Issues powering up with DUEX5 connected to Duet2 crashing

[Dougal1957]

Posted for a friend who is having trouble joining the forum (Details passed separately)

I have tried a bunch of times to register but it keeps saying the captcha is unverified. So I can’t even get on to ask a question. Heh. But this is my issue. When I power up, and both boards are connected to 24v as power in (BUT THE EXPANSION CABLE IS NOT CONNECTED) everything boots up fine and I can surf to the web interface. I’ve done a M115 with just usb connected AND the expansion cable connected and it reports all hardware correctly. I’m running 2.04RC1. When I connect the expansion cable and 24v is also connected, it stalls and the diag light blinks. A few times on the panel due screen it’s popped a motor shorted to ground error and a temp shutdown on a motor. Because of that I also unplugged all motors and repeated with everything connected. Still no go. If I remove just the 24v power, but it’s running on the 5v psu, it will power up and allow me to see the web interface. I’ve done the short and thick ground cable between boards.

This is with 5/12v psu running, cable attached, but no 24v

I’ve checked for grounding issues and power are spot on at 23.99~24v

Thanks for any help

[Dougal1957]

@dougal1957 A bit more info from Nuramori

[Dougal1957]

@dougal1957

[fcwilt]

Hi,

Those power connections look rather thin - what gauge wire are you using?

Frederick

[Dougal1957]

there is no load on them at this stage

[deckingman]

@dougal1957 said in Issues powering up with DUEX5 connected to Duet2 crashing:

there is no load on them at this stage

Even so.....Have you seen this Doug? https://duet3d.dozuki.com/Wiki/Duex2_and_Duex5_Features

Particularly, if you scroll down a bit you'll find quote........

"Important! You must have a reliable low-resistance connection between the negative (ground) terminals of the two VIN terminal blocks. Failure to do so may result in high currents flowing in the ground connectors of the ribbon cable and may cause damage to the Duet 2 or Duex. Preferably, make this connection using either solid core wire, or stranded core wire with ferrules of the correct size crimped on securely. Re-tighten the terminal block screws regularly for the first few days of use, to make sure there is no creep in the wires causing the screws to become loose. "

Might be worth checking. There is some stuff in that link abopt the ribbon cable too (in case you haven't already seen it).

Cheers

[fcwilt]

@dougal1957 said in Issues powering up with DUEX5 connected to Duet2 crashing:

there is no load on them at this stage

Not sure what you mean.

If the boards are powered up there is current flowing through those wires.

The way I have wired all of my Duet2/Duex5 setups is by connecting both the + and - power terminals with solid 14 gauge wire. Then I tap into the middle of each with stranded 14 gauge wire and run those to the power supply.

I have never had anything like the issues mentioned.

Frederick

[Dougal1957]

if you look the main feeds are actually 16 AWG but the link between the 2 boards is quite a bit thicker 12AWG Apparently by No load I mean no motors or anything connected this is weird and is not mine so posted on behalf of a friend who every time he tries to register the Captcha throughs an error and won't complete.

[Dougal1957]

Apparently the issue sort of goes away if he disconnects the 40 way ribbon and further test seem to show that with it all connected up on the power and 40 way stuff the board will boot up and stay up with USB Connected but when the 24V is applied the boards reboot at approximately 15s cycle which is also very weird.

Cables from the PSU are 16AWG and the Ground link between the boards is 12AWG

[fcwilt]

Has he tried a different power supply?

Frederick

[dc42]

Are any endstop switches or other devices connected to the endstop inputs on the DueX5? If yes, disconnect them and try again. If no, then I suspect that the DueX5 is faulty so please ask your supplier for a replacement under warranty.

[Dougal1957]

@dc42 Thank you David I will pass it on this is the user I emailed you and tony about the Captcha is failing to allow him to register.

Doug

[Nuramori]

Hello all. I was able to get registered, and I thank Doug for assisting in being my proxy until I was able to get situated.

I’ll address some of the points raised and the results of the last suggestion.

All power wiring is 16awg; the ground connection between the duet and duex (as seen labeled on the wire in the first picture) is 12awg. I performed the last recommended test (unplugging the E6 endstop connection I had) and rebooted. The result was everything worked perfectly. The connection is a NO switch connection. It’s part of an emergency stop switch that’s NC/NC + NO. The two NC is hardwired to cut power to the 24v psu and the NO is intended to be connected to E6 so when the button is activated power is cut and the duet can respond to the condition (the duet is powered by a separate 12v/5v psu). I then moved the connection all other stops on the duex, and the issue would repeat on each of those, always working fine once it was removed. When I attached it to the E1 endsrop on the Duet, it would work as intended and not exhibit the problems seen when attached to a duex endstop.

I did note that there was a post 1.19 change to configuring endstop use, but when commented out the issues remained, so it wasn’t the config file that contributed to the issue.

[dc42]

What's happening is that your NO endstops are picking up a lot of noise. This is a hazard with NO endstops, and is one of the reasons why NC endstops are preferred. So I suggest you change your endstops to use the NC connections. Alternatively, connecting a resistor of between 1K and 3K between the STP and +3.3V pins of the endstop connector may work; or a capacitor of about 10nF in parallel with the endstop switch.

[Nuramori]

@dc42 thanks for the info! I’ll see if I can find another emergency stop switch that has more than two NC and/or the other two suggestions. Appreciate the help.

[Nuramori]

ok, I split the difference and put a 2.49k pullup resistor between 3.3v and the stop. all is well now!

Thanks to everyone.