Diagnosing hotbed (or controller, or firmware) failure after hotend PT1000 thermistor switch  

RE: Diagnosing hotbed (or controller, or firmware) failure after hotend PT1000 thermistor switch

Skip the self check so it goes to the home screen is it reading room temp or random temperature 

Posted by: @aeberbach

My MK3S has a Bondtech LGX + Mosquito setup. Recently I bought a 50W heater cartridge and a PT1000 thermistor. I installed those, then updated printer firmware to 3.13.0 selecting Firmware For LGX And Mosquito", "If you are using Slice's 500C PT1000 RTD and EINSy10A board" on the Bondtech web site. Then factory reset the printer. All seemed to go well; no trouble disconnecting, connecting or updating.

Then started calibration again; fans, axes and home location passed. When it gets to the hotbed it fails after a minute or so of the '\' spinning. I have measured 23.99 V from both pairs on the power supply at the Einsy board. I have measured the voltage across the hotbed terminals and it appears that there is nothing coming from the Einsy board during that stage of calibration.  The hotbed thermistor measures around 150k ohms, the heater element is a low resistance, not open. All Einsy fuses are good.

• Any ideas? I think I am going to plug in the old hotend thermistor and revert to the old stock thermistor firmware to see if that makes a difference.

RE: Diagnosing hotbed (or controller, or firmware) failure after hotend PT1000 thermistor switch

Yes, both hotend and bed read the same ambient temp on the home screen.

RE: Diagnosing hotbed (or controller, or firmware) failure after hotend PT1000 thermistor switch

Try using preheat function on home menu

RE:

The problem is probably the PT1000 sensor.

The prusa works with an NTC sensor by default.

This one has a negative temperature coefficient.

This means that the electrical resistance decreases as the temperature increases.

With a PT1000 sensor, the electrical resistance increases as the temperature increases.

And the firmware does not understand that. Because it expects resistance to go down instead of up.

If you want to use the PT1000 you will have to modify/program/compile the firmware.

RE: Diagnosing hotbed (or controller, or firmware) failure after hotend PT1000 thermistor switch

I did try to set the heated bed to 50 degrees, caused much alarming and THERMAL ANOMALY.

@rene-3 I did load the specific firmware for the PT1000 thermistor. I don't rule out that it does not work as it says it should but the firmware available from Bondtech is specifically to enable use of that part.

Probably it is time to pull the new thermistor out and hope that it works again with the old one. If it doesn't, I actually have a hardware problem. If it does, perhaps I need to go back to old Bondtech firmware and see if previous versions worked. Or even look at diffs to see what might be wrong in 3.13.0 but I really didn't want to spend hours doing that, just print some hotter filaments 🙁

It is annoying because this was specifically to use some nylon/CF filament where the MK3S just won't go over about 295 without halting.

RE: Diagnosing hotbed (or controller, or firmware) failure after hotend PT1000 thermistor switch

Bondtech uses Prusa’s original firmware as a base for tuning it to their LGX extruder (different rotation ratio/microsteps). The version for Pt1000 modifies as well the thermistor parameters. That shouldn’t be a problem because Bondtech has been doing it for a long time. Besides, the issue seems to be related to the hotbed, so the Pt1000 thermistor and the 50W cartridge of the hotend have nothing to do with it.

 I suspect that the issue resides on the 3.13 firmware, which is well known for giving a lot of trouble with thermal anomaly errors. Which firmware version were you using before updating to the 3.13 ? . 

 I’d try reverting to Bondtech’s 3.11.0 firmware for the LGX/Mosquito/Pt1000 to see if the problem disappears.

RE: Diagnosing hotbed (or controller, or firmware) failure after hotend PT1000 thermistor switch

Bondtech don't have a LGX/Mosquito/PT1000 firmware at 3.11.0 (PT1000 begins with 3.12.0) so I put the old thermistor back in and went back to the Bondtech LGX/Mosquito 3.11 - and still THERMAL ANOMALY. I think there has to be something going on with the Einsy because all of the cables, all of the thermistors and heaters measure good. Have done full factor reset after every flash. When I get some time I need to dismantle the controller section and be sure everything is connected right. If the Rambo is dead that might be it for this printer, I don't know if I want to spend $100 plus on a new one.

RE:

Sorry, my mistake,  I didn't realize that the Pt1000 wasn't supported until 3.12.0. 

My current setup is a MK3S+ with LGX/Mosquito fitted with Prusa 40W heater and thermistor, running Bondtech's 3.11.0 firmware. No thermal anomaly issue of any kind. If I understood correctly, before installing a 50w heater and Pt1000 thermistor,  you were using the same setup and firmware with no issues ?. If that's so, either you're right and there's now something wrong with the Einsy or else the 50w cartridge heats up the nozzle too fast and that triggers the error ?.

RE: Diagnosing hotbed (or controller, or firmware) failure after hotend PT1000 thermistor switch

With everything disassembled I connected the heatbed directly to the power supply and it heats well and evenly so it isn't the heatbed. I'm suspecting Q5 on the board, there is not much else to go wrong there, if in fact something is wrong. I am hoping it all just works once everything has been carefully checked and put back together. But I'm also looking forward to verifying that Q5 is bad and replacing it for $3 rather than buying a new controller board. It's annoying that Ultimaker used Altium for this board rather than something free but I have access to it through my course so it only cost me time to install.

RE: Diagnosing hotbed (or controller, or firmware) failure after hotend PT1000 thermistor switch

After reassembly it is still telling me that I have no heatbed/thermistor connected, even with other thermistors in place of the one on the bottom of the heatbed. I can't think of any explanation other than a bad Q5 MOSFET - next step is to solder a wire to the gate of Q5 and look at the logic level coming from the controller during the calibration wizard when the heatbed test fails. I think I am going to see it try to activate that device but the device doing nothing. The heatbed always has DC 24VC connected but the other terminal floats, and should be connected through to negative via the low-side MOSFET when the controller indicates - this would normally complete the circuit and begin heating the heatbed, resulting in an observable change via the heatbed thermistor. Eventually the controller is seeing no change and failing. If I put my finger on the heatbed thermistor I see instant rise in temperature so no issues there.

RE:

Looking at the schematic it becomes clear that the gate of the Q5 MOSFET is driven with the AND result of the pwm output from the processor and the nAC_FAULT signal from the processor. And once I searched for nAC_FAULT I found that it is really common for this to be caused by the power panic wire.

I disconnected the power panic wire and the printer has started passing self test. Why I don't know but for now it stays disconnected.

(Why doesn't the display show "Power panic failure" instead of telling me the thermistor is disconnected! 🤬)

RE: Diagnosing hotbed (or controller, or firmware) failure after hotend PT1000 thermistor switch

And now it has completed a 14 hour print. Filed a bug asking for a better message on the display when power panic is the cause of the issue.