Cast Toolplate Bed Progress ?

[Dizzwold]

@deckingman @droftarts @fcwilt

Merry Christmas Guys.

@deckingman

Thank you for the link to your video explain the tram scenario. This may play some part in this and I'll have to investigate further.
What isn't shown in the image above is the point measured at the back Left. The point at the back Centre is Z-0.309 where at the back Left was -0.264. I'm not ruling out a tram issue, but if it was, the back Left would have been a greater negative number than the back Centre?
I do actually have some old imperial dial clock somewhere in the garage that belonged to my father, but I've never made a bracket for them.
I might try and use my previous Bed Levelling macro (which took measurements from the four corners). The frame is still basically the original Tronxy X5SA Pro. The original bed had 6 adjusting screws, then mounted on 2 leadscrews. So Maybe I can use this instead of a dial using the BL Touch.
As my funds are limited I'm also still using the original Tronxy Rails (U grove bearing/rails), which would mean if it is a tram issue I'd need to shim the underside of the rails as these aren't blind joints and screwed down onto the end of a frame corner upright.

Dizzwold.

[droftarts]

@Dizzwold You appear to have a kinematic bed, are the three steppers driven from separate drivers? If so, have you set up your bed.g? See https://docs.duet3d.com/en/User_manual/Connecting_hardware/Z_probe_auto_levelling

I also think you need a much finer bed mesh; you're only checking at 9 points.

Ian

[deckingman]

@Dizzwold Will your funds stretch to buying a decent straight edge? Search for "Plasterers Feather Edge". With a strong light behind it, you'll be able to see if the plate is flat or not. But I'd be willing to bet that it's a tramming issue. Ultimately you need to be sure that a plane described to the 3 lead screw attachment points is parallel with a plane described by the XY gantry with errors less than 0.1mm (I got mine to within 0.03mm).

As I said at the end of that video, you can either spend the time and do that, or you can use mesh compensation. But if you do the latter, there is no point in spending the money on expensive tooling plate which is guaranteed flat. (and if you use a continuous belt driving all 3 screws, you'll only have to level the bed once and never again).

[Dizzwold]

@droftarts @deckingman

Hi Ian, and Ian,

There are 3 Z axis motors and all on separate drivers and correct order from M584 drive mapping (Z1 Centre Left, Z2 Front Right, Z3 Rear Right)

So far my testing has been with everything running cold (cold nozzle and cold bed). I started with a 12 x12 bed mesh which initially highlighted the issue, so went to a 3 x 3 for quick tinkering and test.

I'm wondering if it's the bed not completely parallel to the bed frame?

@deckingman said in Cast Toolplate Bed Progress ?:

s I said at the end of that video, you can either spend the time and do that, or you can use mesh compensation. But if you do the latter, there is no point in spending the money on expensive tooling plate which is guaranteed flat

I agree. After using the bed that came with the printer, a bed fastened to a frame with 6 (3 front and 3 rear) adjusting screws, pulling it all over the place, then mounted to 2 leadscrews. This was the idea to invest in a decent bed where once I know it's set correctly, I pretty-much forget about it. More so the stress and strains of the old bed.
Here's the old bed for reference

[Dizzwold]

@deckingman @droftarts

Hi Ian and Ian,

I've ran the 3 x 3 mesh and took the following images;

I've then ran my old Bed adjustment screws Macro which basically take a measurement from each corner;
Front Left, Front Right, Rear Right and finally Rear Left;

G28
M671  X5:278:278:5 Y5:5:302:302 P0.5
G30  P0 X-29 Y1 Z-9999				; probe front left
G30  P1 X279 Y1 Z-9999			; probe front right
G30  P2 X279 Y324 Z-9999			; probe rear right
G30  P3 X-29 Y324 Z-9999 S4  		 ;probe rear left
G1 	 Y0 F12000
G1   X0 Y0 F12000

I then got the following measurements in Console, which would suggest that the bed isn't fully parallel to the bed frame more so in the Rear Left.

I can only assume this is due to a slightly shorter screw or a slightly deeper mounting hole in the bed.

Dizzwold.

[Dizzwold]

@deckingman @droftarts

With the washers and shims I currently have (1mm & 0.5mm), I think this is the best I'm going to get it.

Just ordered some 0.1 and 0.2mm shims, so can hopefully improve this in the new year 'while saving for a new frame'.

Dizzwold.

[sebkritikel]

@Dizzwold definitely carry on with getting your frame and gantry trammel up, but be aware of the published specs (if any) for your cast plate.

As an example, this material and vendor is popular in the states:
https://www.midweststeelsupply.com/store/ATP5Flyer.pdf

Note that for material less than 12.7mm thick, the flatness tolerance is .381mm or better across the entire plate. Example image of what ‘flatness’ means below.

https://media.faro.com/-/media/Project/FARO/FARO/FARO/Images/Resources/2021/01/15/23/03/flatness-ASMEex-969x1030.png?h=692&w=650&rev=97ef025583f14db4a5c22b058fb7b809&hash=F3D4FD2D475501E30BEB379E0BBDA981

Specs vary by manufacturer, and you may well never see the worse case flatness tolerance, but just a tidbit to keep in mind.

[deckingman]

@sebkritikel said in Cast Toolplate Bed Progress ?:

.................... Note that for material less than 12.7mm thick, the flatness tolerance is .381mm or better across the entire plate.................

That's not exactly correct. The specs refer to "guaranteed flatness" not "flatness tolerance". It's a common safety margin that suppliers use to cover themselves in case of litigation. The quoted flatness also applies to the entire sheet which could have a length of 3670mm and a width of 1842mm. That's an area of 6.76 m^2 whereas if the OPs bed was 500mm x 500mm that would only be 0.25 m^2 so it's highly unlikely that his small section of a (machined) sheet would have a deviation anything like that.

[Dizzwold]

@deckingman @droftarts @fcwilt @sebkritikel

Hi Guys,

I'm back with an update after much tinkering, finessing and frustration.

The following image show's a cold 9 point bed compensation. Again I've only used 9 points for quickness to test the flatness of the bed and to try and eliminate any tramming.

The point at the rear centre is now the only point that is out. The rest vary from +0.004 to -0.090.

Going back to the Emergency stop button I have (the mushroom type on a NC circuit).
Is it possible so when the switch is reset, it will reset the Emergency Stop in DWC?
I've tried using M999 in a variety of ways, in config.g as trigger 0 and in sys/trigger2.g and just can't figure this out.
Config.g

; Custom settings are not defined
 ; Emergency Stop
M950 J2 C"!io3.in"
M581 P2 T2 S0 R0
;M582 T0 S0

 ; Z Axis Limit Switch
M950 J3 C"io2.in"
M581 P3 T3 S0 R0
;M582 T2 S0 

Trigger 2.g

;Trigger2.g Emergency Stop Button

M112				;stop all
M582 T2 S0
M999				;restart all

Trigger 3.g

;Trigger3.g Z Axis Limt Switch
M112				;stop all
M582 T3 S0
M999				;restart all

The Z Axis limit switch is a fail-safe for while I'm tinkering as I've already accidentally touched the paneldue after a M564 S0 and crashed the leadscrew nuts into the motor mounts.

Dizzwold.

[deckingman]

@Dizzwold said in Cast Toolplate Bed Progress ?:

Is it possible so when the switch is reset, it will reset the Emergency Stop in DWC?
I've tried using M999 in a variety of ways, in config.g as trigger 0 and in sys/trigger2.g and just can't figure this out.
Config.g

I'm not sure that you can use M999 to reset after an emergency stop. The docs say it'll restart the firmware after a software reset which isn't the same thing. The usual way to reset an emergency stop is to power cycle the board. I have mine wired such that it kills all power to the boards so resetting the emergency stop button effectively does just that. You could maybe do M98 P "config.g" which will run config.g again - that might work as a reset mechanism.

The way to do it would be to have two triggers, one that works on an active to inactive edge, and the other that works on inactive to active. So you can use M581 P2 T0 S0 R0 which will do an emergency stop as if M112 was run - you don't need create a trigger0 macro to do an emergency stop. Then use M581 P2 T2 S1 R0 to run the macro "trigger2" which would contain M98 P "config.g". You might need to swap S0 and S1 around depending on how the switch is wired.

[Dizzwold]

@deckingman

Hi Ian,

@deckingman said in Cast Toolplate Bed Progress ?:

You could maybe do M98 P "config.g"

Now there's an idea, thank you for that.

[Dizzwold]

@deckingman

Hi Ian,

A quick question. Should the M582 line be within the trigger.gor within the main config.g?

Dizzwold.

[deckingman]

@Dizzwold said in Cast Toolplate Bed Progress ?:

@deckingman

Hi Ian,

A quick question. Should the M582 line be within the trigger.gor within the main config.g?

Dizzwold.

Not sure I understand. Triggers are activated when the input defined in M581 changes state. So you don't normally have to do anything as the relevant macro will run "automatically" when the input state changes. AFAIK, M582 is just a way of checking the state of the inputs so I've never used it.

[Dizzwold]

@deckingman

Hi Ian,

Sadly the Trigger 2, M98 P"config.g" has no influance on restarting DWC. It still needs manually resetting.
Config.g

; Custom settings are not defined
 ; Emergency Stop
M950 J2 C"io3.in"
M581 P2 T0 S1 R0
;M582 T2 S0

M581 P2 T2 S0 R0

Trigger2.g

;Trigger2.g Reset Emergency Stop Button

M98 P"config.g"

I've tried many variants with M112, M999, M400 and tried changing the S0, S1, S-1 values, although I have it connected on a NC circuit so S1 on a non-inverted pin triggers T0 (M112 Emergency Stop), then releasing the
Emergency Stop Button S0 Triggers the same non-inverted pinT2, M98 P"config.g"
Have got something wrong somewhere?

[deckingman]

@Dizzwold Dunno. I guess we need to first check that the correct macro is being triggered with the switch in each position. Try doing M581 P2 T3 S1 R0 and keep the existing M581 P2 T2 S0 R0. So we have two triggers (2 and 3) and one should run on a rising edge the other on a falling edge. Then make trigger2.g with just a single command M118 S" This is trigger 2", and trigger3.g with just a single command M118 S"This is trigger 3".

Then exercise your switch and see what happens. Hopefully you'll get a message "This is trigger 2" when the switch changes from one state to another and "This is trigger 3" when the state changes back. Once we are sure that part is working, then we can think about what commands to put in each trigger macro.

[Dizzwold]

@deckingman said in Cast Toolplate Bed Progress ?:

This is trigger 2

Hi Ian,

Hmm... I tried what you suggested above, and i worked as expected;
Trigger 2 when the E Stop Button is pressed S1
Trigger 3 when released S0

So I've then replaced Trigger 2 with
M112
M999
and Trigger 3 with
M98 P"config.g"

Same problem DWC on my computer needs a physical click on the RESET button displayed on the screen (paneldue states HALTED), so it would seem that the M98 P"config.g" is the issue.

Config.g;

; Custom settings are not defined
 ; Emergency Stop
M950 J2 C"io3.in"
M581 P2 T2 S1 R0

;M950 J2 C"io3.in"
M581 P2 T3 S0 R0

 ; Z Axis Limit Switch
;M950 J3 C"io2.in"
;M581 P3 T3 S0 R0

Trigger2.g;

;Trigger2.g Reset Emergency Stop Button
M112
M999

;M118 S"This is Trigger 2"

Trigger3.g

;Trigger3.g Z Axis Limt Switch

;M118 S"This is Trigger 3"
M98 P"config.g"

Maybe it's not possible?

Dizzwold

[deckingman]

@Dizzwold said in Cast Toolplate Bed Progress ?:

Maybe it's not possible?

Maybe. I'd have thought there ought to be a way to recover from an emergency stop but the docs just talk about power cycling the board. I'm surprised running config g doesn't work because that is effectively all that happens when power is restored.......but there is possibly a work around.

The obvious one is do as I do and use the emergency stop button to physically interrupt the 24V power to the board(s).

But failing that, maybe we can do something else. So rather than an emergency stop, think about what you actually want to stop. I guess that's motors and heaters. Heaters are easy - just set the temperatures to zero in one macro and restore them in the other. So what about using M18 instead of M112 to disable motors? It might not be immediate though so might have to wait for any moves to finish which I guess is not what you want an emergency stop to do? M0 is another option but again, I don't think it will work until the current move is complete.

[Phaedrux]

@Dizzwold said in Cast Toolplate Bed Progress ?:

Maybe it's not possible?

A question for @chrishamm I think.

[deckingman]

@Phaedrux said in Cast Toolplate Bed Progress ?:

@Dizzwold said in Cast Toolplate Bed Progress ?:

Maybe it's not possible?

A question for @chrishamm I think.

,..,....or @dc42. Essentially we're trying to find a way to reset a printer after M112 has been sent, without cycling power. The OP is using trigger macros so it can be a series of commands rather than a single command. Apparently running config g via M98 doesn't do the job and I'd have thought it would.

[dc42]

@deckingman sending M999 works. The emergency stop button sends M112 followed a little later by M999.