Thermocouple Faults to 2000

[dc42]

These possibilities occur to me:

1. You have a short between the braiding of the thermistor and one of the two thermocouple wires, but you don't have that on your other machines.

2. The braiding is grounded on your other machines, but not on this one.

3. The hot end metalwork is grounded on your other machines, but not on this one.

[Billsrf]

@dc42 I have swapped all components 3 times. Still same result. I am using an aluminum enclosure to house everything including the power supply. Could it be a grounding problem with the power supply being grounded to the machine?

[dc42]

@billsrf said in Thermocouple Faults to 2000:

@dc42 I have swapped all components 3 times. Still same result. I am using an aluminum enclosure to house everything including the power supply. Could it be a grounding problem with the power supply being grounded to the machine?

If the PSU case is connected to mains ground and to the machine frame but the negative side of the PSU output isn't also connected to mains ground, that could be part of the problem. But if you use a USB connection to the Duet, see the wiki page on USB ground loops.

[Billsrf]

I unbolted the PSU and remounted it with non metallic fasteners, isolating it from the case. Now it works fine! Now, my question is, should I take the ground on the 24 volt side(-) to chassis ground to avoid this in the future???

[dc42]

The chassis of the machine and the PSU mounting plate should be connected to mains ground for safety. the PSU mounting plate is probably connected internally to the mains ground input terminal. So it would be better to leave the PSU bolted to the chassis, check that the chassis is connected to mains ground, and also link mains ground to the negative output of the PSU. This is what I do on my delta.

[Billsrf]

Is this what you are describing?

[dc42]

Not really, I am suggesting that you connect chassis directly to mains ground, and mains ground to the negative output terminal of the PSU. In this image you can see how I have done this on my delta.

[Billsrf]

@dc42 ok thank you. I will try that also. I did put a piece of captan tape on my thermocouple to isolate it. It works, but I’m afraid it will not read properly. I ordered a coupe of Cartridge TC’s from E3D to try out.

[wilriker]

@dc42 said in Thermocouple Faults to 2000:

[...] and also link mains ground to the negative output of the PSU.

What is the rationale behind this specific connection?

[Billsrf]

There isn't a rational thought behind it. I have tried everything I can think of. I wired it as per your delta picture and it is still faulting when i home the machine, and when the Zero sensor activates it goes in to a 2000c fault mode. I put a piece of captan tape on the TC's ring terminal to isolate it and I no longer get a fault. But, I am sure it will not read accurately.

[dc42]

@wilriker said in Thermocouple Faults to 2000:

@dc42 said in Thermocouple Faults to 2000:

[...] and also link mains ground to the negative output of the PSU.

What is the rationale behind this specific connection?

Safety, in case a fault develops inside the PSU. ATX PC PSUs always have this connection internally.

[Billsrf]

Thank you, I did it as per @dc42 suggested. Now my issue is Z-probe not going to middle of table when I call for home all??? LOL It probes the back corner of the build plate on its first probe point. But, it doesn't go to the other points or the center for last probe point?

[dc42]

@billsrf said in Thermocouple Faults to 2000:

Now my issue is Z-probe not going to middle of table when I call for home all???

Please post your homeall.g and homez.g files, and tell us which firmware version you are running.

[Billsrf]

Here are my files.1_1533747130393_homez.g 0_1533747130393_homeall.g

[Phaedrux]

@billsrf It would appear that you need to remove the negative sign from your move to get the probe to the center of the bed.

G90                       ; absolute positioning
G1 X-530 Y-560 F6000      ; go to first bed probe point and home Z
G30                       ; home Z by probing the bed

You're telling it to go to a negative absolute position, which I don't think you intend.

[wilriker]

@dc42 said in Thermocouple Faults to 2000:

Safety, in case a fault develops inside the PSU. ATX PC PSUs always have this connection internally.

OK, I can see that.
But there comes one question instantly to my mind: if I connect negative output to mains ground as shown in your picture wouldn't those ~12-15A needed to heat up the bed also take this route? I ask because I have my PSU connected to mains with a standard PC power cable that has 3x0.75mm² inside. One for L, N and PE. That would mean a single 0.75mm² would then have to take up to 15A of current. That seems like a good idea to start a fire to me.

So what am I missing?

[Billsrf]

@phaedrux , LOL, Thank you!!!

[dc42]

@wilriker said in Thermocouple Faults to 2000:

@dc42 said in Thermocouple Faults to 2000:

Safety, in case a fault develops inside the PSU. ATX PC PSUs always have this connection internally.

OK, I can see that.
But there comes one question instantly to my mind: if I connect negative output to mains ground as shown in your picture wouldn't those ~12-15A needed to heat up the bed also take this route? I ask because I have my PSU connected to mains with a standard PC power cable that has 3x0.75mm² inside. One for L, N and PE. That would mean a single 0.75mm² would then have to take up to 15A of current. That seems like a good idea to start a fire to me.

So what am I missing?

The current will only go down the mains ground wire if the negative wire from the PSU to the Duet breaks and there is another path from mains ground to the Duet ground such as a USB cable.

[wilriker]

@dc42 said in Thermocouple Faults to 2000:

The current will only go down the mains ground wire if the negative wire from the PSU to the Duet breaks and there is another path from mains ground to the Duet ground such as a USB cable.

OK, I have no USB connection to the Duet and my connection PSU<->Duet is redundant with 2x2.5mm² for GND and Vin each (so a total of 4 wires). Also these wires are not moving so it is rather unlikely that this connection will break (but unlikely does not mean impossible).

I still don't understand why only in your described case the current will go to protective earth (not sure how it is done in the UK but in Germany this is different from the Neutral wire) because - with the exception of the rather small wire gauge - this is supposed to have a very low impedance so that current can take this route for safety reasons.
But then again electricity sometimes does things I don't understand and I am nowhere near to be an electrician so I gonna trust you here and create this connection.

[Billsrf]

Thank you both!!! I was "code" blind and completely over looked the (-) in front of my Z-probe home number.