Hi Carsten,
I'm telling you my experience with LFEP:
I had an
old EQ6 with
EQ6 MCU update by Rajiva,
old motors with
Conrad 100:1 reduction ratio.
The only annoyance was the low GoTo speed (40X), because of the stock motors...
So I bought a LFEP; everything
apparently fine except for GoTo speed, still low because of the motors.
So I bought SkyScan motors and found nice Goto but bad tracking:
http://forum.lfep.de/index.php?topic=36.0By assuming the issue related to SkyScan gears I reverted back to old motors&Conrad 100:1 reduction.
But 40x GoTo speed was really annoying; so I tried to figure out what was wrong...
I started to think about LFEP issues...
http://forum.lfep.de/index.php?topic=297.0Then I bought a
Bellincioni mount powered by FS2:
300 teeth on RA worm wheel (in place of 180 on the EQ6);
4:1 reduction ratio by belt/pulleys (so no gears related issues);
Sanyo Denki 103 H548 04500 (nice stepper motor with 1.8° step angle, max bipolar current of 0.9A/phase).
A nice setup.Nothing wrong
expected by controlling it by LFEP in place of FS2.
So I removed the FS2 and tested the mount by LFEP. Tracking was not good.

I had to revert back to FS2.
As I spent a lot of time to fix the ASCOM driver, to implement some nice features, and to overcome some known LFEP bugs by ASCOM itself, I still meant to make use of LFEP in place of FS2.
So a few days ago I made another test on LFEP to figure out definitely what is wrong.
I connected a Sanyo Denki 103 H548 04500 (I own a couple of spare motors bought for EQ6 that I never used) with no load to LFEP.
I observed accurately the pinion rotation.
Someone can argue stepper rotation is not good with no load (the detent torque is expected to affect microstep usefulness). But what I found is not related to speed variability on each full step: I found increasing/decreasing speed (exactly on each full pinion rotation, 50 intervals at lower speed alternating with 50 ones at higher speed). Step angle is 1.8° (i.e. 200 steps).
So no drift, no GoTo issue, good pointing, ... But bad tracking.
Increasing tracking current (e.g. setting as tracking current the current you would use only for GoTo)
seems to reduce the issue.
And this made me thinking about the dead zone issue that affects stepper motors at low currents.
But... I think (I may be wrong) dead zone issue should manifest on each step. And I think (I may be wrong) the same would occur with resistance mismatch on the couple of sense resistors used for each motor.
So the issue I'm referring to seems related to LFEP firmware.
Using an higher overall reduction ratio will reduce the issue but won't solve it.
The amplitude of the error on the pinion side will be the same; the frequency will be greater on the "camera" side; the amplitude of the error in arcsec will be lower; GoTo speed will be affected (you'll be obliged to make motor rotating at higher rpm at sidereal speed - i.e. you'll find motor stall at higher GoTo speeds).
I reverted back to FS2.
Clear Skies!
Armando Beneduce