killed 3 1XD's due to 5V from step and direction?

Dutchprinting

Somehow after a while the 1XD connected to my servo driver for the extruder killed itself, looking at the wiring i was afraid it shorted itself. i fixed the wiring and put a new one on. Today that one gave up out of nowhere, first the out0 worked but the step+dir didnt. after a restart there was just the blue LED and nothing else.

So i disconnected it, grabbed a spare board and put back the connectors. while plugging in the step+dir i saw the red 5v LED come on. As that could be a red flag i turned to the PC and googled why servo drives emmit 5v signal on the step+dir. I found out this was normal bc they are internally pulled up. So to be sure i turned off the servo. connected and powered the 1XD, then turned on the servo. as soon as i turn on the servo the board goes back to only the blue LED. no response anymore.

So my questions are... can i fix the boards? and Is this normal or is the 5v from the servo a problem?
i measure 5v between step+ and step-, same goes for dir+ and dir-

Everything worked fine yesterday....

Kind Regards,

droftarts

@Dutchprinting What make and model of servo driver are you using? Usually the step and direction inputs are opto-isolated, or at least have a diode. Also, as the 1XD supplies 5V, maybe it's the current that the servo driver is outputting that is the issue, as each pin of the DRIVER_OUT connector can source/sink a max of 10mA (see https://docs.duet3d.com/Duet3D_hardware/Duet_3_family/Duet_3_Expansion_1XD#description-of-connections).

I did a google search too, but couldn't find anything talking about the driver emitting 5V. If that's the case it really sounds like the driver is faulty. I'm not sure what will have failed on the 1XD, possibly just the 5V regulator (see https://docs.duet3d.com/en/User_manual/Troubleshooting/Parts#voltage-regulators), but may be the MCU. One for @T3P3Tony.

Ian

Dutchprinting

@droftarts Thank you for your reply, i am using the PARKER COMPACT X3 servo driver, it is a huge beast with a very in depth documentation. (linked below)
I guess i have to find out why the driver is suddently emitting 5v through the step and direction. There is no option for that in the configuration and i didnt change anything

Documentation:
https://download.variodrive.nl/Parker/Compax3/manuals/C3I10T10_eng.pdf

dc42

@Dutchprinting thanks for providing the link to the manual. I've looked at the manual and according to my reading of it, it provides two different STEP/DIR inputs:

Plug X12 provides step/dir inputs that require a 24V signal level. These are intended to connect to PLCs and are not compatible with Duet.
Plug X10 provides an RS422 interface with differential step/dir inputs (step on pins 6 & 7, dir on pins 8 & 12. However, pins 1 and 9 of X10 must be jumpered together to put the port in RS485 mode. If you don't do that, then the port switches into RS232 mode. In that mode it applies +/-12V to some of those pins, which is very likely to damage a 1LC connected to it.

This use of a pin as both a 5V input and a 12V output is very unusual and IMO very likely to result in damage through mis-wiring. When you next connect a 1XD to X10 you may wish to connect series resistors of perhaps 4K7 between the 1XD and the driver to reduce the risk of damage.

Dutchprinting

@dc42 Thank you for the reply, i dont use X10, but instead use X11 (page 62). i have pins 11 and 9 jumped, for other reasons and indeed step and dir on 6,7,8,12
X10 is where the cable to the PC goes into for configuration, i do not have a break out board for that. only for X11 and X12

Could a fix be wiring it as a single ended connection? only wiring the step- and dir- to the 1XD

EDIT: it has always worked in the current config,
EDIT: i have the Compax3S150 i10T10

dc42

@Dutchprinting I'm sorry, I looked at the X11 pinout then when I went back to look at the connector I looked at X10 instead.

Reading +5V on the +step and +dir input looks plausible because of the bias resistor shown in section 3.8.2.2 on page 62. However, applying 5V to a 1XD output through 1K should not have damaged the 1XD if the ground potentials were approximately equal.

I suggest you connect GND on X11 to a ground on the 1XD or to the negative output of the power supply that powers the 1XD.

You could connect series resistors between the 1XD outputs and the X11 inputs to provide better protection. Unfortunately they have applied a very strong bias to the inputs, so the values of the series resistors can't be as high as I would like. Try 470 ohms.