Spikes in RA with PPEC algorithm

[lmcl...@umich.edu]

Hello,

I have an iOptron CEM120 (non-EC) and a Meade 12" LX850 OTA with a focal length of 2435mm.  I'm using an OAG with an ASI174MM Mini guide camera.  I'm having a problem with random RA spikes sometimes as much as 8 arc-sec.  I know the seeing/transparency wasn't the best last night but there was no wind.  And previous nights even with much better seeing the spikes still happened.  Last night's logs are here:
https://openphdguiding.org/logs/dl/PHD2_logs_GZJK.zip

These spikes don't seem to be any relationship to the worm period which is 240 seconds.  I have adjusting the gear mesh several times and it has removed any other spikes and made the peak-to-peak well within specs.  As you can see in the log there is a quite long Guiding Assistant run with no spikes.

I'm using the PPEC algorithm for RA and have been getting random spikes in RA for awhile now and I never have been able to track them down.  They do not occur at regular intervals but most of the time when they do occur, the star mass does take a pretty large drop and then the spike happens.
I have tried using the Star Mass Detection tolerance and even set it all the way down to 10.0.  And even though I will get "Star Lost" a couple of times, right after the large RA spike still occurs.

At first I had the guide camera set to Bin1.  This made the pixel scale 0.49 arc-sec and I thought perhaps that was too small for the algorithm to function well.  So in the log you can see I changed to Bin2 later in the night to bring the pixel scale to almost 1.0 arc-sec.  Unfortunately, making this change did not solve the spikes.

The only thing I was able to do that sometimes got rid of them is to set the MnMo for the RA to something very large like 1.50.  After doing that, I did not get any spikes for over 20 minutes.   But of course, the guiding is not very tight.  Later in the morning I did still get spikes but one was only 4 arc-sec and not the larger 8 arc-sec.

So if you could please look over my logs and let me know what's causing these spikes and what I can do to get rid of them.

Thank you very much,

Lamar

[Bruce Waddington]

To begin, these aren't problems caused by guiding, they are coming from somewhere else.  In this situation, you shouldn't be using PPEC for the RA guide algorithm in any case because you don't want these huge, non-periodic excursions incorporated into the model.  You should revert to hysteresis for RA until you can get this sorted out.  There is a fair bit of periodic error at various harmonic frequencies - 60s, 80s, 120s, and 240s:

But I doubt this is really the problem.  I notice that the big RA excursions are always acting to retard the tracking - meaning that PHD2 has to issue a sizable west correction to recover.  This tells me that something is briefly interrupting the smooth tracking of the mount to the west.  Crud in the RA gear train might do this but it can also be caused by cables that are dragging across a stationary surface.  Especially if the cables have cable ties or are routed through some kind of ribbed tubing, they can "catch and release" as they drag across a stationary surface.  Another possibility is that the mount is balanced west-heavy or can otherwise lose engagement with the drive gear for very short periods of time.  And even with an OAG, you have to be sure that things like extension tubes can't shift around.  Of course there could be many other causes of the problem, the guiding can only tell so much. 

Sorry I can't be of more help,

Bruce