Unable to retrieve can bus uuid

[lyric sable]

Now this ! Error during homing probe: Communication timeout during homing

Yet I can select home and homes fine

[frosty tangle]

that one has a guide page about it: https://canbus.esoterical.online/troubleshooting/timeout_during_homing_probing.html

[lyric sable]

The probe issue and mesh ebb time out happen at exactly the same point which to me doesn't make sense as wiring issues would be varied. I've moved wires I to another position too so would happen at a different time which I doesn't

I'll uninstall obico

[warm sandal]

out of interest, what makes you believe obico is the issue?

[lyric sable]

well only teo things that differ cam added or obico otherwise no issues prior

oh and the flow calibration from orca

otherwise printer reliable with cura cube several times

@frosty tangleade no difference still getting the follow.

Error during homing probe: Communication timeout during homing

[frosty tangle]

what made no difference- what did you do

[lyric sable]

Iade the adjustment you pinned for me. Nice=-10

And the failure happens at exactly the same point again

Obico and webcam removed

We shall see

Ok so exactly same issue at same point again

Give up with this now

[frosty tangle]

how far in?

have we seen a log file recently?

[lyric sable]

it fails as exatly the same pint every time. the voron logo on tool head dies and then comes back. im checking klippy log noew

does look like coms error

im turning off for tonight at ill get back to it tomorrow. the connection from toolhead to octopus is the custom can bus cable not any home made connections

[frosty tangle]

your bytes_retransmit reading does climb towards the end of the log, which does point to a physical cable issue, usually

[lyric sable]

Is it possible there's is a power drop. From psu?

It was just odd the see the RGB led die and then come back or is that indicator or cable issue

[warm sandal]

i doubt that'd cause climbing bytes_retransmit values

[lyric sable]

Ok I shall pull the cabling apart and see what the issue is tomorrow

Thank you all again for advice. Is there any know issue using the moulded can bus cable that comes with the sb2209 rp2040?

I know anything is possible but issue with a moulded cable with zero movement

[frosty tangle]

still benefits from strain relief at the toolhead end, the xt30 connector can be a bit wobbly

[lyric sable]

Ok shall look into it

@warm sandal you say climbing. As in the error gets worse till it affects the communication. Hence what it fails at same point during the bed mesh

[frosty tangle]

it's just a pointer to wiring that's losing CAN messages

[warm sandal]

bytes_retransmit increases when klipper detects that data didn't make it to the mcu and it tries to send it again

[lyric sable]

And would that be similar to an accumulated timer where when it hits a limit it shuts down the communication or reboots tool head ?

[warm sandal]

i'd assume that when bytes_retransmit hits a certain threshold, klipper assumes the mcu is no longer responding and just gives up causing the Error during homing probe: Communication timeout during homing

[digital ocean]

You do not want it retransmitting if you can avoid id. Wiring again

[lyric sable]

Ok so easiest thing first, wires. Was curious to understand why it's giving that info

[pulsar topaz]

some unasked for nerd knowledge:
it's not really that it hits a certain threshold, it's just that a bytes_retransmit means a "packet" had an error in it and had to be retransmitted.
No harm no foul as long as the retransmitted packet can still get to the MCU within the required timeframe. THere is a buffer of queued commands so as long as the packet/informatino is retransmitted before the mcu "needs" it it's all fine.

However, if you're getting any bytes_retransmit then you likely have an underlying issue which will cause more errors. Maybe in the same packet (requiring it to get resent multiple times) or just enough packets so too much overall time is sent retransmitting rather than unique data.

Enough retransmits can cause this timing winow to fail and you get timer too close errors.
Or if the keepalive checks stop coming in because of the connection being down (or maybe because it's spending too much time retransmitting) you get a lost comms error.

"comms timeout during homing" isn't necessarily linked to bytes_retransmit. Homing/probing actions are on a much tighter timing window to normal operations so if the Pi-side is being slow for whatever reason you can miss this 25ms window and get the "comms timeout during homing" even though all the wiring/data is perfectly fine.

That's not the case here as there are clearly bytes_retransmit in the logs, but thought I'd mention that timeout during homing isn't necessarily the same thing as lost comms or TTC (though retransmits can also mess with homing windows)

[warm sandal]

good to know, thank you for the detailed explanation

[lyric sable]

thank you for giving a description of the cause and reason

[lyric sable]

all wires and cables checked and immobilised connections checked and confirmed with continuity, and the tool head stops at exatly the same point

it cannot be wires. the wires are in diffrent place and tied down. so if the issue is wires it would fail at a different point but its exactly the same every time

this is bed mesh config

speed: 300
horizontal_move_z: 15
##--------------------------------------------------------------------
##    Uncomment below for 250mm build
#mesh_min: 40, 40
#mesh_max: 210,210

##    Uncomment for 300mm build
#mesh_min: 40, 40
#mesh_max: 260,260

##    Uncomment for 350mm build
mesh_min: 40, 40
mesh_max: 310,310
##--------------------------------------------------------------------
fade_start: 0.6
fade_end: 10.0
probe_count: 9,9
algorithm: bicubic
zero_reference_position: 150, 150```

now its say lost connection with EBBCan. Ive checked all connection all crimps are doubled up iwht solder to prever fraying. the printer was work flawlessly until i started orca calibrations

as the tool head has to move for homing and QGL and at some point the crosses the same point as where it fails during mesh. So why is it not failing there!

I have removed crowsnest and bed mesh appears to be working but will repeat and see

Nope now get further on mwah and fails again is there no some way of pin pointing the issue!.

New error

Can not update MCU 'mcu' config as it is shutdown
Once the underlying issue is corrected, use the
"FIRMWARE_RESTART" command to reset the firmware, reload the
config, and restart the host software.
Error configuring printer```

Console output

11:55
Klipper state: Disconnect
11:55
FIRMWARE_RESTART
11:54
Klipper state: Disconnect
11:54
Probing failed due to printer shutdown
11:54
Probing failed due to printer shutdown
11:54
Lost communication with MCU 'EBBCan'
11:54
Lost communication with MCU 'EBBCan'
Once the underlying issue is corrected, use the
"FIRMWARE_RESTART" command to reset the firmware, reload the
config, and restart the host software.
Printer is shutdown```

So working the issue, wires and connectors tested and checked. What else can it possibly be?

Spontaneously disconnecting now !

12:06
Klipper state: Disconnect
12:06
FIRMWARE_RESTART
12:05
Klipper state: Disconnect
12:05
Lost communication with MCU 'EBBCan'
12:05
Lost communication with MCU 'EBBCan'
Once the underlying issue is corrected, use the
"FIRMWARE_RESTART" command to reset the firmware, reload the
config, and restart the host software.
Printer is shutdown
12:05
Lost communication with MCU 'EBBCan'
12:05
Lost communication with MCU 'EBBCan'
Once the underlying issue is corrected, use the
"FIRMWARE_RESTART" command to reset the firmware, reload the
config, and restart the host software.
Printer is shutdown
12:05
Lost communication with MCU 'EBBCan'
12:05
Lost communication with MCU 'EBBCan'
Once the underlying issue is corrected, use the
"FIRMWARE_RESTART" command to reset the firmware, reload the
config, and restart the host software.
Printer is shutdown
12:05
Lost communication with MCU 'EBBCan'
12:05
Lost communication with MCU 'EBBCan'
Once the underlying issue is corrected, use the
"FIRMWARE_RESTART" command to reset the firmware, reload the
config, and restart the host software.
Printer is shutdown
12:05
Lost communication with MCU 'EBBCan'
12:05
Lost communication with MCU 'EBBCan'
Once the underlying issue is corrected, use the
"FIRMWARE_RESTART" command to reset the firmware, reload the
config, and restart the host software.
Printer is shutdown```

Is it a faulty tool head board ?.

I've removed Nice=-10 from klipper.service

Same issue no matter what i do! Lost communication with MCU 'EBBCan'
Once the underlying issue is corrected, use the
"FIRMWARE_RESTART" command to reset the firmware, reload the
config, and restart the host software.
Printer is shutdown

12:17 PM
Klipper state: Disconnect
12:17 PM
Probing failed due to printer shutdown
12:17 PM
Probing failed due to printer shutdown
12:17 PM
Lost communication with MCU 'EBBCan'

This is ridiculous. It is not wires the failure is at exactly the same point. And a wire issue would be intermittent. Yes would fail but not at exactly the same place every time !

There must be a better way to figure the issue.

Is there a way to interigate the connection?

I feel it's like a power drop out!

[lyric sable]

Silly question. If there is a communication issue why does it carry on ?

Now this

Printer is not ready
The klippy host software is attempting to connect.  Please
retry in a few moments```

Doesn't change! Has this got something to do with the can bus connection issue right at the start

I have to be logical there has to be a better way to figure this out

@pulsar topaz sale issues and even random disconnect and firmware will not restart. I did update klipper the other day. Maybe I missed something on update

I haven't updated configs etc below is current versions

klipper
v0.13.0-328-g47080385
KlipperScreen
v0.4.6-12-gb3115f9b
led_effect
v0.0.14-0-g83e8c46a
mainsail
v2.14.0
mainsail-config
v1.2.1-1-gff3869a6
moonraker
v0.9.3-121-gf2bb93ff

[pulsar topaz]

Your log will show why it's not starting

[lyric sable]

Klipper doesn't start and then after a while that pops up

Think I'm going to hard wire to toolhead board and avoid connector

[digital ocean]

wiring issue. it cannot talk to the toolhead

[lyric sable]

@digital ocean I've stripped out the canbus cable outer and checked wires you are correct there was a micro fracture

Did simplest thing to do strip wires out and test by gently pulling wires

[lyric sable]

im slowly getting beaten by the mantra Keep It Simple.

[digital ocean]

It helps!

[lyric sable]

But i feel its important i put my hand up when its my fault. Appreciate the guidance.

[kindred whale]

when putting in new wires into the chain make sure they have some slack and restrain them before and after the chain. that way they will have space to move around and not get pinched easily
What kind of wire are you using by the way?

[lyric sable]

i alwsy give strain relief. The damage was 6 inches alone cable inside covering, covering had no evidence of damage let alone a scratch the micro fracture was as the point of manufacture! all fixed now.

[kindred whale]

that's why i ask what kind of wiring do you use, if you use "hard" wiring or housing which isn't meant for movement it will develop these micro fractures very easily

[lyric sable]

I used the canbus cable provided with repaired wire.

[kindred whale]

sadly most provided cables aren't really rated for cable chains 🙁

the small bend radius of the cable chain is what generally kills cables

[digital ocean]

You want multi strand for sure.

If it's solid core it's just a matter of time

[kindred whale]

i mean the btt cable should be multi strand, but the housing itself (of the wires or overall cable) can cause issues if that's not flexible enough

[lyric sable]

@digital ocean multi strand just 6 inches down the cable it was broken

[digital ocean]

K

[lyric sable]

Ultra stupid question I'm struggling to get z offset dialed in for klicky probe. In the rp2040 config there is a z offset parameter. I keep needing to lower the print head during start on the fly to get the right height. The current offset is 5.8mm if I'm lowering the nozzle for right height does that mean I increase or decrease the z offset?

[kindred whale]

Move that probe config block to your printer.cfg and then you can use probe_calibrate. If you hit save and save_config Klipper will update the value accordingly

Why to move that block to the printer.cfg? Klipper won't change these things in included files, it only does that automatic "magic" when the value is your printer.cfg

[lyric sable]

@kindred whale had issues with it saving but do it manually. Just need to I ow if I should increase or decrease the value when I -z adjust

[kindred whale]

I was about to answer to question 🙂 higher probe offset means the nozzle will go lower/more squish

[lyric sable]

So if the current is 5.07 and I adjust -0.07 then I change to 5.00

[kindred whale]

I think you need 5.14 then

With what I mentioned above you don't have to think and just hit save

You don't use a z endstop right?

[lyric sable]

I had issues with saving previously as the probe had to be in the tool head config or it wouldn't work

[lyric sable]

Correct I'm using klicky probe as endstop

[kindred whale]

Good that's correct then

[lyric sable]

Was getting miss matches and issue using original end stop

[kindred whale]

Jeah it can be annoying, klicky is accurate even changing chamber temps unlike the omron probe and gunk on the nozzle won't impact it

[kindred whale]

You may haven't moved the whole block?

[lyric sable]

Yes it's basic but highly repeatable

[lyric sable]

What's whole block?

[kindred whale]

Basically [probe] and everything under it

[lyric sable]

I can copy it across an mainsail will know what the ebbcan gpio is without have to use any kind of redirect code? What the point in have tool head cfg if is all readable on main printer cfg

[raw tartan]

multiple config files is strictly an organizational tool

[kindred whale]

I think it's only necessary that referenced below the include line of the config so the MCU declaration is above referencing that

[raw tartan]

klipper doesn't really care about it (other than the save_config thing)

[lyric sable]

@kindred whale ok will give it a go bit it's not what was recommended with toolhead instructions

[kindred whale]

I like to have everything pin related in the printer.cfg and just put the macros in an included file

[lyric sable]

So I could copy everything out of toolhead configuration and paste to printer config so when I do Input shaping and PID it will save to right place

[kindred whale]

For most people it's easier to just include it.

[kindred whale]

Yes 🙂

[lyric sable]

Ok but if I put EBBCan: gpoi22 for probe in printer config under probe section it will work fine ?

[raw tartan]

might need to check your spelling on that

but yes

[lyric sable]

@raw tartan yeah I'm not at pc just on phone

[raw tartan]

that's the whole reason for the 2 part name: so that klipper knows both the mcu and the pin (which is all it needs to know)

[kindred whale]

pin: gpio23 referenced the pin in the MCU with no name. Mostly the "main" MCU.

pin: can:gpio23 references the mcu named can

[lyric sable]

Ok cool I'll get rid of toolhead configbas it's confusing going back and forth

So pin: EEBCan:gpio22

[kindred whale]

Something that often happens: someone changes a thing in the toolhead config, but there's the same line in the printer.cfg. the line which is lower overwrites the one before

And they are wondering why nothing is changing

[digital ocean]

Unable to retrieve can bus uuid

[lyric sable]

I've copied all functions from tool head config to printer config. Looks like extruder is now under extruding. What is the default e steps for stealth burner again.?

[eager hill]

22.something. Set it to 22 then calibrate them.

[lyric sable]

The small clockwork style stepper is that 16 microsteps?

@eager hill just se to 22 and it's hardly moving

Is the tmc2209 bit at bottom needed?

[extruder]
step_pin: EBBCan:gpio18
dir_pin: !EBBCan:gpio19
enable_pin: !EBBCan:gpio17
microsteps: 16
rotation_distance: 22
nozzle_diameter: 0.400
filament_diameter: 1.750
heater_pin: EBBCan:gpio7
sensor_type: EPCOS 100K B57560G104F
sensor_pin: EBBCan:gpio27
#control = pid
#pid_kp = 26.213
#pid_ki = 1.304
#pid_kd = 131.721
#pid_Kp=35.542
#pid_Ki=8.171
#pid_Kd=38.651
min_temp: 0
max_temp: 300

[tmc2209 extruder]
uart_pin: EBBCan:gpio20
run_current: 0.650
stealthchop_threshold

Asked for 125mm to be extruded but it's only extended 26.25mm

[eager hill]

Oh I see. You’re missing gear_ratio: 50:10

[lyric sable]

Where is that meant to be ?

[eager hill]

Microsteps don’t matter. 16 is fine though. That won’t change the distance moved.

[eager hill]

Anywhere in the [extruder] config. Id recommend below the rotation distance line.

[lyric sable]

That's far better

New rotation is 22.5192 which from what I gather from previous comments is about right

[eager hill]

Yep. Sounds about right to me

[digital ocean]

How did that end up falling in the floor from a copy paste

[lyric sable]

@digital ocean no idea but was a direct copy from tool head cfg to printer cfg. Lots of issues and now have to re calibrate again

[kindred whale]

You might had the config scattered and delete the part which wasn't present in the other config

[lyric sable]

@kindred whale no I used the default toolhead cfg that it came with. If I removed any code it would be commented out not deleted

[kindred whale]

Yes, we figured out that you were missing the gear ratio.

My guess is that your config was "read" from two different [extruder] sections.

If there is a line like for example rotation_distance in both section, the read latter is the one which gets used.

If the btt config was missing the gear ratio but it was present in the your printer.cfg extruder section it would have been used.

If you delete the section in the printer config the gear ratio uses default value of one because the config isn't telling Klipper to use a different one

[lyric sable]

The tool head was working fine with it's own cfg file as that didn't need the 50:10 comment

The printer cfg extruder was uncommented so would not be used.

Is the tmc2209 bit at bottom needed?

[extruder]
step_pin: EBBCan:gpio18
dir_pin: !EBBCan:gpio19
enable_pin: !EBBCan:gpio17
microsteps: 16
rotation_distance: 22
nozzle_diameter: 0.400
filament_diameter: 1.750
heater_pin: EBBCan:gpio7
sensor_type: EPCOS 100K B57560G104F
sensor_pin: EBBCan:gpio27
#control = pid
#pid_kp = 26.213
#pid_ki = 1.304
#pid_kd = 131.721
#pid_Kp=35.542
#pid_Ki=8.171
#pid_Kd=38.651
min_temp: 0
max_temp: 300

[tmc2209 extruder]
uart_pin: EBBCan:gpio20
run_current: 0.650
stealthchop_threshold

Do I also need the [tmc2209 extruder] section

[kindred whale]

Yes

[lyric sable]

Ok

[kindred whale]

That's what I'm trying to say, I think it was present in the printer.cfg

[lyric sable]

No was uncommented as there was a conflict. Mainsail was using code from toolhead cfg to control extruder

[kindred whale]

Ok.

[lyric sable]

Either way it's work but not accurate so will carry on with calibration etc.

[kindred whale]

Do you have a copy of the state of both configs before you moved them together? Might be easier that way to get the same calibration state like before?

[lyric sable]

I do but I've just gone through the calibration and Voron cubes etc. I will have to do that anyway. I've beend ry all my filaments for better accuracy

[lyric sable]

I've now come across belt issues as tension not there. I have the red two trees tensiometer. Do we have a table of expected deflection in mm as I cannot find anything online.

[warm sandal]

that's just a glorified tyre tread depth gauge and is no good for setting belt tensions

the pfmakes gt2 belt tension meter is widely used for tensioning belts

you can buy a kit from onetwo3d

[lyric sable]

thanks for heads up but its what i have in front of me is there not a table of data for it

[warm sandal]

no, because it's useless

[lyric sable]

ok no worries.

[warm sandal]

if two trees can't supply data for it, what hope do we have?! 😆

[warm sandal]

have you tried the frequency method using a phone app?

[lyric sable]

i get that but a massive amount out there is like that anyway

[lyric sable]

no becasue that is very variable too

[warm sandal]

a lot of what gets sold by the chinese companies is just junk, and as soon as one vendor comes up with an idea, all the others copy it

[warm sandal]

it doesn't need to be spot on, just in the ballpark

[lyric sable]

then whats the point ? are we not trying to tune these things, i can do it by ear if its ball park

[warm sandal]

these aren't super accurate machines, it matters more that all the z belts match each other and the a/b belts match each other

[lyric sable]

well from the aspect of tool making yes they are miles out but as for repeatability and quality its best to do all you can, otherwise whats the point of calibration, shapers guages etc. but ill use this to set equal tension and got from there.

[warm sandal]

at the end of the day, we're building a simple robotic hot glue gun

you basically want enough tension to prevent the belts from skipping teeth on the pulleys/idlers, but not too much to put unnecessary strain on the stepper motors and other motion components

if it's 100hz or 120hz when it's supposed to be 110hz, that's fine, if it's 80hz or 200hz, that's a problem

[lyric sable]

absolutly agree but my belts were very loose compare to other machines i have. i agree th guage is basic but it still outputs a repeatable value. so i will use it to at least make sure all the best are within a similar range,

[warm sandal]

i bought a couple of different "digital belt tensioning tools" from various vendors to test out, none of them were what i'd call repeatable

at least with the pfmakes tool, there's data behind it and has been proven to provide good results

[lyric sable]

@warm sandal well with a tolerance

I'm getting withing 0.02-0.05 mm variation

[warm sandal]

question is, what does 0.02-0.05mm variation mean in terms of tension? 5g? 100g? 2000g?

[lyric sable]

The belt holding part of the guage was floating or moving away from guage it's self. The pivot screw was useless as holding guage in place. So glued 3d printed part and guage to prevent unwanted movement. The gauges measures 8mm which equate to 2mm to 2.2mm deflection from what I can see

[warm sandal]

that just goes to prove my point, useless junk

[lyric sable]

ok fair enough. but its measuring the guage deflection. Better than nothing

[warm sandal]

but we don't know what that equates to in tension, it may be not enough, it may be too much, who knows? 🤷🏼‍♂️

arguably, the only thing that meter is good for is comparing belt tensions, not setting the initial tension

[lyric sable]

well we shall see. the reality is the tool has the be used differently.

I would want to know what the above guage is related to actual belt deflection in mm no just a guage value

[warm sandal]

that's the $64,000 question

[lyric sable]

Not really

I'm getting 0.85mm actual belt deflection

[warm sandal]

again, is that too much tension or not enough?

[lyric sable]

But I would like to know what the creators of the above are getting

@warm sandal what does 1.9 actually mean on the above guage

[warm sandal]

they probably didn't even test it, just copied another chinese vendor

[lyric sable]

@warm sandal yes I agree but it's still able to measure a value.

[warm sandal]

yeah, sure, you can measure a value, but what does it mean?

[lyric sable]

So what does 1.9 mean on the above guage

[warm sandal]

i don't know, and i doubt two trees know either; otherwise, they'd have provided the data

[lyric sable]

No I mean the pfmakes gauge

[frosty tangle]

Pounds of tension

[lyric sable]

@frosty tangle thank you and how are pfmakes getting that measurement. By using a calculation based on the deflection of guage correct?

[warm sandal]

there's a table of values on their github

[lyric sable]

@warm sandal correct but how are the getting the values

[warm sandal]

by lots of testing? 🤷🏼‍♂️

[lyric sable]

I see a linear movement turned into a rotation movement.

@warm sandal not I disagree. It has to he calculated otherwise it's not repeatable

So again using the pfmakes guage what does 1.9 mean in mm of belt deflection

[warm sandal]

well, i'm not the designer, so i can't comment, but it's widely used in this community and shown to work

[lyric sable]

@warm sandal ok so the question that has to he asked is how are they getting the lbs of tension related to deflection of belt. Because that's how we measure it.

[warm sandal]

my guess is, they're probably applying various known amounts of force to the belt using weights and then measuring it with their tool, along with data sheets from gates

[lyric sable]

@warm sandal I absolutely agree but it still comes down to belt deflection so plenty as. Using a spring to apply a variable calibrated pressure to then convert a linear motion to a rotational gear dial. Tonnes of variables there. So I would want to know what is the deflection needed for the measurement. Because the crappy Chinese tyre guage can to that quite easily and repeatable.

@warm sandal then there is absolutely no reason to calibrate the tool if it's guess work

[warm sandal]

ok, you crack on with your tyre gauge

[lyric sable]

@warm sandal I shall see

[warm sandal]

it's not guess work, the pfmakes tool is proven by many people here to work

[lyric sable]

@warm sandal no in an engineering sense

Back lash of the making gear would be my first concern.

I'll see what I come up with this crappy Chinese guage

[warm sandal]

i'm not an engineer, i can't answer these questions

[lyric sable]

@warm sandal I can

I would also want to know the tolerances

[warm sandal]

you're overthinking it

[lyric sable]

Again I'm curious so want to see what the differences are mad why

@warm sandal no I want to see how the crappy Chinese took compares

There are far less variables with the crappy Chinese guage regardless of use etc

I'll get back to you

[warm sandal]

well, i've given you my opinion of the chinese gauges, as well as mine and the opinion of many here of the pfmakes tool, do with that what you will

[shut comet]

If you look at the manual, there is an alternative way to calibrate by hanging a weight with a belt and making sure the gauge reads the right number. That correlates tension to a number.

[lyric sable]

@warm sandal agreed.

@shut comet I have seen. But it's all down to deflection as that's how the belt is measured. I'll look into it

[shut comet]

Exactly. The spring provides a linear force and the tension on the belt is the counter force.

[lyric sable]

I get how it works. But what does 1.9lbs equally to in belt deflection

[shut comet]

It depends on the span that you measure it across.

The PFmakes tool measures the deflection between the two pins so it is consistent.

[kindred whale]

The pfmakestool has alternative calibration guide suspending a jug of water on a specificed length. You could use that with your belt meter as reference

[lyric sable]

Yes I can set up all sorts of experiments to determine a value. Bit the question stands. They says 1.9 is the acceptable value for belt tension but what is the in relation to belt deflection because that's what's being measured

[kindred whale]

Exactly what you measure there

3lbs over 250mm is the go-to value. At least you're then in the ballpark

[lyric sable]

Ok again we can come up with a million experiments etc. Bit the question stands what does the guage reference of 1.9lbs equate to in mm of belt deflection

[kindred whale]

We don't know that. You could look into the GT2 sheet and there might be formula.

I was trying to give a simple test to measure that with your tool with a known good value

[shut comet]

I think you are looking at the values the wrong way. We want tension to be set a deflection is just one way to measure the tension in the belt.

If you consult the datasheets for the belts, tension is what is speced.

[lyric sable]

@shut comet yes I see the specs but there is a linear mechanism converting to a rotation. So the Linear has to be defined but a value which is the distance the belt deflects i.e. 1mm of deflection equals 1.9lbs of tension.

[shut comet]

The numbers on the PF makes gauge are arbitrary from memory and is based on the spring rate and calibration.

I mean, the tool works and is relatable and easy to use. 🤷‍♂️

[lyric sable]

But we have not data on why it works or how

And without data nothing is reliable

I'll email

[warm sandal]

at the end of the day, i really don't care about the numbers

i use the tool to tension my belts and i get great looking prints, that's all that matters to me

[limber viper]

By measuring all of the weights suspended from a length of belt with the meter in quarter lb increments and developing a tension table from there. also used a force tension gauge in a jig to confirm the results. then used a gates sonic tension meter to verify those results

[lyric sable]

@warm sandal no worries but I do. So you are welcome to no respond

[warm sandal]

thanks for the tool! 👍🏼

[limber viper]

many springs were tried to find a rate that was as linear as possible for the given belt tension range

[lyric sable]

@limber viper thanks you for jumping in.

I get the tensions and gears etc. so are we saying this is guess work.

[warm sandal]

they're the designer of the tool, listen to them

[warm sandal]

i have more important things in life to worry about

[lyric sable]

@warm sandal no worries take care

[shut comet]

I’d call this engineering work…. #1420505187524804649 message

[lyric sable]

@shut comet I agree

[limber viper]

My original prototype was a vernier gauge version, but I had to use gears to increase resolution, each tick is ~0.09mm of linear deflection

if memory serves

[lyric sable]

@limber viper in getting 0.85mm deflection

Thank you for jumping in and taking the time to answer

[limber viper]

so the issue with clones is that they have no way to calibrate (adjust preload on spring) since no two springs are the same this becomes an issue

[shut comet]

With the PF tool or another tool?

[lyric sable]

@limber viper I agree but that can be figured.

[warm sandal]

a glorified tyre tread depth gauge

[limber viper]

yeah, you can use my manual, the section for alternate calibration. you can develop your own tension table

[lyric sable]

The question I was hoping to find out was with your guage 1.9lbs equates to what in mm of belt deflection

[limber viper]

I would recommend using a water jug and an accurate scale if you go that route

[lyric sable]

I do believe this can be calculated accurately with math

[shut comet]

The center distance may not be the same and the spring may not be the same so they could be wildly different values of tension.

[lyric sable]

Hence why I've been asking continuously what is the deflection with your guage at 1.9lbs

@shut comet correct but math can figure it out

[limber viper]

you would need to test the exact spring rate for the spring that is in your gauge for the math to work out

[lyric sable]

Yes I agree

But I wanted to know 1.9lbs = X mm of belt deflection

Apologies I tend to want to understand how things work and why rather than accepting because someone says

@limber viper what is the compression force required for the main spring. Because the calibration to are performing is just adjusting the spring preload as I see it

[limber viper]

no apology needed, I can probably get this info for you when I'm at home (at work)

[limber viper]

6mm OD x 0.6mm wire dia. x 50mm free length compression spring with 20 active coils (0.35N/mm spring rate)

and also since each spring is slightly different the preload needs to be calibrated

[lyric sable]

@limber viper there is no rush this is curiosity and understanding more than anything

[lyric sable]

See above to the comment about preload

[limber viper]

completely understand, life is demanding without understanding 😅

[lyric sable]

I get a little annoyed when people say that's pointless when ive grown up with calipers and micrometers from a young age. Yes the crappy Chinese guage is that it is but it still makes an accurate measurement. But I would prefer to understand why and how your works. But yes the calibration is just spring preload. I'm not saying anything is better than yours but it's the lateral to rotation which will have plenty of backlash compounded by the inaccuracies of 3d printing. I've always been told what calibrates the calibrator.

[limber viper]

correct, preload is adjusted to hit a a predefined mark at a given weight/deflection be that belt w/ suspended weight or music wire

[lyric sable]

@limber viper yes but ultimately regardless of tool you are measuring deflection

So given a spring of a set force that has preload adjusted will deflect a set length of belt and that deflection is indicated by your guage 0-3

@limber viper I know you are at work but get back to me when you can but I'm curious to understand more

[shut comet]

No, deflection is the end result of the spring force and the tension on the belt over the distance. Add more tension and the deflection will be less.

[lyric sable]

@shut comet absolutely agree

[shut comet]

If you have a weight, you can calibrate your gauge the same way.

[lyric sable]

But the force of the spring is set by preload so it's a specific defind reference but the guage is measuring the deflection

[shut comet]

Anyways. I’m board a plane for VICE so I’m signing off.

[lyric sable]

@shut comet no worries take care

[limber viper]

happy to discuss in DM so as not to veer too far off topic here. I also have to bounce so won't likely be able to reply until monday sometime

[lyric sable]

@limber viper give me a shout on DM when you are next free. Thank you for taking the time

[versed dove]

If you're looking for the math I'm pretty sure gates design manuals will have it (there's 3-4 for gt2 I think and one of them has it elaborated)

https://www.gates.com/content/dam/documents-library/catalogs/light-power-and-precision-manual.pdf

this has it on page 65-66 but isn't the one I'm thinking of

[lyric sable]

@versed dove thank you

[lyric sable]

Is there a common issue with filament delivery. I se to be getting resistance as filament goes from spo to tool head. Any mods ?

[frosty tangle]

the most common thing we see is people using too small an internal diameter PTFE tube for the reverse bowden

other issues are too tight corners

[lyric sable]

I'm using the wide internal diameter PTFE. But still getting issues. It's physically difficult moving the filament through Bowden. Any chance you have image of your tube route.

[frosty tangle]

i can't find a pic and my printer locations aren't conducive to photography

[versed dove]

usually it's the spool holder/roller that has the resistance ime

cut a section and see if its a lot of resistance still

[lyric sable]

@frosty tangle no problem

@versed dove even with a section of just filament in PTFE and no spool there is a ridiculous amount of friction which should be the case as the tube is ptfe

[versed dove]

are there any bends that are sharp like where it enters the rear panel?

[lyric sable]

I have a loop as there isnt any other way to get tube to go in a different direction

Hence why I ask of there are and pics?

[versed dove]

send a pic of your setup

(I don't have a 2.4)

[lyric sable]

How is it meant to go?

I'm going to convert to vertical feed recert to a vertical feed as 90° from Voron fram.is way too tight

[raw tartan]

You're not supposed to jam it up against a wall...

There's a bunch of other nonsense on the back of this machine, but..

[lyric sable]

@raw tartan it's not it's free moving just the angle. Is there a vertical mod?

[raw tartan]

Top feeding a v2 sounds great until you try to print something tall, then it gets really....bad...

(Though I admit, I have been known to do it anyway, when printing soft tpu)

I'd really suggest just getting the printer farther away from that wall, so you can route the tubing more cleanly first

Also, I think your ptfe is sticking out too far over your spool holder, which may be causing some oddities

[lyric sable]

I can't understand why there isn't a plastic guide for the 90° bend.

[raw tartan]

Do you see how nicely mine is sitting?

I don't have any guides in there