heated bed temperature fault

vandermeulen

Sorry to bother with this topic again, but after days of testing I still keep getting heater errors when touching the limit of 90°C.

Heating up goes perfectly in a straight line to the max.. Any value below 90°C and the control works. Only at the max. the heater fault is called.

Changed the sensor and cable, ran autotune, checked previous versions that worked, but no progress.

The 1.800W bed is switched with a solid state relais.

Firmware is 2.05.1 (no time to change to 3.3)

T3P3Tony

@vandermeulen which heater fault is reported? Also what do you mean by "checked previous versions that worked"?

vandermeulen

@t3p3tony Ah of course. The "...x.x°C/s" type is displayed.
And I have made and sold three identical printers, where the problem never occurred. The same settings and config's. Only difference is maybe that we updated this last printer to firmware 2.05.1.

T3P3Tony

@vandermeulen so the temperature is rising too slowly?

Also you mention it rises fine to 90C - I assume thats the set point? and then how long after remaining at 90 does it fault? Do you see any noise on the temperature signal? What happens if you set it to hold the temperature at a lower temp (e.g. 60) how about if you set a higher temp than 90 (e.g. 110?)

vandermeulen

@t3p3tony The limit in the config is set to 90. Setting the temperature to 89 works fine. Only when the setting is identical to the limit, the fault occurs. Heater is strong enough, rises 60 degrees in a few minutes, and the curve is smooth. I've replaced the sensor and cabling, no change. Playing with the safety limits with the M570 command didn't seem to change behavior, not did it help.

T3P3Tony

@vandermeulen please set the limit a bit higher than you want to normally use the heater at, then it should be ok.

vandermeulen

@t3p3tony This is hardly a solution for a printer I professionally sell. And how can it be different from the previous three? I don't get it and want to know what is going on.
Maybe this all sounds a bit grumpy but after hours fiddling and customer pressure I'm not happy.

T3P3Tony

@vandermeulen said in heated bed temperature fault:

@t3p3tony This is hardly a solution for a printer I professionally sell. And how can it be different from the previous three? I don't get it and want to know what is going on.
Maybe this all sounds a bit grumpy but after hours fiddling and customer pressure I'm not happy.

I am not sure why this worked on previous versions of RRF<2.05.1.

Why is it an issue to have the limit higher than the set point? If you set the limit at 90 and the set point at 90 then only a slight deviation above 90 will cause a fault. If you set the limit slightly more (more than potential overshoot) then that slight fluctuation won't cause a fault.

zapta

@vandermeulen said in heated bed temperature fault:

@t3p3tony This is hardly a solution for a printer I professionally sell.

Close loop stabilization involved in above/below errors that are corrected by the loop. That's why you don't want to set your printer such as a small error above will shutdown everything. As @T3P3Tony suggested, you want to have margin, that's normal and reasonable to any control loop.

vandermeulen

@t3p3tony The issue is that the user can set the bed to the maximum and then receive an error. Maybe until now nobody has gotten into this situation and I'm just experiencing it during testing.
It feels strange not allowing the maximum temperature. In a car you run against the rev limiter, but the engine doesn't require a restart.

vandermeulen

@zapta Is this really the case? In the config file you set the extruder temperature limit to eg. 250°C, but then the user is allowed to use 250°C? Is it even possible to set limits in a slicer below the values allowed by the printer?

Stephen6309

@vandermeulen Having the max temp set to 90 When you set the temp to 90, it'll fault since the temp will go above that to maintain 90.

zapta

@vandermeulen, I see what you are saying. I set my nozzle max temp to 260

M143 H1 S260               ; Max temp 260C

PanelDue rejected settings to 260 (too high) but accepted 259. The heater reached 259, overshot a little (that's normal) and triggered a heater fault at 260C. (not good).

I think that the problem is that RRF doesn't distinguish in this case between max requested temperature and max expected temperature.

If you the user sets temperature below max temp (e.g. 200), RRF computes some max expected temperature (e.g. 210) and failed on that temperature (good). However, if the user specifies the max allowed settable temperature, RRF seems to use it also as max expected temperature which causes an false failure condition.

(I am using RRF 3.3)

engikeneer

@vandermeulen the issue is that the temperature limits in RRF are a max temperature limit for the system I.e. if the measured temperature ever goes above that, it throws a fault.

Most likely the PID tuning varied slightly between the different units so the others were slightly overdamped and never overshot the 90C demand. For whatever reason, this third unit tuned differently and overshot the demand slightly on heating so throws a fault.

You can probably play around with multiple M143s with different actions if you want a better solution. E.g. one at 90C as a limit which takes no action, and one at 100C that will fault if exceeded.
https://duet3d.dozuki.com/Wiki/M143

zapta

@engikeneer said in heated bed temperature fault:

You can probably play around with multiple M143s with different actions if you want a better solution. E.g. one at 90C as a limit which takes no action, and one at 100C that will fault if exceeded.

What should be the 'A' values for the two M143 commands?

Also, will the user be limited to request temps up to 90C but not above?

engikeneer

@zapta I was thinking of using the C-1 parameter to disable the monitor

C Condition for temperature event (0: Temperature too high [default] 1: Temperature too low, -1: Monitor is disabled)

Also could set it to use the PS_ON trip if you're not using it. That will trip the PS_ON pin but I think doesn't actually cause a fault to be flagged (I think)

Not tested it at all though...

zapta

@engikeneer , I played a little bit with the monitors. They are also visible in the object model section of the DWC. There are three for each monitor and the ones that are not used are marked as disabled.

This style of configuration via gcodes seems to me more and more obscure as time passes.

vandermeulen

@stephen6309 But that's an impossible user interface design. Just cannot be in a commercial product.

vandermeulen

@engikeneer Okay, I hadn't found the M143 command. Let me try.

zapta

@vandermeulen said in heated bed temperature fault:

@engikeneer Okay, I hadn't found the M143 command. Let me try.

While you play with M143, you can loo at the heater's monitors in the DWC object model section. It will show you the internal affect of your M143 on the RRF internal configuration.