Trying to calibrate extruder setps

[Linebyline]

Ok so i am trying to calibrate my extruder after installing Marlin RC8.
i removed the hotend from the stepper so as not to waist filament and extruded 100mm it only moved about 35mm so i did the math and came up with 227.4. so i updated the steps and extruded 100mm and it moved 100mm but now with the hot end in place it cant move the filament quickly and it skip.i was not sure about the new steps but it did move 100mm when asked to move that amount. so what am i doing wrong

[Walter]

If it moves 100mm without resistance your steps per mm are ok.
I used the word resistance.
As soon as your stepper has some resistance (the nozzle), he has to use more force and cannot deliver that.
Force is created using current.
I would check the reference voltage on the extruder stepper driver and increase that until the stepper stops skipping steps.

[Ron]

It may be a bit of a long shot, but you might want to also check the actual temperature of the hot end tip as opposed to the reported temperature.

If the extruder feed rate is slower then the filament will have more time to liquify as it passes through.
Once corrected for motion it may not be able to heat the filament fast enough or may have too much resistance because the filament will not be liquid enough.
If your step stick voltage is set correctly and there is nothing dragging on the filament and there are no nozzle obstructions, and it still slips/skips, that could be the problem.

I know mine was running much colder than was being reported until I discovered the problem and corrected it by choosing a different thermistor curve.
It was type1 by default, now I have it set to 5 (around line 195 in Configuration.h).
It would have been nice if the firmware had been pre-installed and configured as claimed in the build manual, but it wasn't.
Particularly since you just installed firmware, don't assume it is correct.

[csorrows]

Typically cheap 100k thermistors as provided in the kit use type 1. The 100k thermistor provided with the E3D kit uses type 5 however.

The default setting should be correct for the thermistor provided in the kit.

[Linebyline]

Ok so after some FRUSTRATING moments. i found that it was the stepper missing steps. an increase of stepper voltage fixed the problem. thank you all for the advise

[Ron]

Actually, I had to modify the thermistor type even using the original extruder, as-shipped.
As I recall it would indicate 245C but the heater block was in fact running around 215-220C.
That much of a discrepancy was easily enough to cause all sorts of printing issues.
Perhaps there are variations in the thermistors that are being shipped from FT, I don't know.
Then when I later changed over to the E3D it was coincidentally correct.
Go figure.

The bed sensor type or curves are also incorrect in the firmware, running around 10C too cold.
I corrected this on mine by adding an offset to the measured temperature because I couldn't find a curve that seemed to match.

Glad to hear that the stepper set point fixed the problem.

[csorrows]

If measured on the bed surface it will be lower due to the location of the thermister as it is closer to the heat source.

As the supplied thermisters are cheap Chinese units, I can believe yours were reading wrong.

If you want precise measurements, doing exactly as you have is a great solution.

[Ron]

Yes of course, the losses through the bed surface must be taken into account.
Particularly so because I am also running a mirror tile on top of it (as many users do).
Now it is running perhaps 1-2 degrees warmer at the build surface than is being reported.

BTW, for whatever it may be worth, my extruder calibration was performed over a longer distance (150mm) and came out to 237.8 steps per mm.
The longer the baseline for the measurement, the more accurate the result is likely to be.
I measured and adjusted it several times. It is what it is.
Likely there are production variations in the depth of the knurling on the drive "gear".
So every extruder will be a little different.
But for sure it isn't the value that comes as default with the Marlin firmware.

[csorrows]

You should also fine-tune it afterward using a single wall print. For a 0.4mm nozzle it should produce 0.48mm walls.

[Ron]

Thanks, I'll have to try that.
Trying to get a round thing to fill in a rectangular space.
Some of the support prints are single pass widths so that should be quite easy to measure.
My prints look pretty good with this setting though, and it is within around 0.1mm of the 150mm commanded length.
I presume that the adjustment could be made as an extrusion feed rate (percentage) or as a tweak to the apparent filament diameter?

[csorrows]

Usually fine tuning is done with feed rate in the slicer.

[Walter]

I typically run a cube without top in vase mode 30mm high and measure the walls on the top 10mm and divide the expected value value by the the average. This percentage I feed into the flow rate in the tuning menu. I do not touch the slicer params for that.
This is something I run for a new spool of filament. I write this factor down on the spool label.
example : 072 / ((0,71 + 0,70 + 0,72 +0,71) /4 ) *100 = 101,4% (rounded to 101%)

Of course the cube is sliced with extrusion factors in the slicer at 100%

[csorrows]

Either in the firmware or slicer works, they have the same result. I do not use EEPROM over-ride as I like my settings in the firmware to be used. Thanks for the helpful math.

[Hammerhead]

Chris, I'm at this point too.

Per my FB post > and other post here, I just measured 100mm and used Repetier's 'manual' tab to advance the filament and only 42mm was sent. According to a video linked in the FB post, the math shapes up like this...

Current steps 105 * 100 / 42 = 250.

That means the. rcp file is WAY off for this printer, and this is an issue that's HUGE in importance yet not addressed in the manual. I'd encourage you to please add a section dealing with extraction tuning.

[csorrows]

The last post in this thread was 8 months ago. Please don't necropost.

More than 3 months is considered too long and should be posted in a more recent/active thread or as a new thread.

[csorrows]