Duet 0.8.5, 1.09m and Thermistors

I'm moving a printer from a Smoothie to RepRap Firmware.

On a SmoothieBoard 5XC my heated bed and extruder thermistors were setup as EPCOS100K in Smoothie and read room temp within a degree or so of a temperature probe on my DMM.

With Duet 0.8.5 I'm getting readings 8 or 9 degrees C above room temp as measured on the DMM. Beta value is set to 4200. I can adjust it downward using the H parameter but I have to use like -130 or so which seems like a lot of adjustment.

Any suggestions?

Edited 1 time(s). Last edit at 12/21/2015 05:20PM by WZ9V.

If you have any 100K resistors, connect two of them in place of the thermistors and see what H parameter you need to get them reading 25C. I think 130 may be just within the tolerance of the ADC, but I can't recall hearing of anyone needing a correction greater than about 40.

---

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].

OK, I'll give that try. Not sure I have any 1% ones handy but I can check them with the DMM to see how close they are.

I found a 2% 100K resistor that measured 100K on the DMM so I tried it with that and still had to use pretty large H-130 (bed) and H-115 (extruder) to get the temps to read 25C so I guess that's just what it needs. Is it usual for the values to jump around within 0.5C on those channels?

With the 100K resistor I can see it jump as much as 0.5C

On a Duet 0.6 with 1k series resistors, if you connect 100K thermistors then it's normal to see the temperatures jump around by about 0.5C because at that temperature, you are at the limit of resolution of the ADC. The reading is more stable at higher temperatures where it really matters. However, on a Duet 0.8.5 the reading at 25C should be more stable.

Two questions:

1. If you set H0, what reading do you get?

2. Can you confirm that each thermistor is connected directly to the two pins provided for it, and that the two wires run close to each other over their whole length? Twisted pair is best but not essential.

3. Have you tried disconnecting the thermistors and checking with a multimeter that neither of then has a short to the heater block, the printer frame, or anything else?

---

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].

I'll work on getting H0 numbers. I was just trying to print so now isn't a stable time for that measurement.

Yes thermistors are connected to the correct pins to the best of my knowledge (reads in correct box in web control). The fluctuating reading also happens with the 100K resistor which is really short wiring. I crimped a connector onto each lead and popped it in a 2 pin header.

The thermistors do not have any shorts and the wires are run close to each other (but not twisted).

It does appear to have more stable readings at temperature. Its workable for now, but it will be interesting to see what happens when I swap to the E3D v6 with its thermistor. It could be something funky with these thermistors or wiring. I'll try rigging the other ones up and see how they do. I'm just tying to get this one going enough to put out a few parts I need for the rebuild.

Rebuild = Metal Corners, 1M Verticals, Onyx heated bed (285mm usable), UMBEE Extruder, E3D v6 and IR sensor. Not necessarily all at the same time or in that order.

The H0 values with beta 4200 using the 100K resistor are
Exturder 0 input - varies between 30.8C and 31.1C
Heated Bed input - varies between 33.4C and 34.0C

from config.g
M305 P0 T100000 B4200 R4700 H0 L0
M305 P1 T100000 B4200 R4700 H0 L0

Since the bed thermistor wires were easily accessable I went ahead a twisted them together is a loose twisted pair.

I am tempted to suggest that you ask for a replacement board. With a 100K resistor connected, IMO the temperature reading with no H correction should be closer to 25C than that and should be more stable. Even if your 100K resistor is out by 2%, that would only change the reading by 0.5C. On my delta, the temperatures jump around by 0.1C at room temperature.

I wonder if it is possible that the series resistors on your board are not 4K7 but somewhat lower. Please can you restore the H value to -130 and try one or two other value resistors in place of one of the thermistors. With T100000 R4700 B4200 as you have, you should get the following:

10K 83C
4.7K 107.6C
2.2K 136C
1K 170C
470R 208C
220R 254C

If the readings are about right using a 4.7K resistor but are less accurate with lower values, you can use the L parameter to correct the low resistance/high temperature end. If the readings are grossly inaccurate with a 4.7K resistor, that would suggest that the series resistors on your board are not 4.7K.

If you want to check the markings on the series resistors directly or measure them, they are R19 (E0), R21 (bed) and R53 (E1).

Please also confirm that you do not also have a DueX4 expansion board connected.

---

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].

No DUEX4 expansion. I'll look at the other resistors and see what I can determine. I know they are not the 1K ones because with those settings it read around 60C so I assumed it had the 4.7K ones.

This is a Duet 0.8.5 from Think3DPrimt3D purchased about a month ago if that helps,

Edited 1 time(s). Last edit at 12/24/2015 08:57AM by WZ9V.