Extruder Temp Regulation

[pyronaught]

What is the typical temp swing above and below the set value on the extruder temp? For a set value of 225 I'm seeing the sine wave swing between 222 and 230, which seems way too high. S3D shows a higher resolution graph than Octopi does, and the printer that is doing this is run off S3D while all my other ones are on Octopi. The Octopi machines all seem to hold the temp within one degree, but this one machine has a consistent 8 degree swing. It is not a random swing either, it is a consistent sine wave just like you would normally see with a correctly functioning thermistor. It swings a little higher on the upside. I can't think of what could slow down the response time like this. Bad thermistor contact with the block maybe?

[ednisley]

what could slow down the response time like this

Have you done a PID tuneup (M30x - M500) on that extruder? If not, the P or I terms are probably a bit too high, for whatever reason, and you're watching the control loop oscillate around the setpoint.

If the oscillation is new news, then definitely look for a loose heater or thermistor!

[pyronaught]

I'm not sure if it just started happening or not, I only just started noticing it. The heater block has become loose around the barrel that runs through it though (this is a v4 nozzle). I don't know what normally locks that block in place, but it now spins freely.

[ednisley]

The heater block has become loose around the barrel

Ding! Mechanical problem!

There's the problem: the block isn't firmly connected to the barrel, so it can't transfer heat through the steel cylinder to the filament. Without that "heat sink", the PID loop settings don't match reality: the heater increases the (loose) block temperature too rapidly for the P term to pull it back down and you're watching the I term oscillate.

As soon as the block cools below the lower hysteresis limit, the heater rams it back up again, so it'll always be off by more on the high side.

I think the brass (steel, in your case?) nozzle presses the block onto the barrel; perhaps the nozzle has turned loose by a fraction of a smidge.

[pyronaught]

The heater block has become loose around the barrel

Ding! Mechanical problem!

There's the problem: the block isn't firmly connected to the barrel, so it can't transfer heat through the steel cylinder to the filament. Without that "heat sink", the PID loop settings don't match reality: the heater increases the (loose) block temperature too rapidly for the P term to pull it back down and you're watching the I term oscillate.

As soon as the block cools below the lower hysteresis limit, the heater rams it back up again, so it'll always be off by more on the high side.

I think the brass (steel, in your case?) nozzle presses the block onto the barrel; perhaps the nozzle has turned loose by a fraction of a smidge.

Ah, so the PID settings in the software are based on a specific rate of heat transfer, and that changed when the block came loose. I figured since the thermocouple and the heating element were embeded in the same block that the loose connection wouldn't matter.

Not really sure how to fix it though, if it is a taper fit I could try pressing it back on in an arbor press.

[ednisley]

if it is a taper fit

I'm not a'gonna take mine apart to find out, but since yours is already loose, remove the nozzle (*) and scrutinate the situation: it surely won't heal itself. [grin]

Because the nozzle tightens on the block, methinks a good bit of the heat goes from the block directly into the nozzle, thence to the plastic inside. Most likely, there's a shoulder at the top, with the nozzle pressing the block firmly against the shoulder, with more heat flowing through that joint.

The nozzle is (seems to be?) the only thing holding the block on the barrel, so, taper or shoulder, that's the only knob you have to fix things.

(*) And, yeah, before that, remove the whole hot end. You know that already, but, just in case somebody else reads this without context, wrenching the nozzle with the hot end in place is a great way to break the barrel at that skinny section just under the motor mount.

[pyronaught]

Once this problem happens, you actually have no choice but to leave the hot end attached in order to hold the end while spinning the nozzle off. You can't put a wrench on the block the way you normally would because the barrel will just spin with the nozzle. You just have to be very careful about it and not tighten the nozzle very tightly so that you can get it off with minimal effort next time around.