Large grounded lug
-
I had to install a shielded twisted pair for the PT100 as the extruder generated so much EMF. So it would be nice to have a large grounded lug / screw point for things like shielded cable or voltmeter testing.
-
Having just tried tracking a problem with my power supply I would like this too.
-
Was it a 2 wire or 4 wire cable for your PT100? What did you end up using as shielded cable? I'm about to hook mine up and have been trying to figure out what to use. I was going to go for 4 wire, but didn't think I would need shielding.
-
A general recommendation from having cured a lot of homebuilt CNC machine problems:
Provide your own bolt (not on the Duet board), a tall bolt, for central "star" ground. Run one of everything to it. One from the power supply ground. One from the ground side of the input to the Duet. One from each shielded wire. IMPORTANT: DO NOT ground BOTH ends of a shield, ever. This creates a bevy of problems. Only ground one end of a shield. If you are powering anything "out in the machine", like an optical endstop, one from the ground side of each thing you are are powering. Each wire separate to the boldt/star. No sharing.
Take the time to crimp (not solder) a full ring terminal on each thing you run to the bolt. Stack them all and tighten firmly.
Doing this methodically has, as mentioned in the intro sentence, stabilized many a CNC that has much higher voltages/currents, and large amounts of static generated by the dust collection system.
There is a lot of debate as to whether this single star ground should, or should not, be tied to earth (mains in the UK, wall socket in the US) ground. Personally, I say "no"… unless... absolutely required to make the machine stable (I've seen it).
Anyway, if you are fighting instability, try a star ground system.
And, for the record, I would NOT want the bolt on the Duet.
-
Was it a 2 wire or 4 wire cable for your PT100? What did you end up using as shielded cable? I'm about to hook mine up and have been trying to figure out what to use. I was going to go for 4 wire, but didn't think I would need shielding.
Hi,
2 wire. Currently I am using a 2 core shielded pair for the PT100 that is meant for a stage microphone type use, ie its supposed to be Ok with a lot of movement/flaxing. It has lasted a year so far no issues but the mount point on the hotend is a bit crappy so that is down to fix.
The fix is I am wiring up a CAT6A socket so the cable is easy to replace. This is (obviously) 4 twisted pairs, shielded and stranded 26awg copper, problem is I need 10 wires and I have only 8 so I might put the extruder, cold end fan and part cooling fan circuit through just 1 and leave the PT100's cable as is.
In terms of 2 v 4 wire my understanding is with 2 wire I lose accuracy but that is a constant offset that does not vary in the temperature range?
So I print PLA at 206~208C where I think the offset is around 1.5C but who cares as long as the control is very good around whatever I set? Certainly after monitoring this config for a year the temperature variation is mostly 208C +/- 0.1C with a max variation of 0.2C occasionally seen on the web ui.
In comparison the bed with a conventional thermistor and more mass is varying 0.5C +/- at 65C so I am very happy with the duet's PT100 temperature control in this respect.
-
Oh to clarify this is on a Wanhao D6 which has a known constantly failing 16 way ribbon cable design fault. So the idea is to try and remove this issue or at least make the change quick and cheap.
-
How about using two flat CAT6 cables? One for low voltage stuff (temp sensor and Z probe) and one for heaters and fans?
Mind the heater current doesn't exceed the cable rating.
