New firmware 1.21RC5 available

[deckingman]

@DougJones:

I've upgraded to 1.21 RC5 firmware and 1.21 RC4 webcontrol

When I press the home all button, i get the following error

G28
Error: G0/G1: Insufficient axes homed

I've never gotten this before and the setup as not changed from v1.20. Any idea what's causing this?

Most likely this.

From the release note for RC 3 we have:
"New features:

On Cartesian and CoreXY printers, normal movement commands are no longer permitted until the corresponding axes have been homed"

Then we have:
"On Cartesian and CoreXY printers, normal G0 and G1 moves are no longer allowed before the corresponding axes have been homed. In particular, if your homex.g, homey.g and homeall.g files raise Z a little at the start and lower it at the end, you will need to add the S2 parameter to those G1 Z moves. Otherwise the G1 Z move will be refused unless Z has already been homed and the homing macro will be terminated."

But in the release note for RC4 we have
"Upgrade notes:

As for 1.21RC3.

New features and changed behaviour:

On CoreXZ machines we no longer require Z to be homed before bed probing with G30"

So following that trail of breadcrumbs I'd assume that for Cartesian printers, if there is a Z move at the start of homeall, you need to add S2 to that move.

But we also have:

"Added M564 H0 command to allow axis movement before homing on Cartesian/CoreXY printers"

So I'd guess that adding that M564 H0 to your config.g will restore the printer back to it's previous behaviour.

[DougJones]

Gotcha. It's homing now just fine.

Did you see my question above with the M569 T parameters and the timing chart?

I am using M569 P5 S1 R1 T10:10:7:2. But both channels seem to be the same still. This should give the dir signal being on 7 usec before and staying on 2 usec after the step signal turns off.

And for some reason my images are not showing up correctly.

[dc42]

@DougJones:

I have a question about the M569 T parameters. I have external drivers that I am setting back up after burning out the logic converters my breakout board.

is the image correct? I assume bb is a minimum step pulse interval?

https://ibb.co/fPj9Pc

Yes, your image is correct. The values are the minimum intervals that will be used. In practice they will be rounded up a little; you can send e.g. M569 P5 to see how much.

[DougJones]

Ok. here is the image that I am getting on the scope. yellow is step, blue is direction.

The direction should lead the step by 7 usec and trail the step by 2 usec. The image indicates that they are the roughly the same.

[dc42]

I don't think you can be looking at the step and direction signals, because the timings are too close. RRF never changes step and dir at the same time, even if you haven't specified any timing requirements. The yellow trace in particular looks very odd and its amplitude is lower than for the blue trace; so I think the yellow trace is probably crosstalk from the blue one.

[Drammy]

(continued from the rc4 thread, in case relevant here)

Ok, so if I downgrade back to 1.2.0 everything works as expected and I can see the Endstops hit = No for all axes
Then I upgrade to 1.2.1 (rc4 or 5) and all axes show Endstops hit = yes.

Two wire switches connected to left two pins on duet board (with wifi component bottom left)

Should this be happening? Have I done something wrong?

[robm]

Sorry to be boring, upgraded rc4 to rc5 combined firmware bin and wifi server, still have the rc4 webserver code on cartesian printer. Still works as expected, just like before the upgrade :-))
Thank you for the g30 multi-touch change

[dc42]

@Drammy:

(continued from the rc4 thread, in case relevant here)

Ok, so if I downgrade back to 1.2.0 everything works as expected and I can see the Endstops hit = No for all axes
Then I upgrade to 1.2.1 (rc4 or 5) and all axes show Endstops hit = yes.

Two wire switches connected to left two pins on duet board (with wifi component bottom left)

Should this be happening? Have I done something wrong?

The DWC machine properties page shows the raw endstop input state wirth used wirth firmware 1.21RC5. So "Endstop hit" really means the endstop input is high, otherwise it is low. M119 reports the actual endstop status.

The incorrect readout by DWC doesn't affect the behaviour.

[Drammy]

@dc42:

@Drammy:

(continued from the rc4 thread, in case relevant here)

Ok, so if I downgrade back to 1.2.0 everything works as expected and I can see the Endstops hit = No for all axes
Then I upgrade to 1.2.1 (rc4 or 5) and all axes show Endstops hit = yes.

Two wire switches connected to left two pins on duet board (with wifi component bottom left)

Should this be happening? Have I done something wrong?

The DWC machine properties page shows the raw endstop input state wirth used wirth firmware 1.21RC5. So "Endstop hit" really means the endstop input is high, otherwise it is low. M119 reports the actual endstop status.

The incorrect readout by DWC doesn't affect the behaviour.

Thanks for the tip with M119. I've checked that and its working fine.

I still have the problem though. When I downgrade to 1.2.0 Z homes correctly, when I upgrade to 1.2.1 (rc4 and 5) Z crashes into the bed and ignores the Z switch. Any ideas?

Sorry, a bit new to all this so rather helpless…

[dc42]

You posted your config.g file in the RC4 thread, but I haven't seen your homing files yet. Please post them.

[Drammy]

Here's the homing files. They were on the rc4 thread as well so apologies if I'm double posting

[[language]]
; homex.g
; called to home the X axis
;
; generated by RepRapFirmware Configuration Tool on Wed Mar 07 2018 10:07:16 GMT+0000 (GMT)
G91               ; relative positioning
G1 S2 Z5 F6000    ; lift Z relative to current position
G1 S1 X-205 F6000 ; move quickly to X axis endstop and stop there (first pass)
G1 X5 F6000       ; go back a few mm
G1 S1 X-205 F600  ; move slowly to X axis endstop once more (second pass)
G1 S2 Z-5 F6000   ; lower Z again
G90               ; absolute positioning
 

[[language]]
; homey.g
; called to home the Y axis
;
; generated by RepRapFirmware Configuration Tool on Wed Mar 07 2018 10:07:16 GMT+0000 (GMT)
G91               ; relative positioning
G1 S2 Z5 F6000    ; lift Z relative to current position
G1 S1 Y-205 F6000 ; move quickly to Y axis endstop and stop there (first pass)
G1 Y5 F6000       ; go back a few mm
G1 S1 Y-205 F600  ; move slowly to Y axis endstop once more (second pass)
G1 S2 Z-5 F6000   ; lower Z again
G90               ; absolute positioning
 

[[language]]
; homez.g
; called to home the Z axis
;
; generated by RepRapFirmware Configuration Tool on Wed Mar 07 2018 10:07:16 GMT+0000 (GMT)
G91               ; relative positioning
G1 Z5 F6000       ; lift Z relative to current position
G1 S1 Z-180 F6000 ; move Z down until the switch triggers
G92 Z0            ; set Z position to trigger height
 
; Uncomment the following lines to lift Z after probing
;G91              ; relative positioning
;G1 Z5 F100       ; lift Z relative to current position
;G90              ; absolute positioning
 

[[language]]
; homeall.g
; called to home all axes
;
; generated by RepRapFirmware Configuration Tool on Wed Mar 07 2018 10:07:16 GMT+0000 (GMT)
G91                     ; relative positioning
G1 S2 Z5 F6000          ; lift Z relative to current position
G1 S1 X-205 Y-205 F6000 ; move quickly to X and Y axis endstops and stop there (first pass)
G1 X5 Y5 F6000          ; go back a few mm
G1 S1 X-205 Y-205 F600  ; move slowly to X and Y axis endstops once more (second pass)
G1 S2 Z-180 F6000 S1    ; move Z down stopping at the endstop
G90                     ; absolute positioning
G92 Z0                  ; set new Z position
;G1 Z5 F100             ; uncomment this line to lift the nozzle after homing
 

[dc42]

The only thing I've spotted is that in homez.g, in this line:

G1 Z5 F6000 ; lift Z relative to current position

you need to add parameter S2.

Please run M574 without parameters, to check that the Z endstop configuration is reported correctly.

I suggest you test Z homing by starting with the nozzle a long way above the bed, telling the machine to home Z, then triggering the Z endstop switch manually. If that doesn't stop the Z movement, that will give you time to turn the power off before the nozzle reaches the bed.

[Drammy]

Thanks for the tip of moving Z up first and triggering manually!

So HomeZ now works fine but HomeAll doesn't. I still get the same problem. Is this line correct (from homaall.g)..

[[language]]
G1 S2 Z-180 F6000 S1    ; move Z down stopping at the endstop
 

[dc42]

No, it has both S1 and S2 in it. Remove the S2.

[deckingman]

@dc42:

No, it has both S1 and S2 in it. Remove the S2.

David,

Could you explain what S2 actually does on a Cartesian or CoreXY?

The Wiki is a little vague

We have:

"Snnn Flag to check if an endstop was hit (S1 to check, S0 to ignore, S2+S3 see note, default is S0) 1"

Then referring to that note we have:

"1RepRapFirmware can be set to enable or disable the "sensing" of endstops during a move. Using the S1 or S2 parameter on a delta printer causes the XYZ parameters to refer to the individual tower motor positions instead of the head position, and to enable endstop detection as well if the parameter is S1. If S3 is passed instead, RepRapFirmware will measure the axis length."

So there is an explanation of what S2 and S3 do on a Delta but no explanation of what S2 (or S3) do when used with other kinematics. I guess it just needs a sentence in the Wki to say something like "For other Kinematics S2 does "xyz" "

[dc42]

S2 causes a raw motor move, ignoring any mapping from axis movement to motors, and also disables axis limits, bed compensation and axis skew compensation.

So on a Cartesian printer, S2 is just like S1 except that endstop switches are not checked. On a CoreXY printer, G1 S2 Xnn will cause just the X motor to move, resulting in a diagonal movement. Similarly for G1 S2 Ynn. But the Z motor on a CoreXY printer drives just the Z axis, so G1 S2 Znn will cause just a Z movement

[deckingman]

Thanks David,

Might be an idea to cut and paste that explanation into the Wiki, as and when you get time.

[Drammy]

@dc42:

No, it has both S1 and S2 in it. Remove the S2.

This is working now; thanks for your help

[Danal]

I did a super quick check of the various CNC Work/Machine coordinate changes, and all looks good. I will do a more thorough check as time permits.

However… CNCjs, in "Marlin" mode, doesn't understand how to decode the status report, and always shows machine coordinates in both the Machine and Work readouts. Not sure what the best answer is here...

[DougJones]

Regarding the external drivers. I am still getting random layer shifting in both x and y. I have upgraded to 1.21 RC5 and no change. I am using M569 P5 S1 R1 T10:10:7:2 to create the step and direction signals.

I decided to remove the external drives to see if it was the mechanics of the system and not drive related. I installed traditional steppers on the X and Y axis using the built in drivers and re-ran the 3DBenchy. The results were pretty clear.

While I have a couple other minor issues with cooling to resolve, the print came out much better with no layer shifting. This has been the main problem I've been fighting for weeks. The mechanical side of the printer looks good, I just have something with these external drives that I cannot figure out.

They are 2HSS57 closed loop stepper drives and either something with the step generation or internal position/velocity monitoring loop in the drive is giving a problem. Nothing I did ever seemed to effect any change (inverting signals, changing from rising to falling edge, etc). The only thing that had a small impact was increasing the step signals from 1600 p/rev to 3200 on the drive itself and doubled the steps/mm. This resulted in reduced layer shifting, but it was still there.

Short of replacing all the external closed loop motors/drives with steppers and using the built in drives, I'm not sure really where to go from here.