Curious Delta Homing from PanelDue

Hello - I am trying to set up my Delta using my new Duet (Firmware 1.12a).

All seems to work fine (homing and then moving the axis using the PanelDue). But after the first click on the home, any subsequent requests to home only move two axis (the Y Engine doesn't move while X and Z move towards the end-stop).

Any Idea where to start looking?
Thanks!

I suspect you have a bad connection between the Y homing switch and the Duet, so that sometimes the endstop appears to be permanently activated. Use either M119 or the Machine Properties page of the web interface to read the endstop switch state.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Thank you for your answer - I've now looked intensively in the issue and think I can give a bit more clues, but I do not have an answer here:

the Opto Endstops I use are what seems to be a standard from SainSmart. They have a LED control which lights up when connected and turns down when the switch is closed.

I've tested four different switches, and interchanged all cables: it seems that some switches are more sensitive than others as they seem to operate fine (control LED turning on and off as designed when operating) but the Duet board does not register the switch and keeps registering "closed".

Since it is happening with some switches but not others, could it be that the trigger voltage is just borderline due to resistances on the SainSmart board? Is there anything I can do to adjust that on the Duet?

I think you are right. The Duet has LEDs on the endstop inputs too, so the endstop switches need to be able to sink about 2.1mA of current. This is higher than a RAMPS, but less than a Smoothieboard. I suspect that your opto endstops are not capable of sinking this current.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

So there is no way of using OptoEndstops with the Duet?

I've now been through three different suppliers of Endstops - and it is surprising that it works with some (the two original ones on the machine) but not one, however much I try?

Edited 2 time(s). Last edit at 06/25/2016 05:08PM by Risingphoenix.

You can set the endstops on the Duet to be active high or active low using the M574 command in config.g. If they are set to be active high then the LEDS will light up when the endstops are not triggered. This is the common case using normally closed microswitches. But if you set them up to be active low then the LEDS on the Duet will instead illuminate when the endstop switches are triggered.

Edited 1 time(s). Last edit at 06/25/2016 07:00PM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Sorry, I may I edited my previous reply while you already responded...

My issue is that the Duet firmware does not pick up the state change: the LED's (on the Duet and on the Endstop) change with the state - and on one Endstop it simply does not register any change of state in the firmware despite the LEDs on the board changing.

I have tried multiple Endstops, and since the LEDs work according to expectation I am wondering why the firmware does not pick up any change? Is there anything I can try on the board itself as the cables and endstops have been proven to be error free (through switching components several times)?

I also see now that while powering up the board (and after Emergency reset) the Endstops state are all correctly identified. When triggered, it registeres in the firmware. But once released again, the status does not return back in the firmware.

Edited 1 time(s). Last edit at 06/25/2016 05:16PM by Risingphoenix.

I suspect your endstops are intended for use with 5V electronics. When you run them from the 3.3V endstop supply from the Duet, they will probably have too low a current sink capacity.

You may be able to get then working by changing the resistor in series with the input side of the opto switch for a lower value. What resistance are they at the moment?

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

That would help explain of course

The design is an exact copy of the Gen7 option end stop as per here: Wiki

Is this not suitable for 3.3V?

I think that design may be marginal with a 3.3V supply, although I can't be sure until I have access to a PC and can check the schematics and datasheet for the opto sensor. Reducing the 180 ohm resistors to 100 ohms may be needed for reliable operation on 3.3V.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Thanks for the answer - that could indeed explain how some work sometimes and others don't at all...

Can you confirm that replacing the 180Ohm resistor by a 100Ohm will make this work for 3.3V? I'll put my SMD soldering skills up to date in preparation

I bought very similar opto endstops, though I haven't tried using them with Duet yet. If you prefer not to solder the endstop board, then perhaps actually feeding it 5V and dividing the output down to 3.3 could work? Not terribly convenient, though. Alternatively, if you want to avoid removing the existing 180 ohm resistor, adding a parallel 220 ohm one should get the resulting resistance down to 100 ohm (well, 99), so that might be something to consider.

I've taken a detailed look at that design.

With a 5V supply, the typical forward current of the opto switch is 21mA. Assuming we can tolerate up to 0.5V at the output in the un-interrupted state, the load current is 5.7mA. The specifications of the TCST1103 opto switch give the current transfer ratio as 10% minimum, 20% typical. So with 21mA input current, it can handle a load current of 2.1mA in the worst case, or 4.1mA in the typical case. So this design fails to work properly even in the typical case. In practice the electronics may be happy with 1V input for a LOW instead of 0.5V, which is probably why it typically works.

With a 3.3V supply, things are even worse. The typical forward current is 11.7mA and the load current at 0.5V LOW output is 4.1mA. But the opto switch can only deliver 1.2mA in the worst case and 2.4mA in the typical case. No wonder it doesn't work.

I conclude that this an example of poor electronic design which is alas all too common in the RepRap world.

To get these working well, I suggest the following modifications:

1. For 3.3V operation, reduce the 180 ohm resistor to 100 ohms. If you don't have SMD desoldering equipment available, it may be easier to solder another 180 ohm resistor on top of the existing one, giving 90 ohms (i.e. two 180 ohm resistors in parallel).

2. Increase the value of the 1K resistor to 10K, or remove it altogether - it should work just fine without it.

btw the figures above don't take into account the additional load presented by the electronics. But I think the above mods should be sufficient to work with a Duet, at least in the typical case.

Edited 1 time(s). Last edit at 06/29/2016 05:50PM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Great - I'll try this out, thanks for helping out on this issue that is only tangentially linked to the Duet!