Engines hum and heat up a lot with Kraken and RatOS2.1
I have upgraded my VC3.1 24/48v printer from the Octopus PRO to the Kraken board and installed RatOS2.1 and now the motors keep humming all the time and get very hot. So much so that after a short time the 48v source goes off and I have to restart it to make it work.
I've reinstalled it twice and it's still doing the same thing.
I've bought the same new motors directly from RatRig and it's still doing the same thing.
Can somebody help me?
Thanks in advance.
108 Replies
All the time? Even when the motors are turned off?
No, when I use it and when they rest after using them. If I turn them off it doesn't do it.
If you use the page that lets you see the accelerometer info live, can you see it on the graph
Yes
I have the Z motors disconnected and the vibration is still transmitted through the structure.
I have a similar problem, also a new BTT Kraken on a Vcore 3.1 and 48V and Ratos 2.1. I switched from a Duet3d HC6 with RRF because i wanted the Beacon. Z-motors seem to work, although with a sound i didnt have before. X and Y moved a couple of times and it homed. But now when i try to move X or Y, Klipper shuts down and i get an error message: "tmc 'stepper_x' reports error: gstat: 00000004 uv_cp=1(undervoltage!)". I have checked my cables and connectors, also reverted back to 24V for the drivers, no luck. I noticed that the 4 first TMC2160-drivers on the Kraken have a sense_resistor value of 0.022 and the other 4 drivers 0.075. This is probably part of the problem. I have now measured the resistance in my X and Y steppers and it is about 1,5 ohms in one winding and 0.5 in the other, so i will need new steppers... I will try again with new ones, but it seems that the config for the BTT Kraken isn't correct in Klipper. In the printer.conf it says that the stepper drivers are 5160 and not 2160, but this is maybe just a copy/paste thing? I have only just started with Klipper, coming from RRF, and it still a bit confusing...
I will check it as soon as I arrive home. But I believe the configs for 2160 are exactly the same as for 5160. @miklschmidt can confirm this…
Also, the problem with the kraken is not on the configs. There is a hardware issue of it running nema17 motors… the board was designed to run bigger motors. Exactly what you said about the sense resistors for X and Y.
My advice: do not run it with nema17 until someone else (probably Mikkel) chime in with some info
this is just ADXL345 noise
When I get home I record a video but you can feel the vibration of the motors throughout the structure, perhaps in the graph it looks like a problem with the ADXL but it is something in the configuration of the 2160. Additionally, the motors get very hot.
Klipper shuts down and i get an error message: "tmc 'stepper_x' reports error: gstat: 00000004 uv_cp=1(undervoltage!)"That's usually a wiring or power supply problem: https://www.klipper3d.org/TMC_Drivers.html#tmc-reports-error-uv_cp1undervoltage
I noticed that the 4 first TMC2160-drivers on the Kraken have a sense_resistor value of 0.022 and the other 4 drivers 0.075. This is probably part of the problem.Yes, those 0.022 ohm sense resistors do not make sense for anything below ~5A RMS. I've spoken at length about how i think this board is a mistake design wise for 3d printers. A sense resistor that low will force a super low currentscaler which results in bad microstepping performance. The latter 4 should be fine for 1.5-2.5A RMS.
I have now measured the resistance in my X and Y steppers and it is about 1,5 ohms in one winding and 0.5 in the other, so i will need new steppers...... So somehow your coil soft-shorted / hotspotted? Explains the undervoltage error. I wonder how that's possible though, but i guess the sense_resistor is just too inaccurate at that low current.
but it seems that the config for the BTT Kraken isn't correct in Klipper.As far as i can tell, it's correct. What about it do you believe to be incorrect? There is no 2160 section in klipper, a 2160 is just a 5160 without the motion planner which klipper doesn't support anyway, hence why klipper instructs you to use the 5160 section (you can see this in the official btt config as well).
I will check it as soon as I arrive home. But I believe the configs for 2160 are exactly the same as for 5160. @miklschmidt can confirm this…Correct. It is normal that they hum when standing still, they're receiving the full current (as instructed by the klipper docs, run_current = hold_current) at all times, what you're hearing is the chopper driver trying to keep the motor still at a specific microstep. If you want it to shut up, enable "standstill stealth" in the motion section in the hardware configurator. You'll notice if you move the motor, the hum will change, because it'll hold at a different microstep position. All normal.
The vibration matters little to me, the real problem is that the motors get so hot that it triggers the protection of the power supply and turns off.
Power supply does not have any protection against motor temperature. Only overcurrent / overvoltage etc, the usual.
Can you measure the stepper temperature? and what PSU are you using?
Also, please upload a debug.zip (download it from the sidebar in the configurator)
Sorry for the delay. The power supplies are Meanwell LRS-200, one 24v and one 48v. I have sent the Debug.zip file
It looks like you didn't update RatOS and the configurator, that's why you're having problems
Currently double checking if there's an issue.
Yeah you're missing updates
Thanks for your answer. I don't have any pending updates. Is there a way to install it manually?
or do I directly edit with an overdrive?
Click the refresh button in the update manager panel in the machine tab
You can add an override to printer.cfg but that's not necessary the code will generate correctly. This bug was fixed a month ago.
😱 no. There's another bug.
the sense resistor values were emitted as part of the preset. If there was no applicable preset, it wouldn't render, that's why it's missing.
Shit
I'm fixing right now.
Thank you, it's a relief to know.
Update is ready!
Sorry about that, that's my bad.
I have updated and it is still the same. I'm sending you the new Debug.zip. The motors and even the motor cables become very hot.
After updating I restarted the machine and tried again. Same result. I have increased the amps from 1.1 to 1.768 which is recommended, it does the same thing. Obviously with 1.1 it takes longer to heat up but it does heat up.
Are you sure you updated?
RatOS.cfg hasn't changed
Just checked your moonraker.log, and indeed you're missing 2 commits, you haven't updated yet
(it's the ratos-configurator) package that needs updating.
I got two files to update when you told me and I updated both. I checked to see if there were any more updates and I didn't get any more. It may have been too fast. Now I’m out of my house. Tomorrow I will check it. In any case, thanks for your help.
Aight let me know, you should get
sense_resistor: 0.022 in the X/Y driver definitions in RatOS.cfg, if that doesn't happen, i done goofed somewhere.Hello again. I'm finally back home and I can spend some time with the printer. Sugue without appearing the sense_resist in engines x and y. However, it does appear in z engines.
Gimme a sec, will check everything again.
I don't think you updated:
at least i can't reproduce the problem
I don’t see nothing to upgrade
click the refresh button
I know
Nothing
send me the debug zip
again lol
I know what not to do. But what happens if I put those lines?
It'll work. But we wouldn't have fixed anything, we would just have worked around the problem.
Clearly there's still a bug somewhere
So other people may have the same problem, and it's a very bad issue since it causes insane current levels on the motor
Yeah apparently something is still wrong with the configurator for your particular setup, but i don't know what without the debug-zip
So please send a new one
What surprises me is that no one in the entire world has complained. Could it be that no one uses the Kraken board or they make their own firmware.
Very few use the kraken, it's not really suited for 3d printers. Those first 4 drivers are way overspecced. But that's besides the point. There are actually another user who is having the same problem as you (albeit with the BTT 5160T Plus extenal drivers which are similarly overspecced), but they're not responding so i don't know if the latest fix i pushed fixed it for them.
Oh! Can you try going through the hardware configurator again?
I've already sent it to you again. I don't know if you can see the history, but I have tried changing the value by hand and then deleting it again by hand.
Yes, give a minute
You should never modify RatOS.cfg, always put stuff in printer.cfg
(it says that at the top of the file)
X and Y should look like this
Notice the "PLUS" at the end
you don't have to flash your boards if that's what you're doing... sorry i should've told you that :facepalm: .. You can juts click the "hardware configuration" step in the navigation to the right.
I can’t change
This might be the problem... I wonder why these haven't updated.
Try CTRL + SHIFT + R
That will force the cache to clear and reload
You're not supposed to be able to change them, as the drivers are integrated
But the drivers it thinks are on those ports are wrong
🤔
If necessary I can do a clean installation from scratch. But even if it solves my problem, we would still not know how to solve the problem for other people.
did you try
CTRL + SHIFT +R ?Yes
Same result
If necessary I can do a clean installation from scratchNo that would not help figuring out what's wrong If you go to the hardware configurator, select a different board and then immediately select the kraken again, then go back to that page, does that change it?
That’s is new
It did something.. But what does the drivers say now?
in the motion page
24v for all
Same drivers
reload again
there's some kind of fuckery going on. Maybe the configurator wasn't restarted when you updated the configuration or something.
I updated the board again to try something different and it does the same thing.
Go back to mainsail and click this
then go back to the motion page, and press CTRL + SHIFT + R again (i want to be sure we're not dealing with an old browser cache)
👍
I don't know why my iPhone doesn't focus now.
That's okay, i can see the PLUS is still missing 🙂
I have no idea why, but i'm looking into it.
If only i could reproduce the problem, that would've made it a lot easier
If you want I can give you a teamviewer access to my laptop
Appreciate it! I have a few things to try out first, so no need just yet 🙂
Ok. Good luck
I have it reproduced on a printer now, will let you know when it's fixed!
👏
@PabloMirón Fixed! ... Problem was my damn deployscript.. Code was fine, just never got deployed even though it looked like it did (it deployed from the wrong branch)... :facepalm:
Great, I'm very happy to know that it is now fixed. I'll try it this afternoon and tell you. thank you
I will also try it later today!
Ouuhhh yeaaaah!
But now I have another problem😭
this is just wiring, swap the middle pins on the board side connector on your stepper cables for x/y
I'll try and tell you. One of the things I had to do when going from the Octopus to the Kraken was change those pins. But because in the electrical diagrams they are different.
yup, exactly. Swap the two middle pins and you should be good to go, short of flipping the dir_pin in case it's running in the wrong direction
I'm having all the problems in the world, it's super difficult to remove the pins, a connector broke and when I tried to do a HOME everything restarted.
The photo is of a cable as it comes standard.
Call me useless, but if I swap the red and green it resets everything.
resets everything? What do you mean?
If the bigger connectors for x/y are a bitch to repin, repin the JST-PH on the other end instead.
Hello again, yesterday and today I will be on a work trip.
But I will explain the problem when exchanging the cables:
The motors are two phases, one is the two cables on the right (A1 and A2/Black and Green) and the other is the two on the left (B1 and B2/Red and Blue) .
If I swap the middle ones (A1 and B2/Red and Green) I mix the two phases and that causes a short circuit that causes the board to reset.
Before the software update, it was working, vibrating and getting hot, but it worked. I think the problem is something with the firmware/software.
Don't hate me too much😉
If I swap the middle ones (A1 and B2/Red and Green) I mix the two phases and that causes a short circuit that causes the board to reset.Of course you're right, you should swap the two outer pins on either the left or the right side, the problem is your coils are currently fighting eachother.
Before the software update, it was working, vibrating and getting hot, but it worked. I think the problem is something with the firmware/software.No, software configuration can't do this 🙂
Your wires currently match the default cables that come with the 2504AC's, and thosoe work out of the box with the Octopus, which wants +,-,-,+ but your Kraken wants +,-,+,-
I'm using +/- just to show each end of the coil, of course it has no polarity, but it was the easiest way to show the coil direction, the letters and numbers are not making things easy.
+,-,-,+ == -,+,+,- except the stepper will run in the other direction.
If I understood correctly, if I change the order of only two cables, for example I swap the green and the black, should it work?
I changed the green and black and it does the same thing.
🤔
yes
you need to do it on both x and y
show me your X / Y wires
I changed them on both engines.
It is curious that the three Z axis motors work well and their cables are pin to pin.
There was a new update installed and now the board and the power supply directly jump as soon as I press “home position”
now the board and the power supply directly jump as soon as I press “home position”What do you mean?
At the end of the video you can see that the second power supply, which is the 48 volt one, has been turned off.
And yes, I have installed the screen on the back so I can see both things at the same time.
Your coils are still fighting eachother, but that doesn't explain the shut down. Please send me another debug-zip.
Is uart_pin the same as cs_pin?
In the aliases section of RatOS.cfg we have the same error. The y_uart_pin must be =PD5
What I don't know is if the uart_pin affects SPI mode.
on your board, yes, it's the same pin.
let me check this
It should be
will try and generate a config
indeed, this generates the wrong config.. it's probably looking for the uart_pin and seeing it's not there, then falling back to the board config which is then still wrong.. So two bugs to fix.
@PabloMirón fix deploying, read in 10
Good afternoon. No changes have been applied. Both values remain the same in both files.
Can't reproduce. Are you sure you updated?
Btw, it shouldn't make any difference, the pin map is only for backwards compatibility with 2.0
In about an hour I arrived home and I sent you the debug file
excellent, i probably won't be available until sometime tomorrow though but i'll take a look then!
Don't worry, weekends are for resting.
I have already sent you a file.
thanks will take a look when i get the time 👍