Sorry for the double post in such rapid succession. I’ve got it rigged up finally. The graph on the eeepc is “fuzzy”, and I think what I’m seeing is the microstepping from the stepper motor going through its phases.
We also have a temperature fluctuation with this particular hot end, and those are the large up/down pressure waves that you see. No retraction on this particular print; I’ll get to fiddle with it more on Monday with a temperature stable hot end. I think I need to average out the readings in order to get rid of the microstepping noise “fuzziness” – and also find a way to mount and insulate the circuitry from electrical noise.
The total pressure is important but the slope of the recovery after each step gives you important flow information also that you may not want to just filter out.
@Wylie_Hilliard , the noise is at a high enough frequency that I wouldn’t just be filtering it out. What you see on the screen is happening at 10 refreshes per second. First attempt at filtering will be an average of 3 readings, dropping down to approx 3 refreshes per second. This should allow me better information gathering because things like pressure recovery won’t be hidden in the gigantic amount of microstepping noise.
Also, the info on screen is only ~1000 data points wide. Averaging should allow me a wider time slice to look at.
Depending on the sensor there might be poor filtering. Also hooking it up to your pc might cause some ground loop issues. First try disconnecting the netbook from the wall adapter, it might clear up some noise.
You can do some calculations of maximum steps per second, then calculate a filter capacitor to put on the sensor.
Might also want to isolate the extruder from the machine using rubber washers.
@Stephanie_A , yeah - there’s so many sources of noise here to filter out. Even the head moving back/forth and gravity acting on the bowden tube (we’ve got a long bowden tube!) has an effect on this thing.
I’d do the rubber washers, but I’d be afraid of dampening the load cell too much.
Doing a little more reading, and it’s very important that the excitation voltage be a nice stable reference. I’m not sure what voltage it’s being run at, but there are additional circuits you can add to reduce the noise and eliminate the drift.
