any solutions for 5.6-s oscillation on CEM40EC

[Michael Borland]

I see a 0.6" peak-to-peak oscillation with a 5.6-s period on my CEM40EC. This is with the guide camera (ASI290MM mini) looking directly through my OTA (8" LX90-ACF, FL=2000mm) (even the OAG is out of the way to eliminate any possible confusion). A guide log is attached. The 15-minute-long record is perhaps the most interesting, since I adjusted the RA rate to more or less remove the slow drift.

It seems this oscillation is too fast to correct, since by the time PHD2 sees it, it's probably too late (I get ~1 update rates even with short exposures). For now, I've settled on using a 6-s guide camera exposure time average the oscillation essentially to zero, so PHD2 doesn't chase it. I also found the LowPass2 method seems to work pretty well. However, I still see slightly oblong stars because I'm usually imaging at about 0.5"/pixel. The star shape doesn't depend much on the main camera exposure time, except when it is very short, in which case it oscillates with half 5.6-s period, as I'd expect.

Because the oscillation is very narrow in the frequency domain, it should have constant phase for a significant period of time. In light of this, I wonder if it can be countered by a simple preset series of alternating pushes at the position of the predicted zero crossings. Using a short exposure time on the guide camera, I should see the 5.6-s period showing up if the phase and amplitude of these kicks is not right. Hence, a slow adaptive loop could perform an FFT the last, say, 60s of star positions and use that to adjust the phase and amplitude of the kicks. Is there any way to try this in PHD2?

Thanks in advance for any comments or advice.

--Michael

[Brian Valente]

Hi Michael

i think your setup is causing you a lot of issues, specifically your calibration steps seem off, and you are guiding at 1/2 a second (whereas you typically want 3-5 second range)

if you look at your guiding assistant run, you can see the insane amount of instability in your DEC axis (red) *and* in your RA - this is without the motors moving at all

my suggestion is to do a baseline guiding with reasonable guiding numbers and see how it goes:

https://www.dropbox.com/s/os1thorvswmzaul/How%20to%20create%20a%20baseline%20for%20guiding%20results%20using%20OpenPHD2.pdf?dl=0

Brian

--
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.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/6fa0ae63-94ff-4404-9068-26c1d834d8bc%40googlegroups.com.

--

Brian 

Brian Valente

portfolio brianvalentephotography.com 

[Michael Borland]

Brian,

There is no attempt to perform guiding or imaging with these settings. As stated, I use a 6-s guide camera exposure when imaging. I was led to perform this diagnostic procedure with short exposures to understand why guiding with conventional parameters didn't eliminate my oblong stars when imaging at ~0.5"/pixel. Since the oblateness of the stars didn't go away even with short (e.g, 15s) imaging exposures, I set the guiding interval at 0.5-s deliberately in order to understand the mount behavior. 

The level of instability at 0.29"/pixel is from seeing and wind of course, but in RA is it also from the 5.6-s oscillation, which the FFT pulls out. My question is, given that the 5.6-s oscillation (a known issue in CEM40EC mounts) is so well defined, can't I correct it pre-emptively by kicking the RA at the inferred zero-crossings of the oscillation? 

--Michael

On Sunday, May 17, 2020 at 5:02:20 PM UTC-5, Brian Valente wrote:

Hi Michael

i think your setup is causing you a lot of issues, specifically your calibration steps seem off, and you are guiding at 1/2 a second (whereas you typically want 3-5 second range)

if you look at your guiding assistant run, you can see the insane amount of instability in your DEC axis (red) *and* in your RA - this is without the motors moving at all

my suggestion is to do a baseline guiding with reasonable guiding numbers and see how it goes:

https://www.dropbox.com/s/os1thorvswmzaul/How%20to%20create%20a%20baseline%20for%20guiding%20results%20using%20OpenPHD2.pdf?dl=0

Brian

On Sun, May 17, 2020 at 2:55 PM Michael Borland <michael...@gmail.com> wrote:

I see a 0.6" peak-to-peak oscillation with a 5.6-s period on my CEM40EC. This is with the guide camera (ASI290MM mini) looking directly through my OTA (8" LX90-ACF, FL=2000mm) (even the OAG is out of the way to eliminate any possible confusion). A guide log is attached. The 15-minute-long record is perhaps the most interesting, since I adjusted the RA rate to more or less remove the slow drift.

It seems this oscillation is too fast to correct, since by the time PHD2 sees it, it's probably too late (I get ~1 update rates even with short exposures). For now, I've settled on using a 6-s guide camera exposure time average the oscillation essentially to zero, so PHD2 doesn't chase it. I also found the LowPass2 method seems to work pretty well. However, I still see slightly oblong stars because I'm usually imaging at about 0.5"/pixel. The star shape doesn't depend much on the main camera exposure time, except when it is very short, in which case it oscillates with half 5.6-s period, as I'd expect.

Because the oscillation is very narrow in the frequency domain, it should have constant phase for a significant period of time. In light of this, I wonder if it can be countered by a simple preset series of alternating pushes at the position of the predicted zero crossings. Using a short exposure time on the guide camera, I should see the 5.6-s period showing up if the phase and amplitude of these kicks is not right. Hence, a slow adaptive loop could perform an FFT the last, say, 60s of star positions and use that to adjust the phase and amplitude of the kicks. Is there any way to try this in PHD2?

Thanks in advance for any comments or advice.

--Michael

--
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...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/6fa0ae63-94ff-4404-9068-26c1d834d8bc%40googlegroups.com.

[Michael Borland]

Here's the FFT analysis I mentioned.

--Michael

 

[Brian Valente]

>>>>  I set the guiding interval at 0.5-s deliberately in order to understand the mount behavior. 

Hi Michael

it would be interesting to see your guide log from the longer exposures

I mentioned before, if you look at unguided results during your GA run of your DEC axis, you can see the significant instability. and the DEC motor isn't running at all. So that suggests to me the instability is your seeing conditions, and not a function of the mount performance. 

your RA oscillation of 5.6s is a very small amplitude, and a very fast time period. I doubt you could effectively guide that out predictively, but you could try PPEC with a fixed 5.6sec period with high predictive and low reactive values, and guide at 500ms, but there's so much that could go off the rails there. Might be worth a shot. I did something similar but it was for a 41 sec interval

another possibility is using an AO unit which can guide at that kind of rate fairly well

To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/cc2a1fd4-ee92-482e-aaa9-5811903907a8%40googlegroups.com.

[Michael Borland]

Brian,

Here's a guide log from a week ago. This was my first time imaging with an OAG, but after a bit of screwing around, there's a section where I guided for 2h42m with 6s guide exposures. The ASI290MM mini was on the OAG, while the ASI071MC was used for imaging, both with FL=2000mm, f/10. This worked pretty well, and I accumulated 63 subs of 180s each. I'm still not quite happy with the oblong stars in the RA direction, which you can see in the result:

https://www.flickr.com/photos/128799619@N07/49879310906/

The full moon didn't help, but that's not something PHD2 can solve unless I'm misunderstanding something. ;) Any comments or advice will be appreciated.

The problem I think I would get into with PHD2's PPEC algorithm is that I have other sources of drift to contend with. Would the PPEC handle those as well? 

--Michael

On Sunday, May 17, 2020 at 5:37:13 PM UTC-5, Brian Valente wrote:

>>>>  I set the guiding interval at 0.5-s deliberately in order to understand the mount behavior. 

Hi Michael

it would be interesting to see your guide log from the longer exposures

I mentioned before, if you look at unguided results during your GA run of your DEC axis, you can see the significant instability. and the DEC motor isn't running at all. So that suggests to me the instability is your seeing conditions, and not a function of the mount performance. 

your RA oscillation of 5.6s is a very small amplitude, and a very fast time period. I doubt you could effectively guide that out predictively, but you could try PPEC with a fixed 5.6sec period with high predictive and low reactive values, and guide at 500ms, but there's so much that could go off the rails there. Might be worth a shot. I did something similar but it was for a 41 sec interval

another possibility is using an AO unit which can guide at that kind of rate fairly well

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/cc2a1fd4-ee92-482e-aaa9-5811903907a8%40googlegroups.com.

[bw_msgboard]

Hi Michael.  I think Brian has you on the right track here.  It would help if you could also post one or two raw, full-resolution, unprocessed main camera images from this session including the timestamps when they were taken.  Those will let us evaluate the star elongation quantitatively while matching it against the guiding performance. 

 

Thanks,

Bruce

 

---

To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/6db2cf28-5df4-4a45-8661-122636b8a656%40googlegroups.com.

[Michael Borland]

Bruce,

You can find three unprocessed subs at

https://drive.google.com/open?id=1_-kjZETmLX_KTVBxfv07ca6TiBHW8NJ1

The filenames contain the time stamp in HH-MM-SS format. 

Thanks!

--Michael

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/6db2cf28-5df4-4a45-8661-122636b8a656%40googlegroups.com.

[bw_msgboard]

Hi Michael.  The zip file is giant and I don’t want to get tangled up with the Google tools for unzipping it before downloading.  Trying to download the whole zip file seems to get wedged after Google complains it can’t do a virus scan.  Could you just post one representative FITs file so I can look at that?

 

Thanks, sorry to bother you with this sort of stuff,

To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/896af652-8539-4761-873b-4428a06891f7%40googlegroups.com.

[Michael Borland]

Bruce,

Sorry for the monster ZIP file. I uploaded two FIT files

https://drive.google.com/open?id=16YnbG_KeRD3Sab-eCXD9816UbX0nmKT2

https://drive.google.com/open?id=1bf4carSWir7veWDw57xjABvUKpuPjtr3

--Michael

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/896af652-8539-4761-873b-4428a06891f7%40googlegroups.com.

[Brian Valente]

Hi Michael

thanks that was helpful

your guiding RMS results don't look that bad, particularly if this is an OAG

looking at your 2 hr run, chopping out two hiccups, your overall RMS is 0.45" with RA being 0.34" and DEC at 0.30" - the difference is .13 pixels

i don't know if you would even see that in your results? that was the total RMS for nearly 3 hours

one thing that did jump out is really poor guidestar focus and SNR. For a 6 second exposure i would expect star mass and SNR to be really high, but instead it's oscillating quite a lot (yellow and white)

your guidestar FWHM is nearly 7, so i think if you work on improving your guide camera focus, you will fare better

>>>The problem I think I would get into with PHD2's PPEC algorithm is that I have other sources of drift to contend with. Would the PPEC handle those as well? 

Not in the way I proposed. you are focused on solving a very fast (for guiding) 6 second oscillation. The most ideal place to address this is in the mount. Absent a fix there, you *could* try to brute force PPEC to do it, but it's a very specific application of predictive error at just 6 seconds. the algorithm measures for a cycle and then begins predicting what are the oscillations. If your additional sources are periodic and fall within 6 seconds, it would pick that up, but the things you are describing re: drift probably are not

If your goal in setting a 6 second exposure was to offset the mount's known error, you would be far better to use a shorter exposure like 1.5-2 sec so the mount has time to react to things like that. at 6 seconds, it's already too late for the mount to do anything 

i'd also increase your aggressiveness to 80-90.

But again, looking at your guiding results, improving your guide camera focus and aiming for shorter exposures i think would help an already good guiding result

To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/6db2cf28-5df4-4a45-8661-122636b8a656%40googlegroups.com.

[Brian Valente]

Boy it's really hard to evaluate this with so few stars - maybe you can grab a frame or two from somewhere in the milky way at some point?

i'm wondering if what you're seeing isn't at least partly optical. the star shapes don't appear to be consistent in all corners (again, really hard to tell definitively with so few stars) but looking at a quick curvature analysis, is it possible your back focus is off a little bit?

this it the plot i'm looking at, which is quite a curve for a refractor:

To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/6edc2c69-2472-41a9-ab2e-fbc90240680f%40googlegroups.com.

[bw_msgboard]

Thanks, Michael.  I think this confirms what we suspected, that your star-shape problems are not caused by guiding.  Here are the average aspect ratios of the stars in these two images:

 

 

The aspect ratio is the percentage difference between the lengths of the major and minor axes of the Airy star disks in the image.  By most standards, a value of 10% or less is excellent, generally not noticeable to the human eye.  This corresponds to an ellipticity of 0.4 if you’re more accustomed to that measure.  Taking it a step further, we can see most of the elongation in your images occurs with stars that are nearer the margins of the image, and there is significant variation among the stars in the field.  This points to optical effects of one sort or another because guiding/tracking errors are going to affect all the stars equally.

 

As Brian said, your tracking/guiding performance is quite good with little difference between RA and Dec RMS.  That tells us you should be getting round stars, and I think you essentially are subject to the optical limitations.  So I think that fiddling around with your guiding operations would be at best a waste of time at this point, and you can probably focus your attention on other parts of imaging – post-processing and probably better optical performance.

 

To come back to your original questions – which we never answered – there really isn’t any way in PHD2 to direct a particular guide command at a particular time.  It could be done, I suppose, if you wanted to write your own guide algorithm or develop a companion application that would control the mount itself.  It would be a challenging project because usually there is no way to learn the gear angle at the time you start guiding.  The mount firmware often knows that number but it isn’t reported through any standard interface.  So each time you slew the scope, you would have to repeat the process of deducing where you are in the RA gear phase.  The PPEC algorithm does that but it isn’t an easy problem and it becomes more difficult as the target period gets smaller.  That’s because the signal you’re looking for is increasingly lost in the seeing noise.  You mentioned a couple of times that you were concerned about drift rates.  For guiding, drift is typically a non-problem because it is the easiest thing to correct for and rarely has any effect on the final results.  

 

Hope this helps,

To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/6edc2c69-2472-41a9-ab2e-fbc90240680f%40googlegroups.com.

[Michael Borland]

Bruce and Brian,

Many thanks for the analysis. I had myself convinced that it was a mount problem, but your analysis has me doubting that conclusion.  I analyzed some images of M57, which have somewhat more stars, and these showed similar results. I am still puzzled by the fact that the stars are all elongated in the RA direction, which could be just coincidence of course. Presumably, if it was field curvature, all the ellipses of stars should point toward the center of the FOV. Does that indicate a tilt in the optical system?

A few more observations:

1. I checked for a tilt in the imaging train by rotating the entire imaging train by 90 degrees about the telescope axis. The ellipses kept their orientation relative to the celestial coordinate directions, which means it is in the OTA or mount. (Unfortunately, I can't rotate the OTA relative to the dovetail.)

2. I'm pretty sure my backfocus is within 1 mm of the Meade's specified value of 4.57" for the LX90 ACF. I have a variable T extender that I'll use to play with this and see if I can make any improvements. 

3. Looking at out-of-focus images of stars I concluded that the collimation was quite good, but have an artificial star that I'll try, since that should be more controlled. I guess I can also use this to check the imaging curvature. 

4. With a non-correcting Antares 0.63 reducer, my star images are very round. (I don't often image that way because I'm interested in galaxies this time of year and want higher magnification.) I had concluded that this was just because with a 0.63 reduction factor, my guiding issues were invisible.

5. Some time ago I acquired a series of short exposures (1s I think) in rapid succession using the imaging camera, then analyzed their eccentricity using PixInsight. It showed a sinusoidal variation, which I interpreted as variation in the spot sizes with phase during the 5.6-s oscillation (the effect is expected to be worst at the zero crossings and least at the crests). This still seems convincing to me but I need to repeat the experiment since I've made other improvements since then and I deleted the original images.

Thanks again! I'm relatively new to the on-line astronomy community and I'm very impressed by how helpful people are.

--Michael

[Brian Valente]

If you get some time, it would be interesting to see a few frames in a heavy starfield (think milky way), maybe 3 images of about the same exposure 180s as I recall

that could shed more light on star shape details

Brian

--
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.

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/cfdf4def-a4f8-4807-a13f-01a04de24317%40googlegroups.com.

[Michael Borland]

Brian,

That's definitely on my list of things to try, if it ever stops raining here.

--Michael

On Monday, May 18, 2020 at 8:26:19 PM UTC-5, Brian Valente wrote:

If you get some time, it would be interesting to see a few frames in a heavy starfield (think milky way), maybe 3 images of about the same exposure 180s as I recall

that could shed more light on star shape details

Brian

On Mon, May 18, 2020 at 6:04 PM Michael Borland <michael...@gmail.com> wrote:

Bruce and Brian,

Many thanks for the analysis. I had myself convinced that it was a mount problem, but your analysis has me doubting that conclusion.  I analyzed some images of M57, which have somewhat more stars, and these showed similar results. I am still puzzled by the fact that the stars are all elongated in the RA direction, which could be just coincidence of course. Presumably, if it was field curvature, all the ellipses of stars should point toward the center of the FOV. Does that indicate a tilt in the optical system?

A few more observations:

1. I checked for a tilt in the imaging train by rotating the entire imaging train by 90 degrees about the telescope axis. The ellipses kept their orientation relative to the celestial coordinate directions, which means it is in the OTA or mount. (Unfortunately, I can't rotate the OTA relative to the dovetail.)

2. I'm pretty sure my backfocus is within 1 mm of the Meade's specified value of 4.57" for the LX90 ACF. I have a variable T extender that I'll use to play with this and see if I can make any improvements. 

3. Looking at out-of-focus images of stars I concluded that the collimation was quite good, but have an artificial star that I'll try, since that should be more controlled. I guess I can also use this to check the imaging curvature. 

4. With a non-correcting Antares 0.63 reducer, my star images are very round. (I don't often image that way because I'm interested in galaxies this time of year and want higher magnification.) I had concluded that this was just because with a 0.63 reduction factor, my guiding issues were invisible.

5. Some time ago I acquired a series of short exposures (1s I think) in rapid succession using the imaging camera, then analyzed their eccentricity using PixInsight. It showed a sinusoidal variation, which I interpreted as variation in the spot sizes with phase during the 5.6-s oscillation (the effect is expected to be worst at the zero crossings and least at the crests). This still seems convincing to me but I need to repeat the experiment since I've made other improvements since then and I deleted the original images.

Thanks again! I'm relatively new to the on-line astronomy community and I'm very impressed by how helpful people are.

--Michael

On Sunday, May 17, 2020 at 4:54:58 PM UTC-5, Michael Borland wrote:

I see a 0.6" peak-to-peak oscillation with a 5.6-s period on my CEM40EC. This is with the guide camera (ASI290MM mini) looking directly through my OTA (8" LX90-ACF, FL=2000mm) (even the OAG is out of the way to eliminate any possible confusion). A guide log is attached. The 15-minute-long record is perhaps the most interesting, since I adjusted the RA rate to more or less remove the slow drift.

It seems this oscillation is too fast to correct, since by the time PHD2 sees it, it's probably too late (I get ~1 update rates even with short exposures). For now, I've settled on using a 6-s guide camera exposure time average the oscillation essentially to zero, so PHD2 doesn't chase it. I also found the LowPass2 method seems to work pretty well. However, I still see slightly oblong stars because I'm usually imaging at about 0.5"/pixel. The star shape doesn't depend much on the main camera exposure time, except when it is very short, in which case it oscillates with half 5.6-s period, as I'd expect.

Because the oscillation is very narrow in the frequency domain, it should have constant phase for a significant period of time. In light of this, I wonder if it can be countered by a simple preset series of alternating pushes at the position of the predicted zero crossings. Using a short exposure time on the guide camera, I should see the 5.6-s period showing up if the phase and amplitude of these kicks is not right. Hence, a slow adaptive loop could perform an FFT the last, say, 60s of star positions and use that to adjust the phase and amplitude of the kicks. Is there any way to try this in PHD2?

Thanks in advance for any comments or advice.

--Michael

--
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...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/cfdf4def-a4f8-4807-a13f-01a04de24317%40googlegroups.com.

[sebastian ip]

For what it's worth there's no way to "guide" out those oscillations. 

What happens from what we determined is in various folk's testing is that when you issue a guide command the mount appears to disregard the encoders for a period of time while it execute the guide command. So if you send guide pulses fast enough to counter the SDE oscillation you might a well just not have an encoder. 

[Michael Borland]

I asked ioptron technical support if there was a cure for the oscillation. They said they are working on a firmware upgrade that will reduce it, but from the sound of it, it's a modest improvement for the CEM40EC. If only there was a way to disable the encoder.

Michael

[sebastian ip]

That's the standard response. It's been the same message for the CEM60EC since the original SDE issue surfaced more then 5 years ago.

The good news is that considering normally guided mounts in the mid range even with the SDE often you see no hints of it in the images.

My suggestion is to forget about vaporwear fix from them they have had years to get it right but haven't. And just use the mount or if it really bugs you sell it. For me it's not so bad that I need to get rid of it right now, and I can't in good faith sell it without providing a large discount as I know the SDE is here, so I am using it for now till it makes sense to change for another mount.

[sebastian ip]

Another thing is if it's within return period and this bugs you, take full advantage of it. Post return period the options from iOptron for return or exchange involves significant pro-rating cost based on years of use and restocking fees. 

It's left a bad taste in my mouth and I will not be a repeating customer for them.

[Michael Borland]

Unfortunately, I'm outside the return period given how long it's taken me to understand the problem. Maybe I'll ask them what they'll give me for a trade up on a CEM60. I'm partial to their mounts because they are light for what they carry, which accommodates my physical limitations.

Michael

[sebastian ip]

A cem60 is heavier than you'd think. If weight is a limitation you might want to go and try to lift one before trading.

I think you'll find their offer not appealing. Use it if you can't see star elongated.

[Michael Borland]

Yes, 26 lbs is a bit intimidating for me when it's a precision component and not, say, a box of kitty litter. Maybe they'll throw in a robotic exoskeleton to make it a good trade? ;)

Michael

[Brian Valente]

>>> Maybe they'll throw in a robotic exoskeleton to make it a good trade? ;)

okay that made me laugh :)

On Tue, May 19, 2020 at 8:02 PM Michael Borland <michael....@gmail.com> wrote:

Yes, 26 lbs is a bit intimidating for me when it's a precision component and not, say, a box of kitty litter. Maybe they'll throw in a robotic exoskeleton to make it a good trade? ;)

Michael

--
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.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/5bed21ad-4722-4ee3-ab14-252e84569dd5%40googlegroups.com.

[Al Moncayo]

Hi Michael,

I struggled for months, patiently, with exactly your issue. I tried every single beta FW that iOptron sent me. None worked to eliminate the elongated stars along RA. I confirmed across cameras and scopes (even down at 660 mm FL) that this elongation is the SDE that Sebastian is referring to.

My solutions to the issue were:

1. Avoid large FL imaging as much as possible.. I stay at or below 660 mm for the time being, with image scales >1"/px

2. Dedicated the 40EC largely for wide-field by using it almost exclusively with a WO Redcat, 250 mm FL scope. (Although recently I've piggy-backed the WO onto a 110 mm refractor; with the combination topping the scale at over 23 lbs; and that worked pretty well with the latest FW.)

3. To point #2, I did confirm that the latest released FW has now been giving me consistently better star shapes than other FW versions. 660 mm was hit and miss last year. Now I can use it consistently.

4. Finally, last year when I was imaging with an RC8 at 1600 mm (well below 1"/px), I found that the only way to achieve round stars was to worsen Dec guiding to about 40-50% worse error than RA - as reported in PHD. As an example, if RA is at 0.4" RMS, then shoot for Dec to get up to about 0.6" RMS. From my testing, this worked relatively well. I had to tune the Dec axis to absolutely minimize backlash, then force lots of Dec corrections by using tiny min-mo.

Sorry, PHD folks, my responses have little to do with PHD.. well, except perhaps #4.

Good luck Michael...

Al

[sarg314 sarge]

Michael:
I have a cem40ec with the same problem.  Period is 5.6 sec. amplitude is about 3 arcsec. See attached output from Guiding Assistant runs. The mount had the problem from the very first time I used it last September.  I have been going back and forth with iOptron ever since. Very frustrating.  Recently a different person started responding to me (management?) and implied (did not say directly) that they might be able to replace some parts to fix it. The graphs below are  from the 800 sec. G.A. run he asked me to send to him.  I'm still waiting to hear back.  They are very hard to deal with on this problem.

Virus-free. www.avast.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.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/23d79d2c-b4d0-404c-a072-cca2af821b64%40googlegroups.com.

--

Tom Sargent

[Michael Borland]

Al,

Thanks for the encouragement. I'm using the latest firmware (which definitely helped). I use a 0.63 reducer on my OTA (FL=1260mm, 0.78"/pixel), everything is good enough. However, that's just not what I usually want to do, since galaxies and globular clusters need the long FL, typically. I've asked iOptron (and the original retailer) what they can do by way of a trade-in. We'll see how bad the deal is...

--Michael

[Michael Borland]

Tom,

Your problem is about three times worse than mine! That really sounds hard to tolerate.

With so many people having these problems, I'm surprised there hasn't been a class action lawsuit. Perhaps working through the Better Business Bureau would be helpful.

--Michael

To unsubscribe from this group and stop receiving emails from it, send an email to open-phd...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/23d79d2c-b4d0-404c-a072-cca2af821b64%40googlegroups.com.

--

Tom Sargent

[Michael Borland]

Al,

By the way, I meant to ask, how do you worsen the DEC guiding? That might work for me as a stop-gap at FL=2000mm.

Thanks.

--Michael

[Al Moncayo]

sarge314,

Your mount looks to be a bit worse than mine. I wrote a matlab utility to do a bit more processing than PHD. Per multiple runs I did last year, my 5.6 second oscillation is solidly at 0.7" RMS. Yours looks to be around 0.9" RMS from the data you showed.

Al

To unsubscribe from this group and stop receiving emails from it, send an email to open-phd...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/23d79d2c-b4d0-404c-a072-cca2af821b64%40googlegroups.com.

--

Tom Sargent

[Al Moncayo]

Michael,

One option is to trade down to a standard CEM40. If you are up for a bit of experimentation, email me separately - or join the CEM user group. I have some info that may be of interest to you before going down that path.

Al

[Al Moncayo]

Michael,

As I noted in my other post, I first tuned the Dec axis to insure the belt was snug and gear mesh optimum, so as to minimize backlash. Then I used 100% aggressiveness together with an unreasonably small Dec MinMo. Note that it doesn't always work. But when it has worked, I was able to achieve acceptably round stars at 1600 mm FL.

Al

[Michael Borland]

Al,

How do I find the CEM40 user group? I couldn't it on google or google groups.

--Michael

[sarg314 sarge]

I am using an OAG on a Celestron 8" f/10 so guide focal length  is 2032mm.  It almost works if the guiding integration time is short.  At 0.5 sec or 1.0sec I can often get guiding errors in the 0.6 - 0.7 arcsec RMS range, though often the star images aren't quite round.  Any longer integration period than that and the guide period starts to beat with the 5.6 sec oscillation and it goes to hell pretty quick.  It's not really usable.  I've been effectively shut down since I bought this thing.

The Dec. in this run was a bit worse than normal.  I was experimenting with a shorter Dec. minmove.  It was too short. 

--

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.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/f59e180b-0ac6-4cae-aabb-8b66f9999838%40googlegroups.com.

--

Tom Sargent

[Michael Borland]

Tom,

I'm also using an OAG on a 2000mm FL SCT. I tried what you suggest but found that my seeing was never good enough to make it stable. I got the best results by using a 6s exposure, which averages out the oscillation almost completely and at least allows guiding out any other problems. (My mount also exhibits linear RA drift that varies with pointing direction. I've seen it as bad as 2" per minute.)

Michael

[sarg314 sarge]

Interesting.  I tried 6 sec. and it was terrible.   I don't have RA drift.  Although, every time I run the Guiding Assistant, the RA rate is way off for the first 25 seconds or so, as you can see from the plot. After that, RA rate stabilizes on average, but then the 5.6 sec. oscillation begins. 

They wouldn't even admit it was a known problem until I told them I had communicated with other iOptron owners with the same problem.  They are very good at running out the clock.

--
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.

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/e70d1de7-0aac-4632-b9ac-f307855744ad%40googlegroups.com.

--

Tom Sargent

[sarg314 sarge]

Michael:

Last Friday iOptron got their head straight(er) and gave me an RMA number to send my mount back.  They have "some parts" from China and will try to repair the mount.  It only took 8 months for them to think of this.

--

Tom Sargent

[ray.h...@gmail.com]

I have the same problem on my CEM40ec..about 10 cycles/minute.  I’ve fiddled with Phd2 parameters and tightened the RA belt but to avail.  Please keep us posted as to your success.  I Really don’t want to send my mount back to MA!

Ray

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/CAN2tpX4y9Q5M6Sp1OHqMxDkADeX3Mb9yTDYSoudUZZ6cJ9Oz2A%40mail.gmail.com.

[sarg314 sarge]

Ray:

I think most cem40 owners are very happy with their mount.  A few of us get the ones with the oscillation problem. I don't know of any one who has had the problem fixed (maybe there are some, but I haven't run across them).  So you should hope that sending it back will work.  I think it's your only hope of having a working mount.

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/004c01d63e80%244e0d2a70%24ea277f50%24%40gmail.com.

--

Tom Sargent

[Al Moncayo]

Michael, Ray, Tom,

Last year, in working closely (at least so I thought) with iOptron, I tested all of their beta releases during the summer months. To date, I've had no success with focal lengths >660 mm on the CEM40EC. My goal was to use it with an 8" RC at 1600 mm FL. I've since given up and use the 40EC almost exclusively with my 250 mm FL Redcat51. With that scope I can even image unguided if I'm careful and take my time to minimize PA error. When I do guide, I guide with 7-10 second exposure and very low RA and Dec aggressiveness.

What PHD reports, irrespective of the scope FL, is what it has integrated over time. That integration has filtered out the SDE oscillation, which is there regardless of what PHD reports. When I did attempt imaging at 1600 mm FL, I would see total guiding error reported by PHD of <0.5" RMS, with nearly equal RA and Dec individual errors. One would be led to conclude that the stars would be round. That was never the case. Both my brother and I (we bought the 40EC mount at the same time) tried several scope/camera combinations, including on-axis and off-axis guiding. The star elongation along the RA axis was always visible.

Having said all that, I did receive one copy of FW from iOptron that was very interesting. As best I was able to determine from my testing, it basically disabled the encoder. The mount appeared to work and respond as if there was no encoder. Unfortunately, at that time I didn't have the opportunity to test it at large focal lengths.

If any of you would like to try testing that version of FW at large focal length, please contact me privately. My brother is not so eager to do any testing as he mainly images with a small StellarVue refractor. With that setup, he has no issues. Unfortunately for me, he's the one with the 8" RC that I handed down to him. I had intended to upgrade to a carbon fiber 8" RC, but that has not yet materialized -- in large part due to the issues I've had with the 40EC.

Al

On Tuesday, June 9, 2020 at 8:10:06 PM UTC-7, sarg314 sarge wrote:

Ray:

I think most cem40 owners are very happy with their mount.  A few of us get the ones with the oscillation problem. I don't know of any one who has had the problem fixed (maybe there are some, but I haven't run across them).  So you should hope that sending it back will work.  I think it's your only hope of having a working mount.

[sarg314 sarge]

I have tried 2 or 3 different firmware versions including a beta release.  They kept telling me to wait for the next firmware release, though they couldn't say when that would be.  I never saw any improvement.  

--
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.

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/d9916532-262d-4a19-b281-261499d9fddfo%40googlegroups.com.

--

Tom Sargent