This is my very heavily modified machine, which started as a Tronxy X5SA Pro.

The head carriage is totally replaced with a wider version based around a machined 8mm aluminium plate that has space for an E3D toolchange system between the slideway bearing mounts.

That has a printed "cap" which has the anchor clamps for the belts built in to it, plus channels for the various cables, mount points for the X limit and a BLTouch probe, with slots to take captive nuts for the various retaining screws.

(The cable end anchors etc. and cable support strips are the E3D ones.)

The Z drives are independent and all endstop switches are now optical.

The toolholders are mounted on bases fitted to the rear (originally front) extrusion, and to make space above that the drive belt idler pulleys are relocated back by 34mm using short sections of extrusion fitted on spacers, so the original idler mount locating screw position is now in line with the cross extrusion mounting screw.

The Duet 6HC board is in an aluminium enclosure in the base, at present just resting in place. There is clearance to add a couple of thin rails across the frame to mount it on, and the PSU will probably go on a side extrusion.

There is a tool distribution board in a plastic box on the rear of the frame, and toolboard cables plug in to that - or a dummy plug can be fitted to bypass the CAN chain if one is removed.

The tools are, right to left:
V6 bowden with hardened nozzle and hemera extruder, to allow for carbon filled materials.
V6 bowden using the original tronxy titan extruder for general use.
Hemera direct, to allow for flexible materials
Plus a part build E3D ASMBL.

I'm still spending more time doing fine adjustments than printing so far; the original bed had quite a significant sag in the middle, between its support rails, so I have added a glass bed over that which has given a great improvement, but still a slight cornerways tilt I'm trying to adjust out.

All the modifications are plug-in or bolt-on, the machine can be completely restored to its original specification. It has lost a bit more printing area than I originally expected, but still a lot bigger than my Overlord Pro.

The entire setup could be moved to or used on one of the larger TronXY machines, eg. a 500 or possibly a 600, if the head slide uses the same rail section.

If anyone wants more details of the head mods or the files for those, or any other info about it, let me know.

The convention for lathes and vertical borers is that the spindle axis is Z and the radial axis is X. There is no Y axis.

With two tools, the second tool movement parallel to Z is W and parallel to X is U.

If you use the spindle / chuck for rotary positioning, then the angular setting is normally C axis

If you want to keep compatibility with commercial machines, I'd suggest staying with those axis names.

@alastair1
Make sure you use "run as administrator" when starting Balena Etcher on windows.

It may fail otherwise.

I did my first print using your slicer version yesterday afternoon/evening.

The time estimate was spot on.

And, the print quality is vastly better than the old Cura slicer I use with my Overlord Pro and initially used for the Duet machine!
The item sides have no noticeable layer lines or ripples, just a really good result.

Thanks!

From the paneldue info:

Firmware configuration
RepRapFirmware 3.x and later: Using a PanelDue with Duet 3 and Duet 2 and RRF3 will require M575 P1 S1 B57600 being present in config.g.

RepRapFirmware 2.x and earlier: In RRF2 the command above was assumed as the default, so no additional configuration is required.

I'm guessing you have updated the Duet from 2.x to 3.x, so it needs the extra line in config.g to enable the Paneldue communications:

M575 P1 S1 B57600

The T-1 is presumably supposed to be what causes the system to return any existing tool before picking up the new one??

However you are spot on correct, the toolchange system works perfectly if that line is removed / commented out!

It was in the E3D toolchanger template files I found, and it worked fine with those files, just changed tool positions, when I first set up the Duet on V3.3, so I have no idea at all why it messes things up now? That was the really confusing thing.

Oh well, I can start recalibrating everything now and hopefully get it printing.

Thanks!

Can you see the control board? There should be some LEDs illuminating on it near one edge of the board, if power is on.

The manual for the machine is here, starting at the commissioning section:
https://reprapltd.com/documentation/fisher-build-instructions/commissioning/

And the Duet 2 Maestro control board here:
https://reprapltd.com/documentation/fisher-build-instructions/commissioning/

Does it have an SD card in the card slot? I believe it needs the correct one in for the board to start up properly, but I've no experience with Duet 2 boards.

Is the red light on the toolboard flashing in time with the one on the main board? If so, it may be working OK.

Try sending M122 B121? The board address will have been reset to its default by the two button reset.

If it's not working I'd double-check all the wiring.

Do you have the termination jumper installed on the tool board next to CAN_OUT, and the CAN cable to CAN_IN ?

If it's a generic RJ11 cable, check it is wired identically at both ends - some reverse the wire sequence end to end.

Is it using the same power supply as the main board? If not, the grounds (PSU -) must be linked between the different power sources.

I found a photo of an Artillery end stop switch that has a manufacturers part number visible, a GL-8H

According to the datasheet for those, they are intended for 12 to 24V operation, with NPN (switch to 0V) outputs.

The switches probably need the power connection (brown or orange) changing from the on board endstop connectors to an external supply.

Refs.
https://www.3djake.com/artillery/x-axis-limit-switch
https://www.aliexpress.com/item/4000139818315.html
https://www.makerstore.com.au/product/elec-gl8h/

You appear to be using the default PWM frequency, but with an SSR?

That is unlikely to work well, as the PWM will interact with the AC power frequency, which is all the SSR can switch at.

That could be causing the "stutter" as well.

Try setting the PWM frequency to eg. 5Hz or enabling bang-bang mode instead?

"The Pi Hut" has Pi 4 B 4GB and 8GB in stock - I've just got one for my next printer build, a Kossel XL+ style.

https://thepihut.com/products/raspberry-pi-4-model-b

(They are an official reseller, according the the Raspberry Pi site).

@hayseed_byte said in Which Crimpers for Duet 2:

and adding a dab of solder to a crimped connection
Your wording, not mine!

I never use small crimp connectors in anything we produce, and for a repair, if no equivalent solder connector is available, then properly soldering a wire to a crimp insert is infinitely better than leaving it as a crimp alone, when the correct manufacturers tooling is not available.

There is no quality or reliability problem with soldered joints as long as they are not under mechanical stress.

Small crimp connectors exist because they are faster and cheaper in mass production than soldering.

Soldering is technically better, for small wires:-
https://www.osti.gov/servlets/purl/1288669

If you want to only crimp, then the only way to be 100% sure of it being reliable is to buy something like this for each different make and series of connector - which is ludicrously impractical for servicing multiple different makes of equipment.

For Molex alone, there are around FIFTY different specific crimp tools for different connector and pin/insert types.

https://uk.rs-online.com/web/p/crimp-tools/6634246

As I said before, how you do them is your choice, it's only information, no compulsion.
I'm done.

I did my first print using your slicer version yesterday afternoon/evening.

The time estimate was spot on.

And, the print quality is vastly better than the old Cura slicer I use with my Overlord Pro and initially used for the Duet machine!
The item sides have no noticeable layer lines or ripples, just a really good result.

Thanks!

@hayseed_byte said in Which Crimpers for Duet 2:

I'm going to have to go ahead and disagree with that completely.

That is of course your right.

I'm just working from near 50 years of designing and repairing industrial control systems, and seeing a significant number of poor connections and long-term failures with that general type of small crimp.

The fact is that if you can ever pull a wire out after crimping without it ripping strands, it's not truly crimped and can fail eventually.

https://smcontact.eu/good-crimp-connection/

Anyway, each to their own it's only information!

@stubbedk said in Print 12 mm over bed:

M558 P5 C"^io7.in" H20 F120 T6000 ; set Z probe type to switch and the dive height + speeds
G31 P500 X0 Y0 Z19.027 ; set Z probe trigger value, offset and trigger height

If you are using the Z probe to define the zero height, I believe you need to adjust its trigger height up somewhat to bring the zero down?

@whosrdaddy said in Strange layer offset on coreXY:

This is a mechanical problem, not an electronics problem.

The "cube" appears to be slanted, not just deformed scale-wise - or is it just the photo angle?

Any mechanical setup or steps per mm error would give x-y distortion, but not drift, surely?
A given X-Y point should always be the same, no matter how far off true position it is??

If there is no slant or layer drift, and only shape distortion, then scaling and mechanics are the likely cause.

I use a tool I got from Amazon along with some assorted connectors.
It's a decent fit on many types, and works reasonably well.

https://www.amazon.co.uk/gp/product/B07QNPZDTW/

However; for small contacts such as these, I do not consider any tool other than the connector manufacturers own exact type to be a valid "crimp" - ie. guaranteeing sufficient force in the correct areas to slightly extrude the metal and form a gas-tight connection.

I always solder them after fitting to the wire, by heating the back of the contact until a small amount of solder will flow in to the wire connection area.

ps. Any non-compound action tool, such as those PA-09 etc., cannot possibly give enough force to form a valid crimp, and you must also solder the wire if you want long term reliability.

@stephen6309 said in Strange layer offset on coreXY:

I use M906 X800 Y800 Z800 E800 I30 for the stepper currents on my CoreXY. I believe the 300 your using is not enough.

He says early on that he is using external stepper drivers.

Back to the question - another thing to try; reduce the maximum speed from 18000 down to eg. 10000 to be sure the drivers are not missing steps?

Hi,
just for info, on one topic I was getting alert popups every time I scrolled near the bottom of the page.
[I know.. "Don't scroll near the bottom of the page!" ]

Can you see the control board? There should be some LEDs illuminating on it near one edge of the board, if power is on.

The manual for the machine is here, starting at the commissioning section:
https://reprapltd.com/documentation/fisher-build-instructions/commissioning/

And the Duet 2 Maestro control board here:
https://reprapltd.com/documentation/fisher-build-instructions/commissioning/

Does it have an SD card in the card slot? I believe it needs the correct one in for the board to start up properly, but I've no experience with Duet 2 boards.