Faults being reported involving the hot end heater

bjackson

I've been experiencing some heater shutdowns regarding the hot end. What has happened twice at least is I get both the bed and the hot end pre-heated, and then start a print job. The print starts, and may run for several minutes until suddenly it says:
Error: Heating fault on heater 1, temperature excursion exceeded 15.0 Deg C.

Now for some recent history:
I performed the M303 H1 S210 to get auto tuning done. The process completed, and it warned that Heater 1 appeared to be over powered. If left on a full power, it's temperature is predicted to reach 651 Deg C. Auto tune heater 1 completed in 275 seconds.

I did the M500 command to write data to config-override.g, and also did M301 H1 to take a look at the results. What it came back with was:
gain: 626.9, time constant: 188.2, dead time 12.6, max PWM 1.00, calibration voltage 24.0, mode PID, inverted no, frequency default.
The computed parameters were:
P4.2, I0.087, D37.5. Computed PID parameters for load change: P4.2

After doing auto tune on the hot end, I did the same process on the bed.

I have attached my config.g and config-override.g in case there are clues within.

Other things I have done were to double check the integrity of the connections to the heater cartridge and the thermistor near the hot end. At one point, I disconnected the heater cartridge from the system so I could measure its resistance at room temperature. It read about 14.8 ohms, which would make it about 38.9 watts, and suggests that it is made for 24 volt use.

The hot end components are still rather new, having seen little use to date.

Here are the configuration files:

config.g config-override.g

I guess I have some redundancy there in my configuration insofar as there's tuning parameters in the M307 lines of config.g, and also the data in config-override.g that would be invoked with the M501 command near the end of config.g.

jay_s_uk

Do you have a sock on your hotend?
Do you have a part cooling fan? And does that part cooling fan start around the same time your hotend faults?
Did you tune with the part cooling fan on and with the hotend close to the bed?

bjackson

Do you have a sock on your hotend?

No sock at this time. I'm still using the totally stock FT-5 extruder (evidently a version of the MK8 hot end), and FolgerTech did not provide a sock.

Do you have a part cooling fan?
I recently added a layer cooler fan shroud and fan. I don't really know if the fan is too strong or not.

When I tuned it last night, the layer cooler was off. I don't recall what the situation with the bed was.

Did you tune with the part cooling fan on and with the hotend close to the bed?

This morning, I did another tuning, and this time I had the nozzle close to the bed, and the bed was at 55 Deg C, and the layer cooler was on at 50% power.

I certainly got a different set of model parameters:
A353.2 C117.2 D10.9, so I loaded those into the M307 command and preheated the hot end. During the pre-heating, the layer cooler was off, as it normally would be during the pre-heat and first layer.

I got faults to occur just on the preheating to 210 Deg C. I just watched it, and shortly after the temp read slightly above 225 Deg C, the fault message appeared.

Overshooting the temperature set point seems to be the primary problem here. The layer cooler fan activity certainly has an effect on things, making me wonder if it's too much air.

As an experiment, I tried increasing the A parameter by 10 %, but it would still overshoot enough to fault. I also tried just changing the S parameter to 0.6 to see what the system would do. I did not run a tune like that, I just changed it in the config.

I guess that effectively reduces the wattage of the heater cartridge. It did not overshoot enough to fault, but it seemed like the temperature drops when the heater was off seemed to go a bit lower.

I'm wondering if things would get better if I did get a sock for the heater block. It seems like that would insulate the block quite a bit from the effects of air blowing from the layer cooler output ports.

Two questions come to mind as a result of this issue. When printing with PLA, how much excursion above and below the set point (210 for example) is tolerable before you will be able to see effects in the print?

If the effects of the layer cooler are too strong (I found no guidance so far on what fan I should be using), then I'm wondering if I could make changes in the slicer to call for less cooling that it normally would.

Alternatively, if it really blows to hard, I suppose one could make an alternate adjustable port on the fan shroud that it would allow some of the air to just blow out where it does nothing, thus reducing the flow rate right at the nozzle where it matters. Of course, I could get another fan.

Phaedrux

A silicone sock on the hotend will help a lot. It will retain heat better and will shield it from the part cooling fan. Not to mention keep it cleaner.

If your duct is blowing on the hotend directly that will certainly have an effect. Try adjusting the duct so that it's minimized.

M307 H0 A116.4 C605.8 D2.7 B0                  ; sets bed heater (H0) to use PID tuning parameters, and not bang-bang
M307 H1 A626.9 C188.2 D12.6 B0                 ; set extruder heater (H1) to use PID tuning parameters, and not bang-bang

I'm surprised that the deadtime on your bed heater is only 2.7 seconds but on your hotend is 12.6 seconds. I would expect the reverse situation.

bjackson

A silicone sock on the hotend will help a lot.

I should be ordering a few supplies soon, and I will get some socks. My little Ender 3 has one I noticed.

If your duct is blowing on the hotend directly that will certainly have an effect.
I studied the design of the fan shroud outlet ports, and it does look like the designer was definitely trying to direct the air down on the part, and not toward the heated block.

On the subject of this layer cooler: I decided to reduce its output. I thought of a really easy way to do that since it's powered by a 40mm x 40mm fan bolted to a fan mounting port. I just put some washers between on each of the four screws. Now some of the air escapes through the four narrow slots created, and the flow is a bit less near the hot end. If I want to reduce it further, I can add four more washers.

I would expect the reverse situation.
Not sure about that; those are just the auto tune numbers I got. It's the stock FT-5 bed with a piece of mirror glass on top.

In an effort to get things stable enough that I would not have those "in excess of the 15 Deg C" overshoots that were causing faults, I tuned it again using 0.8 for the PWM parameter. I got a new set of numbers for the M307 command. So far so good insofar as the part that aborted yesterday evening looks like it's going to finish today.