First Light CEM120 Mount

[mark matzner]

Hi Group,

This is the first light from the iOptron CEM120 mount supporting my existing 12LX200 scope and its peripherals.

1.  Guidelog 2018-05-04_205240:

This was a mish-mash of Calibration while doing Drift Align.  The initial physical alignment of the mount was quite a bit off and took a lot of adjustment

a)  The early non orthogonal (about 75 degrees inclusive) calibration graphs were prior to final polar alignment.

b)  The last calibration after PA (Log Section 22) showed an included angle of about 100 degrees.

c)  As has been my experience previously, with both the CGE Pro and now the iOptron CEM120, the calibration results include cautions.  No clean calibration.  I've have assumed this was an electro-mechanical issue, but might it be an optics issue?  Or just something I'm doing wrong?

2.  Guidelog 2018-05-04_215613:

This was about 1 1/2 hr of imaging at 10 minute subs, all on Log Section 1.  Log Section 2 was post flip and PHD2 locked on a hot pixel, due to no stars from fog and dew.  

Guiding was somewhat disappointing since this was a brand new mount which is supposed to be the "Cat's Meow".

Some potential positives-----

a)  No PEC training was done prior to this session, so that might help in the future.

b)  Seeing was crummy with Star mass/SNR deteriorating as we approached flip.  Flip was successful.

c) Scatter looks fairly general with not too much preference between RA and Dec.

Would appreciate just a little nudge in the right direction.

Thanks,

Mark

https://openphdguiding.org/logs/dl/PHD2_logs_mFT7.zip

Equipment (payload weight 50 lb, not including 44 lb c'wt):

iOptron CEM120 (non-encoder) mount

12LX200 Classic Scope, Starizona SCT Corrector

Starlight Feathertouch Focuser, Starizona Microtouch Wired Focuser

Mitsuboshi OAG5, SX Lodestar Guide Cam

Atik EFW2 Filter Wheel, Atik 383L+ mono CCD

Exploradome 8'

Software:

MS Win 10 w/ recent updates

ASCOM Platform v6.3

SGPro v3.0.2.81

PHD2 v2.6.5

ATIK v4.2.0.7 ASCOM Driver

iOptron Commander 6.0.1.2 ASCOM Driver

LesvedomeNet v6.0.1.12  ASCOM Driver

[Brian Valente]

>>> Guiding was somewhat disappointing

 

Hi mark

 

Anything specific regarding your disappointment?

 

Overall the guiding looked pretty good on 215613 first run. ½ pix accuracy, though I can see a few places for improvement

 

I’d suggest you use dec backlash compensation, it’s taking too long to reverse the DEC direction. That will help your DEC quite a bit

 

You might also try Predictive PEC for your RA algorithm

 

Those should help bring it down a bit

 

Thanks

 

Brian

 

 

Brian Valente

Brianvalentephotography.com

--
You received this message because you are subscribed to the Google Groups "Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[bw_msgboard]

Hi Mark.  I think with a new mount, the best way to start is to run the Guiding Assistant for 10-15 minutes to see what the mount is doing on its own.  Then you can know what steps to take to improve performance.  It looks like you might have about 15 arc-sec peak-peak RA periodic error so it would probably be a good idea to tackle that.  If you increase the guide speeds in the mount up closer to 1x sidereal, you’ll improve on the response times for Dec to reverse.  Just guessing, but with a higher guide speed, you’ll probably be able to use PHD2 Dec backlash compensation.  

 

I don’t know quite what to make of the last calibration – the RA rate is actually higher than expected, which is a bit unusual.  It’s not a big deal but it would be worth looking at again to see if it persists.  That assumes, of course, that the reported mount guide speeds of 0.5x sidereal are what’s actually being used.  As a general rule, calibration errors are rarely caused by anything you’re doing.  The most common problem is the presence of Dec backlash which can trigger a variety of alerts.  To be on the safe side, you can always forcibly move the mount north before starting a calibration.

 

Of course, it’s difficult to measure guiding performance under bad sky and seeing conditions.  In the 22:00 guiding session, you should have gotten nice round stars but they were probably somewhat larger than you would like.  So I don’t think there’s anything here that should cause you to be discouraged.  I think you’ll just need to chip away at things under better seeing conditions to see how much more improvement you can get.  You can use the Clear Sky Clock or other astro web sites to preview seeing conditions so you can know what to expect.  

 

Hope this helps,

Bruce

 

---

From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of mark matzner
Sent: Saturday, May 05, 2018 8:59 AM
To: Open PHD Guiding
Subject: [open-phd-guiding] First Light CEM120 Mount

 

Hi Group,

--

[mark matzner]

Brian, Bruce

Pretty decent nudge!  

I assumed - wrongly - that PHD2 was using PPEC and Backlash Comp.  Did not even think to check it.

1.  Both those will be operating tonight.

2.  I will attempt to also do PEC training and playback before tonight's session.

3.  The iOptron handset does display 0.50 sidereal for both RA and Dec, so I will leave those values alone for this trial

4.  Last night was the first clear sky in weeks.  And that was after a nearly full day of rain.  I'm hoping things are more favorable tonight.

I'll follow-up on how things go.

Thanks,

Mark

[bw_msgboard]

Hi Mark.  See below…

 

---

From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of mark matzner

Sent: Saturday, May 05, 2018 12:36 PM
To: Open PHD Guiding

Subject: Re: [open-phd-guiding] First Light CEM120 Mount

Brian, Bruce

 

Pretty decent nudge!  

 

I assumed - wrongly - that PHD2 was using PPEC and Backlash Comp.  Did not even think to check it.

1.  Both those will be operating tonight.

2.  I will attempt to also do PEC training and playback before tonight's session.

3.  The iOptron handset does display 0.50 sidereal for both RA and Dec, so I will leave those values alone for this trial

 

Don’t know why you wouldn’t increase these right away, it will give you a better chance of getting good results from BLC.  It’s not controversial or risky to change these values.

 

4.      Last night was the first clear sky in weeks.  And that was after a nearly full day of rain.  I'm hoping things are more favorable tonight.

 

Yes, seeing is often the worst immediately after a storm system has moved out.  If you run the GA we’ll be able to get an independent measure of tonight’s seeing, and of course that’s the best way to get a good starting point for BLC.

 

I'll follow-up on how things go.

 

Sounds good – hope conditions are good for you.

Bruce

 

Thanks,

 

Mark

[mark matzner]

Brian and Bruce,

here's the latest.  These steps done in the following order:

1.  Set PHD2 algorithm to PPEC.

2.  Set PHD2 algorithm to Backlash Compensation.

3.  Set CEM120 Guider rate from 0.5X sidereal to 0.75X (RA and Dec).

4,  Recalibrated due to guide rate increase.

5.  Ran Guiding Assistant.  Updated parameters per GA's recommendations.

6.  Conducted PEC Training within CEM120 mount.  Included PEC Playback during SGPro NGC4725 capture sequence.

7.  Changed from 2s guide cam exposure to 4s after first 10 minute sub of NGC4725.

A)  The new calibration following steps 1 thru 3 (above) went well.  Can't recall the last time I had a clean calibration.

 (see Log File 2018-05-05_204021).

B)  Guiding Asistant (step 5) looked awfully promising (see attached screen shot).  One of the (many) touted characteristics of the CEM120 mount is there is zero Dec backlash.

 (see log File 2018-05-05-2018_204021 and screen shot "GuideAssistant May05 2018-2).

C)  PEC Training is simple assuming I did it correctly.  I selected a guide star at roughly the meridian and celestial equator and started guiding with PHD2.  Then turned on the CEM120 PEC Training.  The worm cycle is 240s.  During this period the CEM120 mount recorded the guiding and applied it to the RA drive.  Or at least I think it did.  Then this PEC Playback was activated for the NGC4725 imaging session.

 (see Log File 2018-05-05_210038).

D)  I do not believe there were improvements in the PHD2 guiding at the start of tonight's run, so after the first 10 minute sub, I changed the 2s guide cam exposure to 4s.  the longer exposure will tend to mitigate poor seeing,  so I understand.  However, I do not see any improvement.

 (see Log File 2018-05-05_211150).

E)  This is the second time there was an issue with the process not continuing properly after pier flip.  This happened both on May 4 and again on May 5 (last night).  I witnessed pier flip on May 4 and agree there was no guide star available post flip.  Even with better seeing conditions last night, there may be no stars available after flip for this target.

So Calibration and the Guide Assistant looked pretty good (at least to my uneducated eye).  

While the resulting images are OK, the PHD2 guiding  does not show to be particularly good.  Several examples of RA exceeding 4" arc -sec.  

(See Log File 2018-05-05_211150) 

Can I get a recommendation on what to set the "Minimum star HFD (pixels)" value, if this is relevant.   The guide logs from last night and the previous should provide some insight.  I'm pretty certain that a hot pixel was mascherating as a star, post flip, on both nights.

Thanks,

Mark

Sorry about the DropBox on the log files.  Could not get the PHD2 Upload to work today.

https://www.dropbox.com/sh/dbwr2e2u0xdrj57/AADx3Rip8Ngm5TydPQm4Psaba?dl=0

[Brian Valente]

Hi Mark

 

I looked at the dropbox zip file, I don’t see the files below mentioned in that?

 

Also I would personally avoid using PEC on the mount until you have gotten everything else dialed in. Try changing one thing at a time. PEC is notorious for being easy to mess up (I mess it up constantly), so I’d rather not have that extra bit in the mix. I know Bruce suggested otherwise and he’s waaay smarter than me, but I think changing one thing at a time would be easier to diagnose

 

Thanks

 

Brian

 

 

Brian Valente

Brianvalentephotography.com

 

From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of mark matzner
Sent: Sunday, May 6, 2018 12:14 PM
To: Open PHD Guiding <open-phd...@googlegroups.com>
Subject: Re: [open-phd-guiding] First Light CEM120 Mount

 

Brian and Bruce,

--

[mark matzner]

Sorry about that.  DropBox can be confusing, especially on Sunday after I've had several adult beverages while playing at golf.

I think this is the right one.

https://www.dropbox.com/sh/pps4c7kekazcd1u/AADJmUps2IAwYlOc2LUh8K_ua?dl=0

I take your point about PEC.  I never saw any improvement on my CGE Pro when trying it.  

And that will be a shame if it doesn't improve the CEM120.  It's soooo easy to create.

Mark

[Brian Valente]

It’s easy to create, it’s not as easy to create effective and correct PEC ;)

 

 

 

Thanks

 

Brian

 

 

Brian Valente

Brianvalentephotography.com

 

From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of mark matzner
Sent: Sunday, May 6, 2018 1:02 PM
To: Open PHD Guiding <open-phd...@googlegroups.com>
Subject: Re: [open-phd-guiding] First Light CEM120 Mount

 

Sorry about that.  DropBox can be confusing, especially on Sunday after I've had several adult beverages while playing at golf.

--

[Brian Valente]

I think I am eating my own words I just said ;)

 

Log 0038 shows really good improvements – both your RA and DEC show good improvements, including substantial reduction in that nasty PE you are seeing in RA.

 

DEC backlash looks pretty good too.

 

RA is still close to double of DEC so you will get some eccentricity to your stars, but that can be easy to fix.

 

I’d say that’s the most promising log of the lot I’ve looked at so far, just really short

 

 

 

Thanks

 

Brian

 

 

Brian Valente

Brianvalentephotography.com

 

From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of mark matzner
Sent: Sunday, May 6, 2018 1:02 PM
To: Open PHD Guiding <open-phd...@googlegroups.com>
Subject: Re: [open-phd-guiding] First Light CEM120 Mount

 

Sorry about that.  DropBox can be confusing, especially on Sunday after I've had several adult beverages while playing at golf.

--

[bw_msgboard]

Hi Brian.  I’m not clear on why this is such a problem for you.  Is there something about your mount that just doesn’t respond well to PEC?  AFAIK, there are only a few well-known traps:

 

1. Not taking data over a sufficient number of worm cycles
2. Not grinding through the calibration process
3. Moving the mount between the time the data are collected and the PEC is programmed
4. Having PemPro compute a “backwards” correction curve.  I’ve never understood why this happens but there is at least a “reverse” function available
5. Including non-harmonic frequencies in the correction curve – i.e. over-fitting the curve.  This can really be a killer and is counter-intuitive to many users

 

My philosophy is to go ahead and create a PEC curve, then immediately do a before-and-after measurement of the residual PEC.  If you don’t like the result, you can disable PEC and move ahead assuming that you can’t see an immediate way to improve it.  All you’ve lost is a little bit of time and you might get a significant benefit.  But of course this assumes the mount error really is periodic and the mount firmware handles PEC correctly.  My experience has pretty much been the same with both a Meade fork mount and an AP GEM mount – a significant improvement in both cases.

 

Maybe you can shed some light on other issues I don’t know about.

 

Bruce

 

---

[Brian Valente]

Let’s not get too far afield from Mark’s topic, but briefly the Losmandy G11 and G11T (maybe other losmandys? Don’t know) have some periodic errors whose periodicity are more than one worm cycle. There may also be an issue somewhere that introduces drift in the PE correction sync such that it goes in and out of sync and as you can guess create more issues than solutions. I am just guessing here, but that is the short answer to why it’s more of an issue for me than possibly elsewhere.

 

It’s not a huge deal for me, as I am getting somewhere around 0.5” guided, but I feel I have to perform slightly unnatural acts to make that happe

 

 

Thanks

 

Brian

 

 

Brian Valente

Brianvalentephotography.com

 

[bw_msgboard]

Hi Mark – I’ll make some comments on what I see in your logs.  We can start with the good stuff – as you say, the Dec backlash looks good and should be easily handled by PHD2 backlash compensation if you even want to bother with that.  The clean calibration is also a welcome event.  You can probably run with this calibration for an extended time unless you rotate the guide camera or make other significant changes.  

 

The star-selection problem at the end of the last session certainly looks like a clump of hot pixels, a fairly common thing with the LodeStar cameras, especially in warmer weather.  Although we talk about a “hot pixel”, the problem often shows up as a small clump of hot or warm pixels.  That bogus “star” had a HFD of 1.26 pixels compared to the earlier sessions where legitimate guide stars had HFD values of well over 3.  So you could set a min-HFD around 2.0 px and see how that works.  You can even use a text editor on older guide logs and do a search on ‘HFD’ to see what is typical for your system.  

 

This brings us back to the overall guiding performance which, as you say, was a little disappointing.  What I see in both the 5/4 and 5/5 guiding sessions is a repeating, very quick RA excursion at intervals of about 4 minutes.  Here are two views, the first on 5/4, the second on 5/5 (RA only shown in red):

 

 

 

 

You can see the abrupt RA displacements, mostly to the east (down).  Since you didn’t have PEC enabled on the 5/4 session, it seems unlikely this is due to a bad PE correction.  You can verify that, of course, by disabling PEC in the mount the next time you test.  The problem here is that these excursions are very fast and pretty large.  Their size is probably the limiting factor on the RA guiding, and they happen too quickly for guiding to handle very effectively.  

 

This looks to me like a mechanical issue – perhaps a simple one – and I’d suggest getting some help from iOptron.  These graphs might be helpful so long as they understand that this isn’t a guiding issue.

 

Hope this helps,

Bruce

 

---

From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of mark matzner
Sent: Sunday, May 06, 2018 12:14 PM
To: Open PHD Guiding
Subject: Re: [open-phd-guiding] First Light CEM120 Mount

 

Brian and Bruce,

--

[mark matzner]

Bruce, Brian,

If possible, I would like you to look at the logs from last night (May 6).

Guidelog_2018-05-06_201121:

1. "Minimum star HFD" set to a value of 1.5.  I did this by testing the value of the hot pixel that PHD2 had selected after pier flip on previous sessions.  Its value was 1.3.  This was done to help the guide star selection post flip.

2.  Selected 2s for the guide exposure, just because.

3.  Executed PEC Playback on the mount.

This was the same setup as the previous nights.  It started around 8:50 in the evening.  As with previous sessions, this was with the mount at sop West.  

I watched the first sub complete and the second start, and then I went on to other things.  I came back to for the 11:25p flip, and saw that the second image had failed less than one minute prior to completion.  Something caused the RA to do a large excursion.

Besides the above, I also noticed that the mount was warring with PHD in the RA axis.  RA kept wandering north of the abscissa, at least on the graph.  I had changed nothing from the previous sessions.

Guidelog_2018-05-06_233047:

1.  Did a clean restart on computer and mount, around 11:40p.

2.  PEC Playback was turned OFF.

3.  Since this was post pier flip time the mount orientation was sop East upon restart.  PHD ignored the problematic hot pixel from previous sessions and chose a real star, albeit very dim.  Because this star was very close to the FOV edge,  I re-selected another (also very dim), but more centrally located.

4.  I thought the guiding might improve with the mount at sop East due to c'wt imbalance.  Didn't appear to be the case.

5.  Might some of these problems be due to the low guide star SNR?

I will approach iOptron on these issues.  I'm posting to you today to see if last night's sessions offer any further clarity.

Thanks,

Mark

https://openphdguiding.org/logs/dl/PHD2_logs_VhHK.zip

[bw_msgboard]

Hi Mark.  I’m going to be without power most of the day so I’ve just done a quick look.  I see the same RA behavior as before.  The second run, with PEC disabled, is probably the best reference for talking to iOptron.  Look at all the strong, impulse-type moves in RA toward the east (down):

 

 

This shouldn’t have anything to do with star SNR, this looks like a mount tracking issue to me.

 

HTH,

Bruce

 

---

From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of mark matzner
Sent: Monday, May 07, 2018 7:35 AM
To: Open PHD Guiding
Subject: Re: [open-phd-guiding] First Light CEM120 Mount

 

Bruce, Brian,

--

[mark matzner]

Thanks Bruce. I have been working with iOptron and running multiple trials over the last few days. Nothing definitive yet.

one of the sessions last night used ST-4 guiding and guide camera binning of 2x2. First time to try the binning. And I’ve a question about the it.

What is gained and what is lost by binning? Does the guide cam’s FOV change? How so?

Thanks,

Mark

[Bruce Waddington]

Hi Mark. When you bin the camera, the FOV stays the same but the dimensions of the image are downsized by the binning factor. If you're binning 2x2, the camera is combining 4 pixels of data to produce one binned pixel, but all the pixels on the sensor are included. The point of binning - guiding or imaging - is to reduce the percentage contribution of read noise in the image - in other words, to increase the SNR. For short exposures like we use for guiding, this can be especially useful because read noise is normally the biggest overall noise contributor. The price you pay is a loss of resolution because your image scale (arc-sec/px) goes up by the binning factor. If your un-binned image scale is already large (coarse) compared to the image scale on your main camera, you might not want to do that. For the sake of thinking about it, assume that PHD2 can determine a star position at a level of 0.1 px - un-binned pixel, binned pixel, doesn't matter.

I'll give you a couple of examples from my own experience. When I image at 2540mm focal length with an OAG, the un-binned guider image scale is 0.5 arc-sec/px. The image scale on the main camera is 0.6 arc-sec/px. So I typically bin the guide camera and guide at 1.0 arc-sec/px. That lets me use fainter stars for guiding without having to extend the exposure times. When I switch to a separate guide scope and image at 1680mm focal length, the un-binned guider image scale is 3.8 arc-sec/px and the main camera image scale is 0.93 arc-sec/px. So I definitely don't want to bin the guide camera in this scenario because the guiding image scale would jump up to 7.6 arc-sec/px.

One other thing about binning is that you might find that bright stars saturate and won't be good choices as guide stars. If you're going to experiment with binning, you should create separate profiles for binned vs. un-binned operation. Many, many things are different - calibration data, calibration step-sizes, guiding parameters, etc. So having all that kept in separate profiles will makes things go a lot easier. The latest versions of the new-profile-wizard support binning choices so it's easy to do.

Hope this helps,
Bruce

-----Original Message-----
From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of mark matzner

[mark matzner]

At the request of iOptron, I ran a total of 5 subs (5 min. ea.) at the beginning of the session.  The first 4 were guided and the 5th was un-guided.  

The first 4 demonstrate the problem this mount is experiencing when it exhibits the RA excursions on most every sub.  

For now my question is specifically regarding the 5th sub which is unguided.  What are your thoughts on what the substantial wandering in Ra is displaying?  Does this argue that the polar alignment is off?

I have attached a screen shot of the results from May 5th Guiding Assistant run.  It shows the polar alignment error to be 1.9 arc-min, which I thought was pretty good.

Mark

https://openphdguiding.org/logs/dl/PHD2_logs_trRB.zip

[Brian Valente]

The first runs show a roughly 4 min periodic error that’s pretty substantial. PHD handles it as well as it can, but it’s jumping 2.5-3” ish. Bruce probably said that before but it probably can’t be addressed only with PHD.

 

As for the wandering, I look at it like this: the DEC highlights polar mis-align (which looks pretty reasonable) and the RA mostly shows the unguided periodic error, which looks pretty substantial.

 

 

 

Thanks

 

Brian

 

 

Brian Valente

Brianvalentephotography.com

 

From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of mark matzner

Sent: Friday, May 11, 2018 12:00 PM
To: Open PHD Guiding <open-phd...@googlegroups.com>

--

[bw_msgboard]

Hi Mark.  The RA curve shown in the GA results represents a combination of two things: 1) A roughly 9 arc-sec peak to peak periodic error superimposed on 2) a general RA drift of about 1.4 px/min.  Since you had just done a calibration with good results, we can be pretty sure the RA drift has nothing to do with polar alignment.  Polar alignment error principally affects Dec drift.  In my experience, the most likely cause of RA drift like this is sagging in the camera/OAG assembly hanging on the back of the scope.  You were working from the west side of the pier and the direction of RA drift was also west.  That’s consistent with gravity tugging on the gear and “pulling it down” (toward the west).  This isn’t really much of a problem, this sort of slow steady drift is pretty easily corrected by guiding.  In the context of the other problems you’re trying to identify, I wouldn’t worry about it at this point.  OTOH, you might take a quick look and see if something has loosened up in that area.

 

Cheers,

Bruce

 

---

From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of mark matzner

Sent: Friday, May 11, 2018 12:00 PM
To: Open PHD Guiding

--

[mark matzner]

Hi Bruce, Brian et al,original poster here.  

I've been out of country for the last month and am just back.

  

The following is the most recent guiding material sent to iOptron for my CEM120 (not encoded) mount with the Meade 12" LX200 Classic scope at F7.16 (2182 mm FL), along with the Hutech OAG5 with 0.5x reducer and Lodestar guide cam (1192 mm FL).

This mount exhibited the RA excursions right out of the box, so I have been working with iOptron to try to solve.  Given their guidance we have 

1.  Updated to the latest drivers and then reverted to previous.

2.  Tested guided and unguided conditions several times.

3.  Utilized both ST-4 guiding and Pulsed guiding, although iOptron prefers ST-4.

4.  Done a lot of stuff I've since forgotten.

I think it's fair that all guiding tests showed the same (or very similar) behavior as the attached Guidelog from May 12.  This one is a bit cleaner than the previous (probable reason:  guide star, sky conditions).  And in my mind shows clearly the periodic nature of the RA excursions.

These occur every 4 minutes (worm period) within a few seconds each time.

So iOptron's response to the material I've sent can (I think) be condensed down to the following:

"Back to the spikes observed during guiding on your mount, it looks like that the mount is over guiding. So please try to reduce the guiding rate and use less aggressive settings."

"Another possible cause is that the RA worm/gear meshing factory setting is too tight. You may loosen it a little bit by following attached instruction."

Given the setup conditions below, do either (or both) of these sound credible?  And which do you feel is more likely? 

Guiding Summary for May 13 2018

C).  GuideLog_2018-05-12_212805:  ST-4 GUIDING (ON CAMERA)

1.  Guiding and imaging started at 9:33 pm and was terminated around 10:12 pm.  Star mass and SNR stayed steady (i.e. no clouds).

2.  Mount on sop West at approximately 60 deg elevation.

3.  Image subs were of 10 minute duration.  Seven (7) in all.

4.  RA excursions are abrupt and short lived.  They appear to occur every 4 minutes.

SETUP-----

a)  PHD2 camera exposure time was 4s.

b)  PHD2 algorithm Set to Hysteresis

c)   PHD2 Dec algorithm set to RESIST SWITCH with Backlash Compensation Set to OFF.

d)  PHD2 Minimum star HFD Set to 1.5

e)  PHD2 Star Mass Detection Set to DISABLED

f)   PHD2 Camera (Lodestar) binning set to 2x2

g)   CEM120 PEC PLAYBACK Set to OFF

h)   CEM120 Guider rate Set to 0.90x sidereal (RA and Dec)

j)   CEM120 HC & Mount Firmwares reverted to v180103 and v171128

k)   CEM120 mount iOptron Commander reverted to v6.0.0.3

https://www.dropbox.com/sh/fsvceor8vve7mor/AADaXsvob-JFo9nZTK68kA_xa?dl=0

[bw_msgboard]

Hi Mark.  It’s pretty clear to me these spikes aren’t caused by guiding.  So the first option they gave you doesn’t make sense to me and suggests they didn’t look at the data carefully or don’t understand what you’re showing them.  Let’s just zoom in on one of these spikes – they all look pretty much the same:

 

 

The arrow I’ve added points to the time when the RA drive makes its big departure.  There were no guide commands sent at that time.  The guide commands are shown by the small red rectangles, and those only show up after the fact as PHD2 is trying to wrestle the mount back into line. If guiding had triggered the problem, you would see a big rectangle in the “up” direction located where the point of the arrow is.  And of course, the striking repetition that matches the worm period seems like a pretty strong clue that it isn’t a guiding problem., PHD2 didn’t know anything about the worm period.

 

So I would vote for this being a mechanical problem in the RA drive system.  We have seen cases in the past – can’t remember whether they were iOptron -  where the gear mesh was too tight and caused this sort of large, impulse-type errors.  At least with RA, you don’t have to worry about creating too much backlash for the purpose of guiding.  Obviously, you don’t want a huge amount of slack in there or you could start having problems with slewing operations.  And of course, there is the unspoken option that the machining for the RA axis was botched and can’t be corrected by simply adjusting mesh.  Hopefully, that won’t be the case though.

 

It looks like you’ve done a very careful and systematic job of trouble-shooting and documenting what happened – nicely done!  I hope you’re soon rewarded by seeing this problem disappear.

 

Good luck,

Bruce

 

 

 

---

From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of mark matzner

Sent: Friday, June 15, 2018 7:36 AM
To: Open PHD Guiding

--

[Al Moncayo]

Hi Mark,

It seems like I am following in your footsteps. My new CEM120 looks like it behaves very similarly to yours in spite of different hardware attached to it. Looking at the guide log (and FFTs of the same) I see very similar signatures in my mount.

Has iOptron responded to you with something more concrete? I've only invested about 5 nights with mine trying different configurations, algorithms, and settings. So far, I'm not too thrilled about the guiding. Other than that, I like the mount.

I've attached a text file containing what I have sent to iOptron recently.

Al

[mark matzner]

Hi Al,

would you mind sharing some of your Guide Logs? My equipment is imaging at 2182mm and OAG guiding at 1192mm, so your longer FL stuff would be interesting for comparison.

Thanks,
Mark

[Al Moncayo]

Hi Mark,

I've attached a zip file with several logs:

180504 - Atlas EQ-G at 1600 mm FL (OAG)

180507 - Atlas EQ-G at 1600 mm FL (OAG)

180508 - Atlas EQ-G at 1600 mm FL (OAG)

180615 - Atlas EQ-G at 1170 mm FL (OAG); Here I had modified my EQ-G to upgrade it to belt drive, but I hadn't completely refined the RA meshing; I also added my AP67 reducer

180620 - iOptron CEM120 at 1170 mm FL (OAG, with reducer)

180625 - iOptron CEM120 at 240 mm FL (on axis with 60 mm guide scope)

180628 - iOptron CEM120 at 1600 mm FL (OAG, minus reducer)

I'm hoping that I can get the CEM120 to perform better than my EQ-G. Even if it remained at par, I'd be happy assuming that I could put 80 lbs on the CEM120 and still be at par.

Al

[mark matzner]

Thanks Al,

Here's my take.  Sorry in advance for any errors in my thinking.

180628:  Log Section 2 --

a)  The magnitude of the RA plot is less than what I am seeing on my mount.

b)  No 4 minute RA period is evident.

180625:  Log Section 3 --

a)  RA magnitudes may be a bit higher than 18068.

b)  No 4 minute RA period evident

18620:  Log Section 10 --

a)  RA magnitude about the same as 180628.

b)  No 4 minute RA period evident.

I have no experience with the Atlas mount and therefore won't comment. 

And sorry, I don't have anything to offer regarding improving your guiding.  But I suspect the good folks here can help substantially.  I'll be watching closely for suggestions also.

I am encouraged by the absence of the 4 minute RA period on your mount (at least to my eye).  My mount has just gone back to iOptron for repairs.  For now the plan is to replace the RA worm drive and, with that done, I feel better about a positive outcome after seeing your results.

Mark

[bw_msgboard]

Hi Al.  I don’t think you were asking for any help so maybe I shouldn’t offer any. J  But… the unguided performance of your mount looks quite good – here’s the 10+ minute run of the Guiding Assistant:

 

 

The peak-to-peak error is only about 6 arc-sec and the curve looks pretty smooth.  The residual period is still around 240 sec, so PPEC will probably help you out here.  The Dec backlash is also quite small, only 200ms at whatever guide speed you’re using.  I’m not sure why you were fiddling around with uni-directional guiding or trying to throttle the Max-Duration in Dec though.  The latter won’t have any effect on guiding this mount, and with such a small backlash you should have no trouble guiding in both Dec directions.  I’m also not clear on why you don’t want to use an ASCOM connection to the mount – seems like you’re just going to waste time on calibrations every time the scope is slewed in addition to opening yourself up to eventual problems with the guide cable.  

 

But if you’re happy with the results and your main-camera images look good, I guess that’s what matters.

 

Have fun,

Bruce

 

---

From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of Al Moncayo
Sent: Saturday, June 30, 2018 10:10 AM
To: Open PHD Guiding
Subject: [open-phd-guiding] Re: First Light CEM120 Mount

 

Hi Mark,

--

[Al Moncayo]

Hi Bruce,

Thanks for your reply. I did notice that - at least twice - the unguided performance "looked" good with 10-20 minute runs of GA. 5-6" matches what I believe that iOptron claims for this mount's intrinsic tracking.

Regarding ASCOM connection, funny you should mention that. It's not that I was unwilling to connect through ASCOM, I just had not yet gotten around to it. A couple of nights ago, I went ahead and connected through ASCOM directly to the mount. So far so good.

I also switched back to try PPEC/ResistSwitch for RA/DEC. I began to see significantly better trends. Tonight, even though I had clouds, I figured why not see what happens. I was thoroughly impressed that at 1600 mm FL, with my 178 camera (2.4um pixels), PHD could still guide even though I looked out toward the area I was pointing at and saw nothing but cloud cover!

I'm attaching the log here. Please have a peek at it. Also, if you wouldn't mind, comment on the ~30 second period that shows up prominently... In spite of such dismal conditions (and not-so-great calibration), guiding over 30 minutes yielded ~1" rms total.

I'm very encouraged. :)

Mark,

I'll be eager to learn how your iOptron repair experience was.

Al

[bw_msgboard]

Hi Al – see below…

 

---

From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of Al Moncayo

Sent: Monday, July 02, 2018 10:59 PM
To: Open PHD Guiding

Subject: Re: [open-phd-guiding] Re: First Light CEM120 Mount

Hi Bruce,

 

Thanks for your reply. I did notice that - at least twice - the unguided performance "looked" good with 10-20 minute runs of GA. 5-6" matches what I believe that iOptron claims for this mount's intrinsic tracking.

 

Regarding ASCOM connection, funny you should mention that. It's not that I was unwilling to connect through ASCOM, I just had not yet gotten around to it. A couple of nights ago, I went ahead and connected through ASCOM directly to the mount. So far so good.

 

Ok, didn’t mean to nag you, it’s just that we get a lot more info from an ASCOM-connected mount.  Along those lines, here are a couple of suggestions:

1.       It looks like you did your calibration “on-target” at a Dec of -27 degrees.  It’s better to do the calibration closer to the celestial equator and then just re-use it.  Because PHD2 now knows where the scope is pointing, there’s no need to keep re-doing calibrations – just get one with no alerts near the celestial equator and PHD2 will handle the rest.

2.       The guide speed in the mount looks very low, 4.5 arc-sec/sec.  We recommend guiding at a faster speed because it makes the mount more responsive to guide commands.  Try changing the guide speed in the mount to something closer to 1x sidereal (15 arc-/sec) or at least something over 0.5x sidereal (7.5 arc-sec/sec).  When you change the guide speed in the mount, you’ll need to re-do the calibration using a revised calibration step-size.  You can use the ‘Calculate’ button in the Guiding tab of Advanced Dialog to compute a new value – it’s just a linear adjustment, so going from 4.5 arc-sec/sec to 15 arc-sec/sec is an increase of 3.3x, so you would reduce the calibration step-size by about 1/3.  There are instructions in the Help docs.

 

 

I also switched back to try PPEC/ResistSwitch for RA/DEC. I began to see significantly better trends. Tonight, even though I had clouds, I figured why not see what happens. I was thoroughly impressed that at 1600 mm FL, with my 178 camera (2.4um pixels), PHD could still guide even though I looked out toward the area I was pointing at and saw nothing but cloud cover!

 

I'm attaching the log here. Please have a peek at it. Also, if you wouldn't mind, comment on the ~30 second period that shows up prominently... In spite of such dismal conditions (and not-so-great calibration), guiding over 30 minutes yielded ~1" rms total.

 

Yes, there’s definitely a 30-second residual periodic error in RA:

 

 

You could ask iOptron about this if you want to pursue it.  OTOH, you should be getting nice round stars in your main images and your guiding is mostly below 1 arc-sec total RMS, so maybe you don’t want to mess with it now.  Unless you’ve been doing this for a fairly long time, the guiding is unlikely to be the limiting condition in the quality of your images.  

 

I'm very encouraged. :)

 

Yes, things look pretty good.  In your situation, I would probably focus on all the other hard problems associated with capturing good data and then processing it into high-quality final images.

 

Good luck,

Bruce

 

Mark,

 

I'll be eager to learn how your iOptron repair experience was.

 

 

Al

 

On Saturday, June 30, 2018 at 7:13:43 PM UTC-7, Bruce Waddington wrote:

Hi Al.  I don’t think you were asking for any help so maybe I shouldn’t offer any. J  But… the unguided performance of your mount looks quite good – here’s the 10+ minute run of the Guiding Assistant:

 

 

The peak-to-peak error is only about 6 arc-sec and the curve looks pretty smooth.  The residual period is still around 240 sec, so PPEC will probably help you out here.  The Dec backlash is also quite small, only 200ms at whatever guide speed you’re using.  I’m not sure why you were fiddling around with uni-directional guiding or trying to throttle the Max-Duration in Dec though.  The latter won’t have any effect on guiding this mount, and with such a small backlash you should have no trouble guiding in both Dec directions.  I’m also not clear on why you don’t want to use an ASCOM connection to the mount – seems like you’re just going to waste time on calibrations every time the scope is slewed in addition to opening yourself up to eventual problems with the guide cable.  

 

But if you’re happy with the results and your main-camera images look good, I guess that’s what matters.

 

Have fun,

Bruce

--

[Al Moncayo]

Hi Bruce,

I was able to do another round of testing with your thoughtful suggestions. I'm attaching two logs. The calibration was on a night with high humidity (and in my opinion, not-so-great seeing). The following night I re-used the calibration and just started guiding. Seeing was a bit better. Results are getting better.

The reason for my update here is that I've looked through many posts and often do not see closure. At this point I'm pleased and see a promising path to achieve my goal of 0.5" rms total guiding error with the CEM120.

There are a few lingering concerns, like the 78 second peak in RA that I wonder about. Hopefully iOptron has additional info for me. If any other CEM120 users have insight, please share...

:)

[bw_msgboard]

Thanks for reporting back, Al.  You’re right, threads are often left unresolved for one reason or another.  We often can’t tell whether it’s because the problem was resolved or the person just got tired of talking to us. J

 

With regard to your guiding goal, unless you’re imaging at a fine image scale (say less that 0.7 arc-sec/px) on a night of good seeing, there’s not likely to be any meaningful difference in results between 0.5 a-s total RMS and 0.6 a-s (which is what you were getting in the last run).  And at those levels, you’re likely to be seeing-limited in any case.  What you’re getting looks good to me.

 

Regards,

Bruce

--

[Rainer Ehlert]

Hi,

I saw your graph of the periodical error at ~ 77 seconds. From my experience with other mounts and 240 seconds full worm cycle that could be a problem with a Ball Bearing but in order to confirm that we would need to know how many balls are in the bearing. The error I know is a ball bearing with 6 balls and that gives a periodic error of 76 seconds.

regards Rainer

[mark matzner]

Hi Group, original poster back.
iOptron replaced the RA worm drive on my CEM120. While the tech Kevin indicated that adjustments probably would have worked, replacement was quicker.
They returned the repaired mount in short order and the periodic excursions in RA appear mostly gone, or at least, lost in the noise.
The RA curve in the PHD2 log file now looks more like a “random walk”.
As mentioned in my PEC best practices post, we are taking a ponding hear in Texas with the heat, haze and night high clouds. So I may be getting a little bit ahead of myself. But the mount’s performance appears substantially improved.

Assuming better weather in our future, my next goal will be to tame the RA a bit more to better match the amplitude of the DEC.

Mark

[Al Moncayo]

Hi Mark,

Good to hear that your mount is back and working better.

My most recent guiding results have me very optimistic. I'm eager to see updated results from you when you have better weather. Out here in "sunny" California, we've been having waves of hot, very humid weather. To boot, we've now had "June gloom" pretty consistently since May to present. I've had very few decent nights so far. Like you, I'm hoping for good weather in the near future!

Here's a trimmed log of guiding on one of the best recent nights with the CEM120. I did, however, switch to on-axis so my guider image scale is a whopping 5.16"/px. I'll eventually switch back to off-axis...

I'm also curious to know if you see that spike in the FFT at about 77 seconds. I've now been seeing it pretty consistently, with amplitude on par with the main worm's at 240 seconds.

Al

[Rainer Ehlert]

Hi Al,

Nice to hear that there is another CEM 120 mount user. Congrats.

I did take a look at your FFT analysis and looks interesting. Apart from the 240sec peak which is one worm cycle there are 4 little peaks with an Interval of 0.1, 0.2, 0.4 and 0.6 being the 0.6 at 77sec.

Have you contacted iOptron and asked about his ? They will for sure analyse it or tell yu what is involved. But I do not think those should be a problem after seeing your guide graph and scatter result. I would contact them. They are very helpful.

I can not comment on the spikes as I have two mounts with EC2 version and there are zero suspicious peaks. OK, that is what encoders are for.

regards Rainer