Linear Quadrature Encoder Calibration Error

btmclain

Hello,

We are trying to get a linear encoder and magnetic encoder setup with our 1HCL V1.0a expansion board. We have the magnetic shaft encoder setup and are trying to get the linear encoder configured. We are using a Phidgets ENC4110_0 quadrature encoder.

The process we have been using is as follows:

Boot duet and home axis in open loop configuration.
Switch to closed loop configuration.
Perform calibration tuning move on magnetic encoder.
Perform calibration tuning move on linear encoder. This is where the error arises. The error we receive is "M569.6 P50.0 V1 Error: M569.6: Driver 50.0 basic tuning failed, the measured motion was inconsistent Warning: Driver 50.0 warning: tuning/calibration failed".

Our configuration file along with m122 for attached board is below. Thank you in advance!

code_text
```M569.1 P50.0 T1 S200 R40 I0 D0 C1000; Configure the 1HCL board at CAN address 50 with a quad shaft encoder, 
;M569 P50.0 D4 S1;set drive 0 go forward, address 50, spreadcycle, CLOSED  LOOP

M569 P50.0 S1 D2;set drive 0 go forward, address 50, spreadcycle, OPEN LOOP

M584 X50.0; set X drivers
M917 X0 ; Set the closed loop axes to have a holding current of zero

M350 X16 ; 16 microstepping
M92 X640.00 ; set steps per mm

M566 X100.00  ; set maximum instantaneous speed changes (mm/min)
M203 X3000.00 ; set maximum speeds (mm/min)
M201 X10.00  ; set accelerations (mm/s^2)

M906 X4000 I50 ; set motor currents (mA) and motor idle factor in per cent

M574 X2 S1 P"!50.io0.in"
M915 X S3 F0 R0

The commands to perform calibration are as follows

M569 P50.0 D4 S1 ;closed loop
M569.6 P50.0 V2 ;magnetic encoder calibration
M569.6 P50.0 V1;quadrature encoder calibration

And our m122 response is:

M122 B50
Diagnostics for board 50:
Duet EXP1HCL rev 1.0a or earlier firmware version 3.5.4 (2024-11-24 10:39:28)
Bootloader ID: SAME5x bootloader version 2.4 (2021-12-10)
All averaging filters OK
Never used RAM 45676, free system stack 178 words
Tasks: EncCal(1,nWait 6,0.1%,175) Move(3,nWait 7,0.0%,131) CLSend(3,nWait 6,0.0%,147) HEAT(2,nWait 6,0.0%,105) CanAsync(5,nWait 4,0.0%,66) CanRecv(3,nWait 1,0.0%,31) CanClock(5,nWait 1,0.0%,64) TMC(4,nWait 6,62.3%,321) MAIN(1,running,36.5%,253) IDLE(0,ready,0.1%,29) AIN(2,nWait 2,1.0%,255), total 100.0%
Owned mutexes:
Last reset 00:15:31 ago, cause: software
Last software reset data not available
Driver 0: pos -200674, 640.0 steps/mm, tuning/calibration failed, SG min 26, mspos 1000, reads 44823, writes 44595 timeouts 0, steps req 0 done 13441
Moves scheduled 1, completed 1, in progress 0, hiccups 0, segs 3, step errors 0, maxLate 0 maxPrep 14, maxOverdue 0, maxInc 0, mcErrs 0, gcmErrs 0, ebfmin 0.00 max 0.00
Peak sync jitter -6/8, peak Rx sync delay 185, resyncs 0/0, next timer interrupt due in 4 ticks, enabled, next step interrupt due in 3691560387 ticks, disabled
VIN voltage: min 24.2, current 24.2, max 24.3
V12 voltage: min 12.0, current 12.1, max 12.1
MCU temperature: min 31.9C, current 39.2C, max 39.2C
Last sensors broadcast 0x00000000 found 0 207 ticks ago, 0 ordering errs, loop time 0
CAN messages queued 7691, send timeouts 0, received 4882, lost 0, errs 0, boc 0, free buffers 38, min 38, error reg 0
dup 0, oos 0/0/0/0, bm 0, wbm 0, rxMotionDelay 278, adv 37137/37137
Closed loop driver 0 mode: closed loop, pre-error threshold: 2.00, error threshold: 4.00, encoder type linearComposite, position -32
Shaft: Encoder reverse polarity: no, full rotations 0, last angle 10911, minCorrection=-30.7, maxCorrection=26.1, agc 117, mag 4559, no error
Lin: Encoder reverse polarity: no, raw count 65504
Tuning mode: 0, tuning error: 0x10, collecting data: no
Control loop runtime (us): min=5, max=66, frequency (Hz): min=24, max=17857
Accelerometer: none
I2C bus errors 12, nak

dc42

@btmclain we've seen this once before. The cause was identified as high backlash in the drive chain between the motor and the axis. In the version 3.6.0-rc.1+3 binaries at https://www.dropbox.com/scl/fo/geqmn8gbn97n6b587mkbk/AJ2hBIqO-L57_QpR_uqHJ9c?rlkey=fw37wycbp2gil8rvxhe7aopy7&dl=0 I've increased the amount of backlash allowed when using a linear composite encoder, so that version might work for you. To use it you will need to update your whole system to firmware 3.6. If you are running in SBC mode then update to 3.6.0-rc.1 first, then you can install the 3.6.0-rc.1+3 binary on the 1HCL.

btmclain

@dc42 Thank you for the reply!

We went through the steps that you recommended and tried the binaries but still are facing the same error. We have noticed that there is a jerk at the beginning of the quadrature calibration that may be causing us to skip steps. We do not see this jerky motion during normal operation or during the magnetic encoder calibration. Do you have any idea of what could be causing this? Our driver configurations are all listed above.

ironhydroxide

Have you tried the calibration steps manually?

once you have the magnetic encoder calibrated once you should not have to calibrate it again unless you change the mounting or magnet.

btmclain

@ironhydroxide Is there a way to manually perform the calibration for the linear encoder? It is my understanding that we cannot use the system until both encoders have been calibrated. Thank you for the reply!

ironhydroxide

@btmclain By "manually" i mean not in a homing macro, but by sending the gcode manually either in the DWC console or connected SBC.
But after rereading your first post, I assume that's what you've done.

I think your issue may lie in the initial configuration of the encoder. the C parameter should be the number of cycles from the quadrature encoder per motor revolution.
I don't see anywhere you specify the motion relationship between motor and linear encoder, the linear encoder you specify has a 5um/pulse output, (20um/cycle)
With your C value of 1000 that works out to a movement of 20mm/motor rotation. is that correct?

Are you doing a small motor movement in open loop mode before attempting calibrations?

When playing with my setup I was able to somehow get the system to move in closed loop with only a magnetic encoder calibration, then I was able to confirm the linear encoder cycles/rotation value (i had to iterate a few times as the documentation says Pulses per revolution, I believe it should say cycles/revolution instead as that is what I ended up at to get the correct movement) Once I had the C value in the ballpark I was able to get the system to calibrate the Quadrature encoder (M569.6 V1) successfully.

dc42

@btmclain there should not be a jerk at the star of encoder calibration. Please post a video showing exactly what you mean.

If you configure the system to use just the magnetic encoder, does your system behave correctly, and is motion smooth?

btmclain

@ironhydroxide Thank you for the reply. Our system is currently setup to do 5mm/rev of travel. Yes, we have been doing some small motor movements before beginning the calibration. We tried your suggestion of implementing cycles/rev instead of counts/rev for the C parameter but unfortunately no luck.

btmclain

@dc42 We have recorded a video showing both calibration routines but I am unable to post a link because of my reputation. Do you have another option for getting the video link to you?

The motion is fine and smooth for operation during the magnetic cycle. You can hear a loud jerk at the beginning of the linear encoder calibration cycle, quickly followed by the calibration failing.

Thank you!

droftarts

@btmclain said in Linear Quadrature Encoder Calibration Error:

We have recorded a video showing both calibration routines but I am unable to post a link because of my reputation.

I've upvoted a couple of your posts, so you should be able to post it. It's a forum anti spam measure.

Ian

btmclain

@droftarts Excellent thank you!

Here is the video. It first runs through the magnetic encoder calibration which sounds smooth. You can hear the jerk at the beginning of the linear encoder calibration.

https://photos.app.goo.gl/6C1YT96pYNYoVGq28

ironhydroxide

@btmclain I see you have standstill current at 0% (M917 X0), have you tried commenting out this line, or increasing the standstill current value?

on my system when I add the M917 X0 I also get this clunk. Removing it and I have no clunk when tuning.