Author Topic: Losmandy G-11 PMC-Eight w/ Telescope Drive Master - UNGUIDED 600sec image  (Read 327 times)

wetrerede

  • Active Astronomer
  • ***
  • Posts: 121
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: Losmandy G-11 PMC-Eight w/ Telescope Drive Master - UNGUIDED 600sec image
« Reply #15 on: January 15, 2018, 08:39:37 PM »
Yes, my program TDM TE Recorder, implements a PID controller using the value read through the serial port on the TDM and outputs corrections via pulse guiding on my EQ6 Pro mount.

Jason Muse

  • Active Astronomer
  • ***
  • Posts: 104
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: Losmandy G-11 PMC-Eight w/ Telescope Drive Master - UNGUIDED 600sec image
« Reply #16 on: January 20, 2018, 09:16:04 PM »
Quote
Jerry, I have no doubt this is possible.

I was pointing out that the TDM + variable rate tracking would be way more useful.

I believe the TDM has a serial port, so the mount computer can talk to it (instead of relying on ST4 inputs from it). So it should be possible to pre-calculate the required tracking rate on the mount computer, then validate (via the TDM) that this tracking rate is being achieved.

Only a matter of programming

Since tracking is at 15 degrees/hour, even if you re-calculate the tracking rate every 5 minutes, that should be plenty. Mel Bartels has a nice simplified equation for determining the refraction-corrected tracking rate.

FYI - The ExploreStars application that comes with the PMC-Eight system implements the standard 2 and 3 star alignment routines when plopping the mount down close to polar aligned, but not great, maybe 5 - 10 degrees off. In this case the pointing is managed by the 2 or 3 star alignment, but he tracking is managed dynamically by a mode we call Point mode. In this mode, the position of the object is continuously (3 time per second) calculated and the scope is "moved" to the updated position. In this way a dynamic rate on both the RA and DEC axes is the result. The PMC-Eight firmware and command language supports this type of dynamic pointing/tracking mode out of the box.

Lcs King

  • Active Astronomer
  • ***
  • Posts: 132
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: Losmandy G-11 PMC-Eight w/ Telescope Drive Master - UNGUIDED 600sec image
« Reply #17 on: January 21, 2018, 02:39:06 AM »
Quote
Quote

Jerry, I have no doubt this is possible.

I was pointing out that the TDM + variable rate tracking would be way more useful.

I believe the TDM has a serial port, so the mount computer can talk to it (instead of relying on ST4 inputs from it). So it should be possible to pre-calculate the required tracking rate on the mount computer, then validate (via the TDM) that this tracking rate is being achieved.

Only a matter of programming

Since tracking is at 15 degrees/hour, even if you re-calculate the tracking rate every 5 minutes, that should be plenty. Mel Bartels has a nice simplified equation for determining the refraction-corrected tracking rate.

FYI - The ExploreStars application that comes with the PMC-Eight system implements the standard 2 and 3 star alignment routines when plopping the mount down close to polar aligned, but not great, maybe 5 - 10 degrees off. In this case the pointing is managed by the 2 or 3 star alignment, but he tracking is managed dynamically by a mode we call Point mode. In this mode, the position of the object is continuously (3 time per second) calculated and the scope is "moved" to the updated position. In this way a dynamic rate on both the RA and DEC axes is the result. The PMC-Eight firmware and command language supports this type of dynamic pointing/tracking mode out of the box.
i have a CGE wtih TDM and was wondering if it was possible , asorlyandico suggested, to dynamically change(using a software) the tracking rate instead of physically chaging it on the controller. i have a remote observatory and this feature would be awsome.

thanks

trualolalun

  • Active Astronomer
  • ***
  • Posts: 112
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: Losmandy G-11 PMC-Eight w/ Telescope Drive Master - UNGUIDED 600sec image
« Reply #18 on: January 23, 2018, 07:36:03 AM »
The TDM has a selection for the average King rate which works with altitudes around 30 - 45 degrees but of course doesn't do as you suggest. I could see one thinking that it would be an easy matter to implement the polynomial equation in the TDM to compensate for the refractive error, but typically microcontrollers don't have the trigonometric routines needed. There are math co-processors available to provide that functionality, but an easier although not as accurate solution would be to simply provide a lookup table. Having said that, you still need to provide the altitude value as an input to the controller. The TDM has no idea where the mount is pointing, it could not care less. Adding that input would be a bit cumbersome and would require two-way communications with the controller in real-time. My solution to this would be to integrate the high-resolution feedback from the RA axis into an existing mount controller such as the PMC-Eight and do away with the separate box. The beauty of the TDM is that it is a set and forget type of system where it provides a stand-alone, isolated solution that can be retrofitted to a number of different mounts. That is also it's limitation.

The only other possible way with the existing TDM design would be to hack the rate switch and figure out how the change in rate is impemented when switched. If a resistor value is changed the theoretically, you could substitute any resistor value in to set the rate to a dynamic value. I could see a separate module consisting of a voltage controlled resistor within a fixed resistor network scaled appropriately could be built to interface to the TDM (maybe, I don't know if that is possible, I would need to see the schematic to determine that). That interface would then need to be driven by an appropriate analog computing circuit (basically some op-amps resistors and capacitors) to develop the correct voltage values based on the altitude input from some source connected to the mount. As you can see, based on my description it would be a bit of a kluge, although a very interesting project in analog computing.

Thanks

contpeeresto

  • Active Astronomer
  • ***
  • Posts: 114
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: Losmandy G-11 PMC-Eight w/ Telescope Drive Master - UNGUIDED 600sec image
« Reply #19 on: January 23, 2018, 09:03:48 AM »
I don't think the encoder you are using has accuracy equal to it's resolution. Encoder error will manifest itself as a slow error in velocity. Has this been characterized? Perhaps this can be determined analytically for smooth low order errors. Also I don't see why you wouldn't pass thru autoguide signals which are wire ored. At least Dec axis. Your picture shows a 4 arc sec fwhm but it would be much better to run a phd guiding plot with guiding off and with your system on for a time long relative to the worm period.

asexdalo

  • Active Astronomer
  • ***
  • Posts: 118
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: Losmandy G-11 PMC-Eight w/ Telescope Drive Master - UNGUIDED 600sec image
« Reply #20 on: January 25, 2018, 01:44:48 PM »
For example if the encoder is only accurate to 10 arc min in 360 degrees that works out to a 4 arc second slow error over a 10 minute exposure. So hopefully the encoder is better than that.

Levi Cruse

  • Active Astronomer
  • ***
  • Posts: 126
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: Losmandy G-11 PMC-Eight w/ Telescope Drive Master - UNGUIDED 600sec image
« Reply #21 on: January 30, 2018, 04:20:45 AM »
Quote
For example if the encoder is only accurate to 10 arc min in 360 degrees that works out to a 4 arc second slow error over a 10 minute exposure. So hopefully the encoder is better than that.

The Heidenhain encoder has more than enough accuracy... the incremental rate is resolved to +/- 0.125arc-second (nominal). The limiting factor is not the encoder, but the accuracy of the internal clock/timing frequency. This internal clock is also somewhat temperature dependent but this has not been characterized. When I purchased my TDM initially in 2011, I did some measurements and wrote a paper that is shared on the MDA-Telescoop website. You can see what I found out about another aspect of the tracking accuracy there.

http://mda-telescoop...vestigation.pdf

niososerso

  • Active Astronomer
  • ***
  • Posts: 120
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: Losmandy G-11 PMC-Eight w/ Telescope Drive Master - UNGUIDED 600sec image
« Reply #22 on: February 02, 2018, 08:27:46 PM »
t
Quote

For example if the encoder is only accurate to 10 arc min in 360 degrees that works out to a 4 arc second slow error over a 10 minute exposure. So hopefully the encoder is better than that.

The Heidenhain encoder has more than enough accuracy... the incremental rate is resolved to +/- 0.125arc-second (nominal). The limiting factor is not the encoder, but the accuracy of the internal clock/timing frequency. This internal clock is also somewhat temperature dependent but this has not been characterized. When I purchased my TDM initially in 2011, I did some measurements and wrote a paper that is shared on the MDA-Telescoop website. You can see what I found out about another aspect of the tracking accuracy there.

http://mda-telescoop...vestigation.pdf[/quote]
Thanks for the 2011 paper reference. The number you cite above is resolution and not accuracy. I agree that clock accuracy is also important. How accurate are the clock source and the encoder?

Paul Syring

  • Active Astronomer
  • ***
  • Posts: 116
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: Losmandy G-11 PMC-Eight w/ Telescope Drive Master - UNGUIDED 600sec image
« Reply #23 on: February 04, 2018, 12:39:18 PM »
That Heidenhain has 5000ppr. It is not accurate to +/- 0.125" - physics prevents it.

(I have this from the Heidenhain reps themselves) the actual physical accuracy is in the +/- 10" range, and interpolation is used to get the smaller figure.

This is not to say that the interpolation is not accurate: but there always will be interpolation error and SDE (subdivisional error).

Nick Ellis

  • Active Astronomer
  • ***
  • Posts: 140
  • Activity:
    0%
  • Reputation: +0/-0
    • View Profile
Re: Losmandy G-11 PMC-Eight w/ Telescope Drive Master - UNGUIDED 600sec image
« Reply #24 on: February 09, 2018, 10:55:14 AM »
Quote
That Heidenhain has 5000ppr. It is not accurate to +/- 0.125" - physics prevents it.

(I have this from the Heidenhain reps themselves) the actual physical accuracy is in the +/- 10" range, and interpolation is used to get the smaller figure.

This is not to say that the interpolation is not accurate: but there always will be interpolation error and SDE (subdivisional error).

I understand Orly, that is accuracy, not precision. The incremental rate calculation relies on a precise incremental measurement, not an accurate position measurement. The sidereal rate is about 15 arc-sec/second so that if you are measuring the rate 5 times per second you would be measuring the elapsed counts from the encoder and getting a value of (15 arcsec/sec x 0.2 seconds x 8 counts/arc-sec) = 24 counts.

If the incremental precision is based on the error you stated, +/- 10 arc-sec which is a linearity error, and the per cycle increment is 259.2 arc-sec (360 deg / 5000 cycles) then the linearity percent error is +/- 3.85 %. If we apply that to the expected counts then the error in counts measured every 0.2 seconds would equal to 24 x 0.0385 or +/- 0.92 counts, or about +/- 1 count.

So if we add the two errors, interpolation and linearity, then we get 24 counts +/- 2 counts. so the measured incremental rate error will vary about +/- 0.25 arc-seconds (we should actually add these errors in quadrature since these are random errors, no systematic errors). Since this is a random error and not a systematic error this error gets averaged out over time as you correct the rate so it amounts to nothing to worry about.

This analysis ignores other errors with the mechanical coupling causing other non-linearity in the incremental rate measurement, but the total measured random rate error for my EQ6 mount,including all the mechanical, electrical, and other errors involved, was about +/- 0.7 arc-seconds/sec (2 sigma, 95%) which is equivalent to an RMS error of 0.35 arc-seconds.

This is proven with the results that I have obtained when using the TDM. The paper I wrote back in 2011 identified a systematic error in the TDM which has been corrected since.
If there is another way to analyze the behavior of using an incremental rate encoder, then I would welcome your analysis.

Thanks!