What's new
VORON Design

Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members!

Issues with PT1000 after moving to EBB36


New member
I have been having issues with my 3D printer ever since I’ve changed my print head from being controlled via the main MCU, to being controlled by an EBB36 over CANBus. It’s not as simple a fix as “just go back to what you had before” - unfortunately I was silly enough to change literally everything about the hotend all at once - AfterBurner -> StealthBurner, Voron Tap and the EBB36… plus I cannibalised some of the harness wiring for other things.

I believe I’ve got most things to the point of being “pretty good”; however the last thing that has been plaguing me has been horrendous stringing of PLA during prints, which is something I’ve pretty much never seen before. After far too much investigation, I’ve discovered that the PT1000s that I’ve been using have been reporting about 20degC under what they should be.

I’ve used both an external temperature probe attached to my multimeter, and separately another PT1000 probe connected to my main MCU, to verify the problem. When I set the temperature of the hotend to 200degC, the validating probe would report it back as ~220degC! (Note: I had two probes inserted into the hotend at this point, so they were reading as close as possible “the same” temperature/environment)

I later also tried plugging the “validating” probe into the EBB36, and was able to rule out the EBB36 as necessarily being a problem; at this point the validating probe was still just a long wire hanging down from the hotend, long enough to be plugged into the main board underneath as needed.

I decided to try using this validating unit for printing, so I cut the wires down to size and crimped it for connecting to the EBB36. I crimped it for use with the MAX31865, and plugged the old robe into the standard 2-wire JST socket in the other side of the board (I had wired up the old probe for both connectors previously).

Lo and behold, suddenly the new probe started reporting the same values as the old probe! Indicating it was at least 20degC under-reporting at the hotter temperatures.

I cut the terminals off and re-crimped, and had a 10-degree difference compared to the other unit. Supposedly this was an “improvement,” however this all feels very suspect to me.

Other information: the temperature probes follow the same curves, so seem to share thermal characteristics? Only the temperature seems to differ. At lower temperatures the difference is not so severe, being only a degree out at most around room temperature (observed during validation). I did PID tune. I did not move the probe between “validation” and “installation,” so the environment hasn’t changed. I can anecdotally say that I’ve seen this issue one three “PT1000” units now (what was originally installed in the Afterburner and moved to the Stealthburner/EBB, the one that replaced it when I thought it was faulty and performing poorly, and now this validating probe).

So, is this an expected problem with PT1000s? Do I maybe not have “real” PT1000s? Are the leads contributing more to the reported value being correct than I had previously anticipated? What can I do to get accurate enough temps so I can print without PLA stringing again?
So, I couldn't really understand. Do you run the PT1000 with or without the MAX31865?

If without, do you have the correct pull-up resistance? Read somewhere where they had wrong and gave faulty readings.
So, I couldn't really understand. Do you run the PT1000 with or without the MAX31865?

If without, do you have the correct pull-up resistance? Read somewhere where they had wrong and gave faulty readings.

@pws I actually have two PT1000s hooked up, one to the MAX and one to the other port. I haven’t checked the ref resistance for the non-MAX port, but I have checked the one for the MAX, and it appears to be correct within margin of error (429x vs 4300).
It should work but as far as I understand the PT1000 has an internal resistance at 1k at 0C and in the charts I have seen around 1.3 k at 250-300C so you will loose some resolution using the usual onboard 4.7k pullup.
The EBB36's have a jumper that you need to install if you want to use a PT-1000. This is what their web page says:
'Users can select the thermistor's pull-up resistor values through a jumper wire, so as to support PT1000 (2.2K pull-up resistors), which makes it convenient for DIY.'

Not sure if this explains your issue, but I would suspect that the PT-1000 would be a bit more accurate with this jumper in place. (Don't know the in's and out's that would explain why this works better.