thing. Justified violence is not.
I disagree, but it's not really the place for a debate. π
No, it is not.
Mmk
At least, not that debate.
How about a description of my simplified decode logic for the LEDs from the charge chip?
π
I realised that since if theres an expected output at all, I can assume that each previous LED is lit if the current one is and just select them via priority
That is, only look at the signals individually and if one that precedes another in priority is on, it shorts the logic to select that outcome
Also much easier to test
Because I can just input one signal at a time
I'm basically doing an inverse decoder and ignoring any signals that come with less precedence, then passing that through a selector for the output lights
then I'll have an input that sets whether the output signal is coming from that select or passing RGB PWM signals straight through from some other inputs
This way it has default charge indicator colors while the system is powered down but it can override and set its own color with PWM if the system is on
And since I'll be doing analog charge monitoring for the system to have a battery percentage display, the system will be able to do smooth color transitions on the charge indicator light as the battery charges while the system is on.
Oh and I think I figured out how to get smoother light output so it doesn't flicker as much. If I have an RC circuit on each LED with a diode in series to the resistor and in the correct orientation and tune the resonance to the PWM frequency, I should be able to have the cap charge as quickly as it can and then discharge slow enough to hold a bit of brightness to the LED and where it holds depends on duty cycle
Anyway point is for now it will have defaults controlled by a 4 bit RGGB "DAC" multiplexed to 3 PWM channels for RGB control from the system
Oh also I think I just figured out the filter system. Resistor between signal and LED, capacitor between VCC and signal.
When signal is active, capacitor immediately charges, when signal is inactive capacitor discharges through resistor and LED
Tune RC to pwm frequency and it should smooth it out pretty well
And if you ran PCM through that it would probably be even smoother.
Something like this
I wonder if the F16V8B can do tri state output... Would make this work a lot better
Since then I could set to 1 for off, 0 for on, and floating for "use stored energy"
Would need to be careful about voltage spikes when switching straight from 1 to 0 though
Since 0 grounds the signal so however much energy is in that cap will dump through the chip to ground
I guess I could do another diode into the cap from the signal
Oh no just a small resistance to act as a bit of ballast buy not so much that it takes too long to charge and discharge
Something like this
Big R meaning a fairly large resistance, little r meaning a fairly small one
Anyway I'm procrastinating again
I should be writing PLD code to do what I'm describing
I don't have enough spoons to catch up on everything I missed, I only read around your @, lol
Unrelated, I "upgraded" one of the big red buttons I just got!
I love it
I just cut out part of a security envelope and put it over the detractor, lol
Use it for the AI stop button so they have a hard time processing what it is
Lol
I mean it would work until they learned to associate anything unknown with potential death...
Just gotta make sure that data is never saved
Only do one at a time
Or one in each location at a time π
I don't have unlimited time in this universe you know π
Not yet
Lol
And until I do I can't assume I will
Woah that's actually a good motivational slogan...
"Until I do I can't assume I will"
What's the resistance on the LED? Or is it just a straight diode connection with no load resistance?
Well shoot... I guess you could just send lower and lower voltage in until it doesn't light up...
Product page is sparse on details outside of dimensions https://www.adafruit.com/product/1185
Yeah, I do have a breadboard PSU I can do that with
I need to start actually reading the full product pages before buying things, lol
Yeah I feel that so hard...
Lol, glad I'm not alone
Once bought what I thought were usb wifi modules but they turned out to just be called wifi modules when really they were keyboard controllers that used the wifi standard to communicate
And I had no keyboards compatible with them.
So completely useless to me
Oof, weird... One time I did get "Bluetooth" modules that were actually just 433MHz radio pairs, they had no pictures and details were buried
But for $2 a piece I wasn't crying XD
Yeah these were $4 each and don't communicate with each other even... And I bought 4 thinking "usb wifi modules for that cheap? No way! I'll get 4" without reading description
Derp... At least mine had use, they were used for wireless serial links for a while
Yeah if they could at least communicate with each other I would have found a use for them.
At least it was $16 and not $1600
True...
well crap, I have a new interest to obsess over
I got sucked into Halo Lore lmao, bought my first few books
Oop
I feel like several here might like this https://www.youtube.com/watch?v=YV__C64Ni50
i procured a hurdy gurdy through mystical means, and i refuse to learn what a laptop is
join my patreon: https://patreon.com/briandavidgilbert
this is where karen han is: https://twitter.com/karenyhan
this is where brian david gilbert is:
https://briandavidgilbert.com
http://twitter.com/briamgilbert
http://instagram.com/briamgilbert
http://...
esp @late fulcrum
for the cosplay at least
Man that got several hearty laughs out of me
And like
Now I need to go watch the we like watching birds song...
Because it will be stuck in my head for the rest of the day otherwise
And probably will just get further stuck by listening
Such is life
Foto de Eudecio Reis
So many GB of internet(per month) avalible in the smartphone.
It feels like I can download all data in the world.
I think I figured out in 5 lines of logic how to express this decoder...
SkyBlueEnable=!L4;
YellowEnable=!(SkyBlueEnable # L3);
OrangeEnable=!(YellowEnable # L2);
OrangeRedEnable=!(OrangeEnable # L1);
NoneEnable=!OrangeRedEnable;
means logical OR, L1 through L4 are the indicator signals encoded as a bar graph and are inverse logic
So I check if the highest bar is lit, if not check next and so on
Which should allow me to test each one simply by activating each highest signal pin one at a time for the different color outputs
And simplifies the decode logic, then I can use those enable signals to map out which bits of the color space to use under each condition with select lines between the decoded of output and a PWM override
(with PWM override mode simply directly mapping each of the PWM input line states directly to each of the color bits, with green maybe even having another pair of select bits to choose between which bit to map the PWM signal to in that mode)
I blinked when I saw #, which is the "mingle" operator in Intercal.
Or I could have misremembered, one reference gives ^, another gives $. However, I count myself glad to have forgotten some of the horror that is Intercal.
Dying
Ah yes, the "programmer too polite" error.
"The C-INTERCAL compiler will automatically include the system library if a DO (1xxx) NEXT statement is used, and if no (1xxx) labels are defined anywhere, where (1xxx) is a label in the 1000-1999 range, inclusive." π₯΄
that sounds... intuitive
I know the entire song (that's in the video) by heart
oh intercal
i love esolangs
Yes I enjoy this
12GB, you can download a Windows update XD
Or 6 jetbrains IDEs π
(But the installers are only ~700mb, the installation size is ~2gb)
Lol
So on thursday I recorded myself playing that entire song I've been working very hard on. All 5 minutes of playing
And I only made 2 or 3 obvious mistakes which I could've hidden by not stopping when I made them (a performance habit I need to fix)
For me I skip back a measure which is even worse π
Oh, dear
Here, I'll send the video for those curious
The piece is Debussy - Reverie
You know I really have to thank my teacher for never telling me my performance is good
And always finding something wrong with it, then explaining how to work to fix it
He tells me when he sees improvement, but never that it's enough
Glad I have the maturity to understand how good of a teacher that makes him
Recording is so difficult because of the tendency towards perfectionism
I am trying to shed my habit of stopping on "mistakes" but it's a long road
Fwiw, I think classical music culture is way too intolerant of mistakes as being an inherent part of live performance
One of my favorite performances ever is the Miles Davis quintet performing "It Never Entered My Mind"
And he is so out of tune
I imagine him getting a zero from Simon Cowbell (ok I switched genres, and thank you phone for that excellent autocorrection)
Turns out the way Miles does it is wonderful
Hm
@wooden schooner
READ MORE BELOWβ¬
First recording of Debussy's Arabesque No.2 with the new 4th harmonic pedal invented by Debussy's fellow countryman Denis de La Rochefordière in Paris. The harmonic pedal allows for the first time controllable sympathetic reverberation in staccato sections.
Follow with score: https://www.youtube.com/watch?v=RMjfUIKQCQM
My ...
Go to 2:30
Isn't that transition beautiful
Outside on phone, let me get to a quiet place
haha, the delicate little dragonfly trailing behind the rest of the stampede
I hear that prancing melody as footsteps
It's beautiful
It sounds like an underwater theme to me
The depth of the bass, the simplistic beauty of the melody
it's the same theme but in an airy (or as you say, underwater) delicate presentation
I love the way this pianist's hands move across the piano -- there's a beautiful ease in it
I guess this image probably came from Saint-Saens, Carnival of the Animals. But that one is more of a zoo trip than a stampede
Hey @delicate stream did you still want a beta Bread 2040?
Yes! :D
Also yes. Lol
Is it okay if the headers are soldered on?
I only have one unfortunately π¦
All good.
Sure, I'll probably use it with headers/unmodified
If they're already soldered
Only one of the GPIO (pin 12 I believe) doesnβt work which from what I can tell is just a lack of enough solder on the pin for the IC. But everything else works positively great π
I might be able to fix that. I think I can survive without a pin XD
Iβm doing the circuitpython PR this weekend
Yay!
I hope this is the beginning of the Skerr Empire... That one day you'll be successful like Ladyada, and I'll be able to say "yeah, I was there at the beginning, I got this beta Bread board!"
lol
hopefully after i finish grad school I'll have better means to relaunch oak dev with better boards and designs π
all the current designs will be open source and available for anyone to make
I like open sauce
I'll send you my chicken tikka masala recipe. π
but anyway message me or email your details so i can send it. email: hello@oakdev.tech
Yay! DM incoming
π
I've already got two Picos, so I think one will go into a keyboard.
I ordered 5 pico from digikey on thursday lol
i like to have a good variety of boards
I just knocked 6 Picos off my desk
xD
Picos for DAYS
I don't have any 3rd party 2040s though
