Rainbow Astro RST-135E help

[Andrew Burwell]

I've recently started using the new Rainbow Astro RST-135E incremental encoder version of their mount. I've guided with it for two nights. The first night I played around with the settings and ran guiding assistant. The mount has almost no backlash but the assistant suggested adding it. Anyhow, I didn't stick with this setting (guiding was worse) and kept adjusting things to get to where I was at the end of the first log on 7/17. For the second log on 7/18 I used the same settings at the end of the 7/17 and let it run all night.

I'm guiding a 8" Edge reduced to 1442mm focal length, with an ASI174 guide camera on an OAG. I'm not sure if I'm fighting with the incremental encoder, or the strain wave gear, or the combination. The guiding was ok at .64 RMS. There are strange DEC oscillations that keep appearing, My pixel scale is .67, so I'd like to get RMS down to at least .5 reliably if possible to have more of a safe zone for imaging.

The standard 135 mount has a large +/- 30 PE per cycle, so on that mount you guide at a high rate to keep up with corrections. The incremental encoders added to this new mount are there to tame the large PE and produce a smaller +/- 2.5 PE per cycle. There is no mount modeling for this mount like 10Microns, so there's no detailed pointing model and expected corrections for location in the sky.  I don't know enough information about PHD2 and the settings to make educated guesses. I've just been looking at the tool tips and trying to interpret how they might need to change based on what I'm seeing.

Any help would be appreciated.

[Bruce Waddington]

Hi Andrew.  Some mounts with absolute encoders seem to require a slightly different approach to guiding.  The issue is that the encoders can “silently” continue to move the mount even after the mount driver has reported completion of a guide command.  From a software point of view, this is of course undesirable because the driver is unintentionally misrepresenting the status of the guiding operation.  On the other hand, if the tracking accuracy is very good, there is no need to use a high guiding rate (cadence) to make mount corrections.  One approach is to use a normal guide camera exposure – in the range of 1-2 seconds to retain high a SNR – but to force a delay between the completion of a guiding operation and the start of the next guide camera exposure.  The theory is that this will give the encoders time to complete their activity before the next guide camera exposure is started, thus bypassing the driver’s inability to know what’s going on.  This can be accomplished with the time-lapse control in the Advanced Settings dialog:

 

 

Users have reported that some well-constructed, absolute-encoder mounts can work well with guide commands being issued at something like 10-second intervals.  This allows the guiding software to correct for polar alignment error, flexure, and atmospheric refraction without running afoul of the encoder activity.  I don’t know if this would apply to your mount, but you could certainly try it.

 

Further, I think the LowPass2 algorithm might be a better choice for declination assuming that guiding corrections are only needed for the slow tracking errors mentioned above, particularly polar alignment error.  You can also experiment with further reducing the aggressiveness settings to further reduce the likelihood of the guiding conflicting with the encoder activity. With respect to your goal of guiding below 0.5 arc-sec RMS, that will depend a lot on your local seeing conditions.  Judging from your short Guiding Assistant run, I’m not convinced your seeing conditions on 2/17 were good enough for that.  That might also explain the relatively large star-sizes you were getting (greater than 4 px) assuming you’ve done a careful job of focusing the guide camera.

 

If you decide to experiment with these changes, please let us know how things go.  Even with what you’ve done to this point, you’re likely to be getting good results on your main camera images with nicely round stars.  Very rapid deflections and immediate corrections like you’re seeing here are quite often invisible in the final images although I understand your interest in reducing them. 

 

Regards,

Bruce

--
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/b4d8943e-a5b1-4a88-bda2-492daa529917n%40googlegroups.com.

[Andrew Burwell]

Thanks Bruce, I'll give it a try and see how things go.

[Bill Blanshan]

Andrew, did you run the guiding assistance so it can calculate a suggested Min/Mo and aggression settings.

I am trying to find specs to this mount, I dont know what the mechanical accuracy is nor the type of renishaw encoder they use.  Do you have this info?

[Bill Blanshan]

Im wondering if the gear ratio or their Renishaw incremental encoder resolution isn't high enough on the 135 for precise tracking??  The movements are very jagged on RA which should be controlled by the encoder in a closed loop setup.   Comparing your tracking (below image left) to my EQ8-Rh (right) which uses a Renishaw Incremental encoder, I am getting around 0.3" and sometimes lower and that's in Houston TX seeing.   Your DEC is not encoded so if the strain wave gear sticks, you will see sharp jumps which if there was an encoder on it, it would prevent this from happening., and you do have large jumps.   Not sure why (other than cost saving) they dont put encoders on both axis.  Regardless, maybe also try setting your guiding to 2.5 or 3 sec?  Are you using multi-star guiding?  I am still not a fan of strainwave gears yet because of these reasons.   

[Andrew Burwell]

Hi, yes, I did run the guiding assistant in the first log (section 5). Then I applied the settings right after that run. The guiding assistant asked me to implement backlash compensation, which the mount doesn't have, so after running a few minutes with worse guiding, I turned that part off. Then adjusted aggressiveness, the DEC algorithm and min-mo settings to try a few different things before settling at what I considered the best guiding of the night for the last session (15). 

The encoder is the Renishaw ATOM DX https://www.renishaw.com/en/atom-dx-encoder-series--42875 And it is only on the RA axis. Other than the addition of the encoder, the mount is identical to the standard RST-135.

[Andrew Burwell]

Interesting comparison with your EQ8. I'm in Houston as well, hence the poor seeing from my logs :)  I was using multi-star guiding. The main reason they didn't put encoders on both axis is to keep the mount very small. The primary reason for even having it is to tame the PE of the strain wave gear. Rainbow Astro really made this version of the mount to be able to run unguided on small focal length scopes, but because they've reduced the PE so much, it opens up opportunity to use it with long focal length scopes like the 8" Edge that I have on there. I have sent these logs to Rainbow Astro as well, and asked them about the DEC, but have not heard back from them just yet.

[Bill Blanshan]

If they are using the glass disc version, the encoder resolution is WAY to low for sub acr sec tracking.  The Renishaw encoder on the EQ8-Rh has 11million counts per rev.  I am not sure why they even picked that encoder if that's the case. Now I am really curious as to what their encoder resolution even is.  

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/9d7c33a4-2cdd-4651-b920-c0fc4f1fb09cn%40googlegroups.com.

[Bruce Waddington]

This discussion has taken quite a different turn from where I thought we started.  If the Dec axis doesn’t have an absolute encoder, then obviously the Dec deflections you see have nothing to do with encoder activity.  So I don’t think it’s a question of how to fine-tune PHD2 to work with your mount – I think it’s more a matter of understanding the limitations of the mount.  Here’s an illustration of some poor Dec behavior – note this is well after you disabled PHD2 Dec backlash compensation:

 

 

You can see that direction reversals on the axis become unstable.  When the guide pulse is south (down), it takes two consecutive guide pulses to get the axis moving – at which point it overshoots.  For the north (up) guide commands, the mount immediately over-corrects.  This can be caused by imbalance of the scope in Dec and/or by stiction in the Dec drive.    I assume you’ve disabled any form of Dec backlash compensation in the mount controller.  By the way, this also agrees with the GA result you got where a 900ms backlash was measured.  That’s consistent with the need for two consecutive south moves before the axis moves.  So I think you probably need to reset your expectations a bit about the mount’s capabilities, at least in Dec, and expect that it probably exhibits some of the common problems found with non-encoder mounts in its price range.

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/913fc3a6-db8d-4848-85ac-a052a70d9258n%40googlegroups.com.

[Brian Valente]

>>>   I assume you’ve disabled any form of Dec backlash compensation in the mount controller.  

that's what I'm wondering, is there any backlash compensation enabled in the mount? definitely worth confirming and/or disabling.

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/007101d77dd8%2454b90ae0%24fe2b20a0%24%40earthlink.net.

--

Brian 

Brian Valente

portfolio brianvalentephotography.com 

[Andrew Burwell]

I had disabled backlash compensation. Initially Guiding Assistant recommended around 300ms of backlash compensation, but guiding got way worse with it, so I just turned it back off.

I managed to do some new logging tonight, but mainly tests. Section 4 of this log is about 10 minutes unguided at .5s exposures so that the characteristics of the mount can be well seen. A single revolution of the gear is 340s, so about 7 minutes.Then I ran guiding assistant in section 12 for more than a single period of the gear with 5 second exposures. Then Section 13 was my longest section where I was guiding and trying out different aggression settings and exposure lengths but I didn't really settle on anything. Aggression of about 50-55 seems to be the best.

I also ran all of tonight with DEC using the LowPass2 algorithm as Bruce suggested. I did also try adding 100ms of time lapse to the camera settings, but it made things worse, and I wasn't sure if I should adjust it to less or more, so I set it back to 0 and didn't use it again.

[Brian Valente]

Hi Andrew

just to clarify i was referring to disabling any backlash compensation in the mount settings if it exists, i wasn't referring to the backlash compensation feature in PHD

Also the idea behind Bruce's suggestion was to add longer delay via timelapse, say 10 seconds (10000ms) between "regular" guide exposures of approx 1-2 seconds each.

I think 0.5sec exposure is far too short for an encoder mount, at that point you are likely chasing seeing conditions. There's no need for an encoder mount to use that kind of correction. 

I have an astro physics 1600 with absolute encoders. my image scale is 0.53" @ about 3500mm. my guiding is 3-4 second exposures, with 10 second delay between. I guide around 0.2-0.3" total RMS. 

I know it's not the same mount, but just for comparison

Brian

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/46468e1b-705f-487a-b958-a4e9f8f4cde2n%40googlegroups.com.

[Andrew Burwell]

Brian, Ok, I think I understand the delay now. I should be able to test that next clear night. There's no backlash compensation in the mount controller. There is a polar drift compensation setting if you do a manual 5-7 star align, but I think this feature is just for visual, so I have it off.

The only reason I ran the log (unguided) with .5 seconds was to see if we could tell what the encoder was doing. I also have a 10Micron mount with very similar results as your AP mount. 

Getting this mount and scope setup was really a project scope for me to try and achieve long focal length imaging with an ultra light setup. And it's more or less there. I also have a regular RA RST-135 without encoders and have been happily using it for a year or so. But kept wishing you could do long focal length guiding with it (which it does really poorly).

Anyhow, the goal here is to try and characterize the mount, and possible issues, and see how well I can get guiding improved if at all over what I've seen already.

Here's a bit of info on the project if you're interested. https://www.macobservatory.com/blog/2021/7/19/project-dumbbell-first-light-with-the-rainbow-astro-rst-135e-and-edgehd-8

[George Shoup]

Hi Brian

For comparison, can you say what you do and your results with your

 losmandy mounts?  Good information.  

George

Sent from my iPhone

On Jul 20, 2021, at 10:56 PM, Brian Valente <bval...@gmail.com> wrote:



To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i7e128h4JYpN91ajMxtJP%3D2ENje04bN1xvJe4zw90CncQ%40mail.gmail.com.

[George Shoup]

Hi Brian:

Good information.

What settings do you use with your G11 and what results do you get?

George

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i7e128h4JYpN91ajMxtJP%3D2ENje04bN1xvJe4zw90CncQ%40mail.gmail.com.

[Brian Valente]

Hi George

I'm not sure if your post here was supposed to be a personal note to me? But it's sufficiently off topic here that I suggest either starting a new thread or just drop me an email directly

Brian

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/CAKfhV07UhNtAOdQ1s15-PKf3U2NWecn%2Bi%2BCUfVAK2eEeN6%3Dndg%40mail.gmail.com.