Large RA spikes and Oscillations with CEM70 iOptron CEM70 · D. Jung · ... · 14 · 672 · 7

On mount PEC does not work well with guiding on the CEM70. You either use PEC or guiding, not both at the same time. 
if you need PEC. I suggest you try Predictive PEC in PHD2. I used it with excellent result (under 0.4 arcsec total RMS reliably with good PA).

Minh Lết:
The manufacturer's declared PE iss +-7arcsec so yours seems to be in line with the declared PE.

I think it is +-3.5 not +-7.

From the manual (https://www.ioptron.com/v/Manuals/C70_CEM70_Manual.pdf):
The CEM70's performance is demonstrated by its low periodic tracking error: <±3.5 arc seconds for CEM70 and CEM70G, and <0.3 arcsec RMS for CEM70EC. Along with the stability aspect, the CEM70 features an advanced cable management system consisting of more ports in more locations preventing tangle ups and reducing the chance of fractured cables.

The manufacturer's declared PE iss +-7arcsec so yours seems to be in line with the declared PE. 

D. Jung:

Minh Lết:
The manufacturer's declared PE iss +-7arcsec so yours seems to be in line with the declared PE.

It helps for me, while PPEC in PHD gives terrible results. But even with PEC disabled, the problem is there.

I stand corrected. Thought it was 7 arcsec total. However, as of my understanding, the manufacturer only claim a certain part of the drive train with the declared PE (most likely the worm gear) which they measure with an encoder. The Mount's PE comes from other part of the drive train also. So sending this back MIGHT get you a better PE mount orr it wont. I got something similar to yours (about +-8 arcsec) but mine guide exceptionally well.
Looking at your non PEC PE, it seems your mount is quite good (no steep valley-peak that is hard to guide out). The problem seems coming from the way the mount issuing the guide command (the RA seems to be over corrected, which will be the case if PEC enabled and PHD2 trying to guide at the same time) . Could you please screenshot your calibration graph, see if anything is off.
Try turn off the PEC, guide it with hysteresis and screenshot the result along with the calibration graph here.
Also, Before sending it back, try adjusting the gear mesh and try again (too tight gear mesh can gives all kind of weird guiding issue).

Are you actually calibrating below 30deg of altitude? If you do, don't!. Chose a star mid-altitude, say 10deg above the equator and no more than 2 hours from transit. BTW, the calibration graph doesn't look good.

This might be something irrelevant here but out of curiosity what was the temp’s when you were calibrating? One of my friends has the 120 and he has stated that these mounts don’t like the severe cold. Again don’t know if thats what is going on here but just thought I’d through it out there.

Also, I have found the best (the standard for calibrating a mount) is to be as close to 0º Dec as possible and within an hour or so of the meridian. 
@Minh Lết had mentioned to try the “Predictive Pec” algorithum in PHD2. FYI, I’ve had pretty good luck with that setting as well.

Dale

D. Jung:
When i first tried the mount, it was around 10c. Maybe a coincidence, but when it got colder, around 0c, things started getting out of hand. But that's still not particularly cold.
​​​​I had to tighten several screws on the tripod when it got colder, but seems a bit far fetched that that would deteriorate the mount performance like that.

When i do ppec, the graph jumps around between -2 and +2 arc secs.

I’m not saying thats the issue here, I was just throwing it out there, but I agree even 0c isn’t that cold. When its in the -10-15º F range is when I start to hear things get a little slow moving but still even with my AP1200 I don’t have issues, I’ve seen this with several Celestron mounts personally but again it has to be very cold.

I just know that my friend has a lot of similar problems with this 120 and the customer support wasn’t very good when he started tring to get things fixed. Its been a while now since he’s purchased it so if your mount is under warranty I’d defiantly recommend sending it back for service.

Dale

D. Jung:
The CEM70's performance is demonstrated by its low periodic tracking error: <±3.5 arc seconds for CEM70 and CEM70G, and <0.3 arcsec RMS for CEM70EC. Along with the stability aspect, the CEM70 features an advanced cable management system consisting of more ports in more locations preventing tangle ups and reducing the chance of fractured cables

Ioptron is a little deceptive in how they state their PE specification (IMO).  The quality manufacturers (like Software Bisque, Astro-Physics, etc) generally specify their PE peak to peak.  Ioptron knows that and know that at first glance most people are going to read their spec as peak to peak.  In reality, their specification at +/-3.5" IS 7" peak to peak.

PHD2 cannot, nor can a PEC implementation, correct for non-curve/PE.  What that means is that as long as the guide error/PE is a true S curve no problems.  If it is a sudden shift/spike like hitting a speed bump, there is little that a PEC or PHD2 can do.  PHD2 will work to get the guide star back in on target but that does not mean you won't wind up with a bad frame(s) every time that speed bump comes around.

Ioptron are relatively good mounts for their class (mass produced Chinese manufacture).  I had a GEM45 for a while - it worked well until it didn't.  I sent it back after VERY slow response and delays from Ioptron Customer support; they found the lower RA bearing mount had come loose, repaired it and sent it back.  I just couldn't bring myself to trust it again, so sold it a short time later.  QA/QC for all the mounts coming from China (Skywatcher, Ioptron, etc) is something of a lottery.

Ioptron has some serious issues around PEC - their implementation is rudimentary at best even as compared to their similar class competitor SkyWatcher.  I'm assuming they may have made some improvements in the two years since I had my GEM45.  Back then their PE/PEC routine would only capture one worm revolution and could only be done with their software; tools like PEMPRO that do a more complete/complex curve fitting could not be used.  Even SkyWatchers EQMOD PE/PEC routine was more capable than Ioptron by a significant margin.

Which raises the other issue - with Ioptron PEC implemented, it seems to be a 50/50 possibility that the mount won't guide.  In those instances, Ioptron's PEC routine fights any guidance inputs with some pretty bad results.  I've always been suspicious that this contributed to the woes Ioptron had with their half-assed encoder implementation, although that was bad enough in its own right.

Other than what Guide Assistant can tell you about the unguided performance/PE over 2-3 worm cycles, all the guide performance graphs/calibration aren't going to help you with Ioptron customer support.  For Guide Assistant to be of best help, your PA has to be as close and exact as you can get, so that what you see in Guide Assistant is illustrative of the tracking/PE performance and errors from PA are minimized. 

Follow the directions from PHD2 on establishing the baseline.  Create a new profile in PHD2 - you need to have a 'fresh' start and the only way to do that is create a new profile.  The AI's in PHD2 are complex and always 'learning'.  As a consequence you really can't reset all the settings to get back to where you started at, hence the need for a new profile if things get way out of kilter.

You need to pick a star as close to 0dg DEC as possible, and within 10dg or less west of the meridian (minimizes the visual affect of atmosphere and prevents a meridian flip in the middle of the process).  Insure that all your settings in PHD2 or correct (focal length of the guidescope, camera pixel size, etc).  If you have that option and ability, I do recommend an IR block/bandpass filter in you guide scope - this will minimize some of the movement of the stars due to atmospheric conditions/turbulence.

Run a calibration so that PHD2 'sees' how your mount responds to inputs, backlash, etc.  The part that you need for discussions with Ioptron Customer support is the Guide Assistant run that you follow with.  When troubleshooting, I typically have gone 2-4 worm revolutions, depending on the tracking/PE error.  You don't want to continue to the point that PHD2 loses the guide/reference star.  You will want to save that log/graph as that will be what you have to send to Ioptron Support if it demonstrates out of spec PE and/or movements/spikes that cannot be corrected.  For example, if you have a nice sinoid curve going but a sudden 3" spike that jumps the PE even if it is still within spec, there is a problem as no PEC nor PHD2 can correct for that.

PHD2 Support is great about helping providing an in-depth analysis of their log data from guiding sessions, but you need that starting baseline to help them out.  The unguided Guide Assitant plus an accurate polar alignment will tell you what the mount/OTA combination is doing with no external influences of than the rig itself.  The calibration and guide data, tell you how PHD2 and that baseline of your mount are interacting.

FYI - PHD2 support will tell you that if you have the option for an OAG, use it.  There will always be issues with subtle movement, differential flexure, etc with an external guide scope.  Just remember you need different profiles in PHD2 for one vs the other.

One additional piece of info I learned about the hard way.  I experienced a sudden onset of poor performance with my current mount (SB MyT) from one night to the next.  I made several changes, checks and was getting conflicting/inconsistent data.  Then I had a sudden epiphany when I remembered a discussion on the MyT forum about the internally routed user cables.  There is supposed to be a loop left inside so that the cables don't drag or put tension on either axis so that they can move freely without hindering the mount.  When I checked, my cables were pulled tight even though I was certain I had left a loop.  So I fed the cables back in eliminate the tension but I marked them where they entered the mount.  What I discovered is that gravity/mount movement was slowly working the cables out the cable exit at the rear of the mount.  I resolved this by used a ty-wrap to secure/create a strain relief point against the mount/pier to prevent the cable from feeding out.  I haven't had any issues since doing that.

Clayton