@Ryan_Carlyle not as yet but will in a bit…
As a note, this was happening on an earlier Edge version that it was running, unfortunately I’m not sure which, must be at least 5 months ago and was updated to check that it wasn’t firmware causing the issue, same results.
In the Marlin 2.0 development, the devs faced an issue with the Re-Arm/LPC176X chipset.
They had to use software filters to stabilise the temperature.
Maybe the same issue with Smoothieware?
@Arthur_Wolf
Another alternative is to modify the hardware as to filter the temp readings.
See details here:
@Panayiotis_Savva The LPC176X ADC is well-known these days to have utter garbage input noise, and the Smoothie guys already put a ton of filtering on the thermistor reads to deal with that.
Does soothieware support autotemp and is it being used in your G-code?
It looks like “M109 Smintemp Bmaxtemp Ffactor” and it adjusts the temperature upward as the speed increases.
Id try a different ramps. Sounds like, as the drivers demand power its bleeding back into the board. Then going the path of least resistance usually heaters and fans. Id almost bet you hear a noticable difference in fan speed also.
@Ryan_Carlyle the difference in environments and filaments is why you have the PID next to the feed forwards. We have feed forward configurations on the target temperature, the fan speed and the extrusion amount (ignoring retractions). But we only actively use the temperature setpoint one. The rest we use the good old PID controller to stabilize and correct for the unexpected.
I was involved in trying to fix an Aon printer. The hot end would read 15° off when the bed was on, and the reading error varied like this chart shows. The problem was a lack of electrical isolation of the sensor and the frame. Replacing the sensor fixed the problem.
@Jeff_DeMaagd - it’s the same with the bed on and off. The thermistor on the hot end has been changed, I’ll take a look at though.
Maybe the motors are connected to the hotend? I dont know why it should but it could be everything…
@Ryan_Carlyle one way to reduce the filament feed rate contribution on temperature that I found was to increase the heat lost though a much higher cooling fan flow rate.
I have seen on some printers that ground current changes and/or power supply rail sagging caused by motors and hot ends being turned on/off can affect readings.