Printer Model: Optimus P1 with LGX & Mosquito Magnum Plus
MCU / Printerboard: Octopus Pro
Host / SBC: Raspberry Pi 4
klippy.log: N/A
Describe your issue:
Is there any way of configuring dual heater cartridges in a smart way on one hotend (one input, one output) rather than configuring dual extruders or multi_pin? My clients current hotend (Mosquito Magnum Plus) has 2 heating carts and 3 thermistors, can i set up dual pid loops easily? Note that running the carts in parallel is not viable since they are 50W each.
Pretty much any board should be able to handle a 100W load easily. If unsure, you can use a Bed Heater connector instead, those are made for handling 200W+.
Arakon is correct, most cheap off the shelf power MOSFETs can handle something ridiculous like 85A at 30V if properly heatsinked or has good thermal bonding to the PCB.
If you can look at the board you can figure out the power dissipation.
Conduction loss is usually P = I^2 * Rds(on)
and Rds(on) on Power MOSFETs is usually milliohms.
So 16A^2 * 50 mOhm = ~13 watts power disssipation and 50 mOhm in a low voltage MOSFET would be huge and that’s assuming 200 watts at 12V
If it’s a 24V system it would be 8A^2 * 50 mOHm = ~3.5 watts dissipated in the MOSFET
Again, This is assuming 200 WATTS. So your IRL values should be at minimum half that, probably a lot less because I picked a huge Rds(on) value.
There is some switching losses too, but that’s why I chose high values for a safety margin.
Look into danger klipper for dual loop PID, I think there’s a branch of that for it. I’d imagine the 3rd thermistor is in the heatbreak but I know nothing about slice’s hotends
What if I want one PID loop to give an output the opposite of the other and vary them in time to sequentially heat opposite sides of an extruder to minimize power usage but heat both sides synchronized to the thermal conductivity of the metal in the extruder?
Which means we need a new config option under extruder to define the watts per meter-kelvin of the users extruder and it’s shape along with the heaters shape (a cylindrical object will heat differently than a cube/cuboid etc.) and the Klipper code needs to implement Fourier’s law of temperature conduction.
This obviously needs to be implemented in mainline Klipper ASAP. We’ve been missing out on some major functionality.