-
Notifications
You must be signed in to change notification settings - Fork 50
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
MKS-Gen L V1.0 w/5 Fystech 2130 drivers for reference #28
Comments
Hi, |
In bugfix branch of marlin software serial has been implemented. And is also recommended by the marlin guys. By using software serial you remove the traffic over the hardware serial line which can cause connection issues over usb on ramps based boards. Also this allows the full use of both the full graphic lcd and the sd card pins without having to hack or splice into anything. Pins 4 and 5 of the servo have been recommended by other ramps users so I just followed that line of thinking and kept my connectors simple, since I am using 5 steppers in spi configuration. I used pin 65 because it’s close to servo pin and also because I had already made my spi connector for the aux3 pinout so I just moved it over to the aux2 connector keeping the same pinout.
…Sent from my iPhone
On Mar 20, 2018, at 9:44 AM, Budavaril <notifications@github.com<mailto:notifications@github.com>> wrote:
Hi,
Why better to use those irregular pins (example servos pins) than using same standard pins (aux2-aux3)on any other ramps based boards?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub<#28 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AB7M2j5jh65TEbxN_nzgunYvd7iB7Wwzks5tgRXjgaJpZM4SvCS4>.
|
That make sense. So may that's why I had a some trouble with my FLsun i3 during TMC2130 upgrade? |
I was going to do the same thing but octoprint can no longer connect reliably with the mks screen attached and the lcd. So I scrapped the touch screen for a RepRap discount full graphic lcd and haven’t had an issue since.
…Sent from my iPhone
On Mar 20, 2018, at 1:16 PM, Budavaril <notifications@github.com<mailto:notifications@github.com>> wrote:
That make sense. So may that's why I had a some trouble with my FLsun i3 during TMC2130 upgrade?
I using my board with "dual screen" .
One of the 3.2" touch screen has SD and WiFi built in, that's why I thought if I using for "second screen" just an (Anet Discount) 12864 what has no SD feature I should have no any those problems to changing my drivers to TMC2130.
But may yes .. Because I tried first using Marlin v1.1.8 and library TMC2130 v2.2.1 but that combo wasn't succesful. Motors just moved shortly with strong noise and I got few messages:
"00:00: X driver overtemperature warning! (0mA)"
"X driver error detected: short to ground (coil A)"
"Error:Printer halted. kill() called!"
I thing there is some bug and the "AUTOMATIC_CURRENT_CONTROL" (What called "MONITOR_DRIVER_STATUS" in v1.1.8) does nor really worked and my motors got (uncontrolled) maximum current.
After I downgraded to Marlin 1.1.6 and TMC2130 2.0.0 looks everything working. But I still have to troubleshooting because I still loosing some steps on X and Y.
I thought that can be solving with some "current-speed-acceleration tune". But may there is a communication problem?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub<#28 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AB7M2png9qLBkC3Oh8qxfN2di6MKnq-Fks5tgUeTgaJpZM4SvCS4>.
|
If you are still using hardware serial then you are flooding the serial line with information. I’ve found that most of the x any y skipping I’m having has been caused by inadequate cooling of the drivers coupled with fast acceleration/velocity. I can cause the skips right now by either enabling interpolation or steps or by either increasing my acceleration or velocity. I am currently installing some stack headers and more heatsinks and am trying to mitigate excessive wires blocking airflow. I should be able see some improvement by this weekend so I will update once I get my thermal tape in the mail.
…Sent from my iPhone
On Mar 20, 2018, at 1:16 PM, Budavaril <notifications@github.com<mailto:notifications@github.com>> wrote:
That make sense. So may that's why I had a some trouble with my FLsun i3 during TMC2130 upgrade?
I using my board with "dual screen" .
One of the 3.2" touch screen has SD and WiFi built in, that's why I thought if I using for "second screen" just an (Anet Discount) 12864 what has no SD feature I should have no any those problems to changing my drivers to TMC2130.
But may yes .. Because I tried first using Marlin v1.1.8 and library TMC2130 v2.2.1 but that combo wasn't succesful. Motors just moved shortly with strong noise and I got few messages:
"00:00: X driver overtemperature warning! (0mA)"
"X driver error detected: short to ground (coil A)"
"Error:Printer halted. kill() called!"
I thing there is some bug and the "AUTOMATIC_CURRENT_CONTROL" (What called "MONITOR_DRIVER_STATUS" in v1.1.8) does nor really worked and my motors got (uncontrolled) maximum current.
After I downgraded to Marlin 1.1.6 and TMC2130 2.0.0 looks everything working. But I still have to troubleshooting because I still loosing some steps on X and Y.
I thought that can be solving with some "current-speed-acceleration tune". But may there is a communication problem?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub<#28 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AB7M2png9qLBkC3Oh8qxfN2di6MKnq-Fks5tgUeTgaJpZM4SvCS4>.
|
Also I had this issue with the overtemp warning at first as well. What ended up fixing that was to make my wiring harness so that each serial line went directly to the spi pins instead of daisy chaining them. Apparently the spi bus cannot handle more than one device if they are wired up like this. So I threw out my first harness and instead opted for twisting all the wires together and capping them with some heat shrink. I intend to design a better plug, maybe a small pcb to sit on top of the pins I need and then breakout the wires from there. So to reiterate, if you have your spi wires going first to one stepper then to the next stepper you will need to change it so that all of the wires from each individual stepstick go directly to the spi header.
…Sent from my iPhone
On Mar 20, 2018, at 1:16 PM, Budavaril <notifications@github.com<mailto:notifications@github.com>> wrote:
That make sense. So may that's why I had a some trouble with my FLsun i3 during TMC2130 upgrade?
I using my board with "dual screen" .
One of the 3.2" touch screen has SD and WiFi built in, that's why I thought if I using for "second screen" just an (Anet Discount) 12864 what has no SD feature I should have no any those problems to changing my drivers to TMC2130.
But may yes .. Because I tried first using Marlin v1.1.8 and library TMC2130 v2.2.1 but that combo wasn't succesful. Motors just moved shortly with strong noise and I got few messages:
"00:00: X driver overtemperature warning! (0mA)"
"X driver error detected: short to ground (coil A)"
"Error:Printer halted. kill() called!"
I thing there is some bug and the "AUTOMATIC_CURRENT_CONTROL" (What called "MONITOR_DRIVER_STATUS" in v1.1.8) does nor really worked and my motors got (uncontrolled) maximum current.
After I downgraded to Marlin 1.1.6 and TMC2130 2.0.0 looks everything working. But I still have to troubleshooting because I still loosing some steps on X and Y.
I thought that can be solving with some "current-speed-acceleration tune". But may there is a communication problem?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub<#28 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AB7M2png9qLBkC3Oh8qxfN2di6MKnq-Fks5tgUeTgaJpZM4SvCS4>.
|
Marlin v1.1.8 has an issue with setting the pin mode for CS too late and that causes problems with the configuration. The bugfix-1.1.x has this fixed. Also it might be clearer to use "serial" to mean "USART" and not "Serial Peripheral Interface". |
is that correct though? every where its referenced that tmc2130 is connecting in spi mode. 2208's use a uart connection. |
I tried using the bugfix-1.1.x (not enabled the SW serial yet)and steppers are working looks well, but if I start a print (from pc or from the touchscreen's SD-USB) the bed and hotend doesn't heating up, just start immediately moving the head looks printing. |
Double check your configuration file. You might have I commented or accidentally deleted something. Try comparing your files to an example for a similar machine and make sure nothing is out of whack. Also make sure you power cycle everything just to make sure it’s not a random glitch.
…Sent from my iPhone
On Mar 22, 2018, at 12:02 PM, Budavaril <notifications@github.com<mailto:notifications@github.com>> wrote:
I tried using the bugfix-1.1.x (not enabled the SW serial yet)and steppers are working looks well, but if I start a print (from pc or from the touchscreen's SD-USB) the bed and hotend doesn't heating up, just start immediately moving the head looks printing.
Any idea what I missed?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub<#28 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AB7M2sbxfuy6LTD9YQoad3Chi0h79nS3ks5tg9k7gaJpZM4SvCS4>.
|
Thanks your quick response. |
And with everything put back together. I may clip those stack headers down in the future. Currently Ive re-enabled stepper interpolation and am running my current at 650 650 650 700 700. Im printing a nice medium sized part with a lot of circular patterns so I can see if i get any more layer shifting. so far it is nice and quiet and running pretty smooth. only time will tell. |
Definitely take your steppers out of series. Look at my recently posted pictures. If you don’t have plenty of cooling on your step sticks you will start to lose steps. Definitely try out software serial. And if all else fails go into your firmware under the tmc settings and change interpolation to false. This drops it back down to 16th stepping and is what I have been running until I made the recent changes.
…Sent from my iPhone
On Mar 22, 2018, at 3:40 PM, Budavaril <notifications@github.com<mailto:notifications@github.com>> wrote:
Thanks your quick response.
So it was strange. I uploaded previous 1.1.6 version what was working before, but got same problem.
I removed the USB cable, restarted the printer after start printing from SD card normally (first 1.1.6 and after bugfix1.1.x too).
I connected my laptop again, but same problem came back.
Restarted printer and laptop, after working everithing.
So was something problem with the laptop USB but firmware upload was working.
I don't understand, but most important this part have fixed.
I still loos steps, I don't know what I should changing first.. The cables to same lenght (no chain) or change pins and activate SW serial? Finally I want to do both, but I'm courious which one cose in my configuration at the moment.
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub<#28 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AB7M2vHkzvxKlkzkzEbe3CnCzvMY0zguks5thAwygaJpZM4SvCS4>.
|
I had my first successful print with stealthchop and interpolated stepping. Still some fine tuning to do on my slicer side of things but I printed for 5 hours straight at 60mm/s with 30% fast honeycomb infill on a part that is all curved outer perimeters with a separate thin but tall tower and not a single layer shift can be seen. So I can definitely say that cooling is the first step to a successful install. I also noticed that my motors are not getting near as hot as they were even when I was working with plain old 16th stepping. |
From your pictures I would say your drivers are definitely shutting down from over overheating try bringing your acceleration down to 100 and velocity under 500. I would recommend the heatsink mod as I got immediate positive results. Another thing you can try is to increase your psu power from 12v to 14v.
…Sent from my iPhone
On Mar 23, 2018, at 1:05 PM, Budavaril <notifications@github.com<mailto:notifications@github.com>> wrote:
[20180323_172638]<https://user-images.githubusercontent.com/19350571/37844545-158c08b4-2ec0-11e8-944e-c023c322e990.jpg>
I still try all adviced upgrade, but not better.
I tried cooling with a 5015 blower fan from chip side, other 2pcs 40mm from "cable side) but result is same.
Another intresting, the lose steps coming on same points,most of the times between two layers.
Difference between the results, this effect sometimes just on X or on Y, sometimes on both axis coming.
Happening always when accelerating in travel (non printing) and motor(s) give "scratchy" noise.
I can do same if "playing" with the motors, I mean controlling manually by laptop (usb) or from any own (display) controller, if moving 10mm steps fast after each other. No matter, same direction or change direction. About every 10-15 movment give that noise and after of course able to go over the end points because lost the correct home position.
Can I adjust "jerk" or "acceleration" in config.h or if TMC2130 used there is another way?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub<#28 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AB7M2hI3Brrk3hrL7ykkDKSGC9oi2WC-ks5thTlLgaJpZM4SvCS4>.
|
Ok, I have 5pcs brand new "old type" driver and those heatshinks doesn't fitted yet, so I can using them instantly on the chips. Let's try.. |
I was having those same overtemp errors with my wires daisy chained. Also you can turn off the monitoring to keep it from shutting down during those messages. But then your only way to know there is a problem is if you have later skipping. The reason why you think Thomas sandlanderer didn’t have any problems is because he never went any further into the issues. Go to his website and you’ll see his forum has become inundated with people just like you and me that are having these issues. That’s how I found to use the servo pins instead of the lcd pins
…Sent from my iPhone
On Mar 23, 2018, at 2:16 PM, Budavaril <notifications@github.com<mailto:notifications@github.com>> wrote:
Ok, I have 5pcs brand new "old type" driver and those heatshinks doesn't fitted yet, so I can using them instantly on the chips. Let's try..
I just don't understand, if overheating happening, why I can't see that in the log?
I guess any of this error message should be show there same than I got this message with official 1.1.8 which is doesn't adjust the current and after 10mm movement sent message "00:00: X driver overtemperature warning! (0mA)"
Anyway my 12V -> 24V step up converter and the timer modul are alredy on the way for that upgrade. May that will be solving? Until of course I can try to adjust my 12V PSU..
I still courious, how working instantly on the "old" Ramps boards (example in Thomas Sanladerer's video) without extra cooling and why we have this problem on MKS GEN L? What's the difference?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub<#28 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AB7M2uAIKyiHi5HFOaBQeOEhhMsAvkh8ks5thUoFgaJpZM4SvCS4>.
|
All of the modifications I have posted have been recommended by the writer of this library.
…Sent from my iPhone
On Mar 23, 2018, at 2:16 PM, Budavaril <notifications@github.com<mailto:notifications@github.com>> wrote:
Ok, I have 5pcs brand new "old type" driver and those heatshinks doesn't fitted yet, so I can using them instantly on the chips. Let's try..
I just don't understand, if overheating happening, why I can't see that in the log?
I guess any of this error message should be show there same than I got this message with official 1.1.8 which is doesn't adjust the current and after 10mm movement sent message "00:00: X driver overtemperature warning! (0mA)"
Anyway my 12V -> 24V step up converter and the timer modul are alredy on the way for that upgrade. May that will be solving? Until of course I can try to adjust my 12V PSU..
I still courious, how working instantly on the "old" Ramps boards (example in Thomas Sanladerer's video) without extra cooling and why we have this problem on MKS GEN L? What's the difference?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub<#28 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AB7M2uAIKyiHi5HFOaBQeOEhhMsAvkh8ks5thUoFgaJpZM4SvCS4>.
|
I haven't got that "overtemp" error message after changed that 1.1.8, not with my chained wire and not now with "parallel" cables. Not on 1.1.6 and not on bugfix 1.1.x. Now cooled 2side by heatshinks. Uppers blowed 2pcs 40mm fans. Unders (on chip) blowed with 5015. At the moment finished my all ideas.. |
To clear up a few things There is no communication from the motherboard to the driver after the initial boot time, unless you have "Scratchy" noises from the motor indicate a mechanical problem. Check your motion system. An acceleration of 100 would be excessively low. Max velocity of 500 would be excessively high. Please post your M503 and M122 output if you're seeing skipped steps. |
It shouldn’t matter how the wires are tied in, but it does. It’s very hard to get 2 wires into one of those crimpable style connectors. And it was the first problem I encountered and it was almost the first google entry that led me to realize that one bad crimp can damage the whole harness and when I did wire up in series I got the driver overtemp on x no matter what driver I plugged in. The first solution I found was to try parallel using some factory duponts. All I did to remedy this issue was to twist all of them together and heatshrink after that it was on! I hate the cheap crimp style connectors. And it’s the first place to troubleshoot imho. I used some single ended DuPont ribbon cables I had lying around and removed them from the single connectors and applied them to whatever connector I needed. It’s quick and dirty and I plan to route out a simple pcb to make everything cleaner. But my layer skipping is now a thing of the past. Definitely turn off the monitoring if you have a stable setup.
…Sent from my iPhone
On Mar 23, 2018, at 5:08 PM, teemuatlut <notifications@github.com<mailto:notifications@github.com>> wrote:
To clear up a few things
TMC2208 uses Serial
TMC2130 uses SPI
There is no communication from the motherboard to the driver after the initial boot time, unless you have MONITOR_DRIVER_STATUS enabled.
It doesn't matter if you use SW SPI or HW SPI. The only difference is pins used and HW SPI is faster.
It shouldn't matter if the MOSI/MISO/SCK wires are from one driver to another or all starting from the same point. Electrically this is the same.
"Scratchy" noises from the motor indicate a mechanical problem.
An acceleration of 100 would be excessively low. Max velocity of 500 would be excessively high.
Please post your M503 and M122 output if you're seeing skipped steps.
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub<#28 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AB7M2ujSQitB4VLgSN-oD9oaItmq_AbMks5thXJegaJpZM4SvCS4>.
|
Thanks all of that, now I understanding better how works. and M122: |
See M201. |
Wow.. No output.. Possible?? |
In your M503 output you just posted. |
That was the original and last time just set everything back to original. |
M503 shows your current settings. |
Oh yeah... I think that was the key.. |
Just send |
Ok. that's cool. Now became "younger" and to be honest movements more similar how was with the old drivers.. Printing nearly finish and no any skipped steps until now. :) |
Sorry, I realize I had acceleration and velocity settings backwards earlier when I made the recommendation. Lots of settings and numbers to keep up with. This mornings tests are going well. I have doubled all my values for speed and acceleration and so far (fingers crossed) no layer skipping. Hopefully this trend will remain and I can get back to decent printing speeds. My first test at 60mm/s with slow settings was successful but outer perimeters were kinda not where I wanted them, but at 40mm/s we’re almost perfect. |
so it seems that acceleration is kind of flexible, as it doesnt seem to make as much of an impact as velocity does. im stable at or below 150-175. but i have tried accelerations of up to 1000 I havent gotten around to testing any higher values yet. my first test was with x and y at 250 with acceleration of 1000. I was also using a high jerk setting of 30. i had failures at the first sign of tighter shorter movements. so when my base feature was finished and only wall type features were being printed almost every other layer was problematic. stopped and went back to original velocities but increased acceleration had no problems. so ive begun to tune velocity only. |
I received the parts for 24V (12V/24Vstep-up converter and timer modul), but I'm not sure about the delayed method. First powered the MKS GEN L 12V and after the drivers "VM" 24V or first the 24V after the board 12V? |
If you can decide, apply the 24V first. |
Hello. Somehow i think you might be my savior, so i apologize in advance. BIG thanks to you or anybody willing to give me a hand here. |
You likely won't find the help you're looking for at this page. |
Thank you. i wrote here because of the thread topic. Now i am running MKS Gen-L + TMC2130 +SDcard. Installing the sd card module was not that hard actually :) |
here is a rudimentary drawing of how im using my drivers on an MKS-Gen L V1.0 board. Just in case anyone else needs a refernce
The text was updated successfully, but these errors were encountered: