New Tronxy P802M extruder heating strongly delayed and overshoot

Hi Everybody

I am a very new user of a Tronxy P802M which I got myself as a Christmas gift. I am a complete newbee in 3D printing, so please bear with me.

I have assembled the printer according to instructions and some youtube videos and the first check looks promising EXCEPT one strange issue with the extruder heating: If a temperature of eg 190 degrees is set, it takes several seconds until heating starts finally (about 10-20 sec). Then it heats up to >= 270 degrees. When it finally decides to stop heating, the temperature falls back to about 110 degrees to then rise to the >270 degrees again. Sometimes it stops with a "def" as a temperature reading, when the temperature reaches 300 degrees.

I have no idea, why that happens. I tried to get some details by inserting a current meter into the power for the extruder heating. It pretty much complies with the temperature curve I described. I made a video showing this issue: [www.youtube.com]

Can anyone help me here? Is this a firmware problem? Or a hardware problem with the control board?

Thanks a lot (and a Happy New Year!)

Michael

If I had not seen the video would have never believed this could have happened. It looks as though the firmware is not responding to the temperature probe (thermister).
Will do some searching and see if I can find you some information for you. Though someone else may already know right off what may be the problem is.

Some people recommended me to change the PIDs. I therefore did a couple of tests with the PID settings etc. In general all of them did not solve the issue, but some even were not reproducible, which makes me think that either the firmware is completely bork or there is some hardware defect.

Here is what I did (in chronological order with the results):

• changed from "Dead time" to "PID" (without changing PIDs) -> no change
• changed to "BangBang" -> No change
• changed back to "Dead time" and adjusted the PIDs to P=40, I=0, D=0 -> seemed to work!!! Temperature was between 180 and 198 degrees.
• changed I to 2 -> old behaviour
• changed I back to 0 -> still old behaviour!!!

So, the machine is now running with P=40, I=0 and D=0 and it still shows that old behaviour.

Some other interesting facts:

• The delays until the current of the output is adjusted (after the LED turned on or off) differs. Sometimes it only takes 3-5 seconds, sometimes I need to wait >30 sec for a reaction on the current output.
• The state of the current output seems to be "surviving" a complete machine reboot. Eg if the output is ~3 Amps, temp at (say) 260 degrees (so way over the 190), then I turn the machine off for 5 seconds and on again, the MosFET output is active immediately with a continued current flow of 3 Amps (so heating continues). -> This is very strange. The PIN output of the ATMega should have been resetted. So there should be no current at all. (This would lead to the conclusion that either the firmware is really completely unusable, or that there is a hardware defect either in the ATMega or on the board.)

I did check some connections on the board. The Gate of the MosFET is connected correctly with the corresponding pin of the ATMega (so no disconnection of the lines). Unfortunately, I have no equipment to due some measurements during operations. It would be interesting to see the output of the ATMega on an oscilloscope (to see the PWM and to measure the voltage of the PWM signal).

Im interested in this as i had something similar happen after printing for 14months. In my case the temperature was much more eratic. Absolutely no change to the firmware.
Thought it was probably the thermistor but changing the complete hot end only changed the max temp a bit = as you would expect with a different heater and thermistor. So i concluded that its a problem on the control board.
Its hard to see from the video but i noticed on mine that it seemed to be heating when the LED was off. Seems to just max out the heater.
I tried with the bed turned off to check it wasn't using that to drive the hot end.
I couldn't see any dodgy joints on the board. But im suspicious that there is some sort of short/break causing cross talk.
The controller actually switches the earth side so somesort of leakage to earth is a possability.
Those terminal connectors are prone to melting. I would advice you to inspect them every few months. I just rewired mine with seperate 12v direct to the bed, hot end and the board to try and reduce the temperature.
If you connect to a host program such as repetier you can look at the temperature trace and the PWM signal for a clue on what its trying to do.
In your case check what the max temperature setting is in the firmware as this can confuse your investigation. Thats because if you hit the max temperature limit the firmware will turn off the heater instead of the PID control.
Its hard to believe its the firmware as the temperature reported seems real.

Have you tried PID Auto tune. This will take the guess work out of your numbers.

// FIND YOUR OWN: "M303 E0 C8 S190" to run autotune on the Hotend at 910 degreesC for 8 cycles.
// FIND YOUR OWN: "M303 E-1 C8 S90" to run autotune on the bed at 90 degreesC for 8 cycles.

[reprap.org]

I tried PIDs with no I nor D factor (see above). Only P should make it a simple on/off with no interpolation and delay. It did not change anything.

I also flashed now a new firmware successfully. The result was, however, pretty identical with the former experience. Actually, it also repeated the last experience, in which the very first run seemed to be better, so that I thought it's fixed now. But then it soon went back into the odd behaviour. I think that this clearly shows that the fault is a hardware defect. I therefore need to see that I get a replacement board as soon as possible. (I was sooo happy to have the printer built and really wanted to start printing some things now...)

Btw, I would love to know if it is the MosFET or the ATMega. I wonder if I should get some cold spray to check if cooling will help, and if yes, what part it is.

Another question: If in the worst case, I would need to purchase a new board myself, ie Gearbest would not send me a replacement, what board would you guys recommend? Should I then get a different board or stay with the Melzi?

OK, it seems to be the MosFET. I have been able to measure the Gate voltage controlled by the ATMega and concurrently the current output. The ATMega correctly goes to 5V or 0V depending on heating on or off. But the output of the MosFET is not following. So, this seems to be the culprit.

Does anybody know, if this MosFET would be a good replacement? (It's the only one I can get hold of easily.) [www.irf.com]

PROBLEM SOLVED: The MOSFET was defect. I exchanged it with a similar one and now the printer works

And we all learned that a Mosfet can partly work and still be defective.

Edited 1 time(s). Last edit at 01/09/2018 03:09PM by Roberts_Clif.