moshc the backlash compensation seems to be getting applied on every focuser move and not only when reversing like expected.
I just tested a 5V EAF with ASIAIR, and it appears to work properly.
You may remember my post on measuring the 5V EAF when it was first released:
https://bbs.astronomy-imaging-camera.com/d/12562-5v-eaf-linearity-and-backlash
I just lashed up the same angle sensor setup with a (different) 5V EAF that I have around; this time, driven by ASIAIR (v2.1.1).
For this first test, I have set the coarse EAF steps to 100 units, and Backlash to 0. This is the sequence in one direction (EAF steps and angle):
10000 0º
10100 6.3º
10200 12.6º
10300 19.0º
10400 25.3º
10500 31.6º
10600 37.9º
10700 44.1º
I then reversed direction (remember, with backlash set to 0)
10600 39.4º
10500 33.1º
10400 26.8º
10300 20.6º
10200 14.3º
10100 8.0º
10000 1.7º
So, it appears that with this EAF, there is about 1.7º worth of backlash.
Keeping the coarse movement to 100 EAF units per step, but this time settting Backlash to 25 EAF units, I see the following after initially clearing backlash:
(Note: from the above link, my original measuremments of a different EAF produce 32 EAF units of backlash.)
Forward:
10000 0º
10100 6.4º
10200 12.6º
10300 19.1º
10400 25.3º
10500 31.6º
10600 37.9º
10700 44.1º
and coarse stepping in reverse:
10600 37.9º
10500 31.5º
10400 25.2º
10300 19.0º
10200 12.6º
10100 6.4º
10000 0.1º
Notice that the asymtotic backlash has been reduced to 0.1º. Notice too that the steps (both forward and reverse) produce pretty much the same constant 6.3º to 6.4º per step. The backlash value (about 1.6º), unlike what you saw, did not incrementally change the ∆ angles for each step; just the initial reversal step.
To double check, instead of coarse adjustments, I tried fine adjustments, of 10 EAF units per step, and the same 25 step backlash (i.e., backlash much larger than the step size, and should be easily noticeable if there is a bug):
Forward:
10000 0º
10010 0.6º
10020 1.2º
10030 1.8º
10040 2.4º
10050 3.0º
10060 3.7º
10070 4.4º
Backwards:
10060 3.8º
10050 3.2º
10040 2.6º
10030 2.0º
10020 1.4º
10010 0.8º
10000 0.1º
Again, there is a residual 0.1º of backlash (I can probably tune it out by using a backlash value of 28 instead of 25).
So, the results appear to show that the backlash is working normally (I haven't had any problem using the ASIAIR with about 8 different 5V EAFs that I have on my various OTAs).
If your EAF does not behave properly, I would sent it back to the dealer for an exchange. The ASIAIR is not the problem in this case.
Chen
