Astro-Physics Mach1GTO vs 10Micron GM1000HPS

[CCD Imager]

You are right Ollie, I've owned some premier mounts that have had issues. I guess the best is the one that gives the least problems!

Buzz - Maxim doesn't require an annual subscription unless you want to maintain updates after a year, I'm quite happy with my older version for now. Is it compulsory for the Bisque software?

Adrian

[perfrej]

Steve,

Glad to hear that the guiding works fine. What are the results without?

As for running a mount from 10Micron completely without a model, that is really not a supported mode of operation. The mount does not work like other mounts in that respect. A drift is quite normal without a model, and the mount doesn't bother to keep the tracking perfect unless is is a 10Micron GMx000QCI (not HPS). Note that you only need three points to have a model that will make the mount behave as any purely mechanical beast - the stuff we're all used to

This is not a normal mount with encoders added, even though that may be the impression one gets, especially considering the fact that the QCI mounts share the same mechanical design. Encoders are deeply embedded in the design and cannot be disabled. 10Micron found that the precision required to mount an encoder post production makes it unfeasible to do so and to make it an option.

Commercially available encoders will run about USD 3000 per axis. Bisque is continually pushing the launch date of their encoder solution - could this be because of said mechanical and retrofit difficulties? Me thinks so. 10Micron uses a proprietary encoder that is not available commercially. Astro-Physics do not yet have their software for encoders ready and they expect the encoder option to be a user-mounted option for retro-fit. Needless to say, I find that hard to digest, and I wonder why it is not shipping yet. Please interpret this information correctly; I do not work for nor do I have stock in 10Micron, I am just an encoder buff who accidentally tripped on a 10Micron mount and found it extremely good. Have two now...

There are three mounts out there that I know of that have played the encoder game correctly: ASA, Planewave and 10Micron. ASA uses Renishaw encoders with very high resolution as does Planewave. 10Micron, as I said, uses their own encoder specifically designed for mount use and they are, naturally, the only ones that have it. ASA's solution is really good as is 10Micron's. I have some doubts about Planewave but that is strictly from an uninitiated perspective.

In a 10Micron mount, everything the mount does is filtered through the model. In single axis tracking the mount uses a simple skew algorithm to correct the position, in dual axis a complex model algorithm is used to adjust everything to remove things like flexure, cone error, polar error, refraction and more.

As an encoder has a base offset that is calculated from the first three points of a model, the mount will not work correctly without these three points. To verify this, just look at the log from a Model Maker run and note the pointing errors of the first three points compared to the fourth and consecutive ones:

22:08:41 - Model Maker v3.44 (2014-01-31) Start-up22:08:41 - No refraction file k:\_weather\realtime.txt22:08:41 - Reference catalog is c:\gsc1122:09:40 - Telescope: 10micron GM2000HPS connected22:09:41 - Mount firmware is 2.12.522:09:41 - Mount coords are  J200022:09:43 - MaximDL connected22:15:49 - Scope reports as 10micron GM2000HPS22:15:49 - Deleting mount model22:15:49 - Model deleted. Waiting 4s...22:15:53 - Model deletion completed22:16:25 - Slewing to base point 1 at 162, 71.....22:16:40 -     Slew settle 5s (5.00)22:16:41 -     Exposing (5)...22:16:49 -     Solving...22:16:54 -         Ra  diff 68.54"22:16:54 -         Dec diff 447.87"22:16:54 -     Solved OK22:16:54 -     Syncing22:16:55 -     Synced OK22:17:05 - Slewing to base point 2 at 124, 73..22:17:19 -     Slew settle 5s (5.00)22:17:19 -     Exposing (5)...22:17:28 -     Solving...22:17:31 -         Ra  diff 71.54"22:17:31 -         Dec diff 459.58"22:17:32 -     Solved OK22:17:32 -     Syncing22:17:32 -     Synced OK22:17:42 - Slewing to base point 3 at 67, 76...22:17:56 -     Slew settle 5s (5.00)22:17:56 -     Exposing (5)...22:18:05 -     Solving...22:18:14 -         Ra  diff 80.38"22:18:18 -         Dec diff 451.60"22:18:18 -     Solved OK22:18:19 -     Syncing22:18:20 -     Synced OK22:18:31 - Slewing to refine point 1 at 13, 73..22:18:45 -     Slew settle 5s (5.00)22:18:45 -     Exposing (5)...22:18:54 -     Solving...22:19:01 -         Ra  diff 3.30"22:19:02 -         Dec diff 3.65"22:19:02 -     Solved OK22:19:03 -     Syncing22:19:04 -     Synced OK22:19:13 - Slewing to refine point 2 at 10, 63...22:19:56 -     Slew settle 5s (5.00)22:19:56 -     Exposing (5)...22:20:04 -     Solving...22:20:08 -         Ra  diff 0.73"22:20:08 -         Dec diff 1.52"22:20:08 -     Solved OK22:20:08 -     Syncing22:20:09 -     Synced OK

Note the quick "I found where I am" event after the first three points. How well would it work without a model? Not very, so you NEED the three alignment points. After that you can run guided forever and unguided for a bit. More points and you can run unguided forever. I have done five 1-hour subs in a row at FL 1000 mm.

In both modes, single axis and dual axis tracking, the motors themselves are controlled not only by measuring the motor movements (as is customary, but also by measuring the absolute position of the final axis. In this way, periodic error and gear imperfections are removed to a very high degree.

As for firmware versions, I have kept a full history of firmwares which means I can revert to whatever I want in order to make comparative tests. Not that I necessarily have the time, but still If you want to test the current production firmware - which has not been published on their site yet (I have reminded them and by now the slightly over-cooked Italian vacation period should be over) just PM and I'll provide a link for you.

You have the GM1000HPS, and even though all 10Micron mounts - even the QCI - share the same firmware package, the code inside the firmware handles the mounts differently. This means that some firmware revisions correct issues with a specific mount model, or motor version or whatever, but do not change anything in the handling of the other versions. This is all evident in the readme-file for the firmware.

MaximDL and lags, sluggishness etc is unfortunately a reality. I do not know why and I only have version 5, but my general take is that it is a very old-style application that could do a with a complete re-write (personal opinion). Even though it does show some multi-threading, my guess is that there are wait loops etc in the code, or possibly communication parts that run in the GUI thread or so. I have made the same mistakes myself... My Model Maker uses MaximDL for camera control and it is an odd interface that can lock up pretty badly at times. For 10Micron use, and specifically for the early versions of my ASCOM driver, a slow poll is required. I have since re-written parts of the driver to include some nifty intelligent caching that eliminates most of the problems. 10Micron's own ASCOM driver is a local server version which should be less sensitive to Maxim's behavior.

Nothing like a morning rant

All the best,

Per

p.s. If you have MaximDL and PinPoint, modeling with Model Maker is a very good way to do it - and fast. It is a free product that I wrote to improve my life (considerably) as I do not use eye-pieces or DSLRs with live view and run my stuff remotely (I am not a visual guy).

[buzz]

You are right Ollie, I've owned some premier mounts that have had issues. I guess the best is the one that gives the least problems!

Buzz - Maxim doesn't require an annual subscription unless you want to maintain updates after a year, I'm quite happy with my older version for now. Is it compulsory for the Bisque software?

Adrian

You are right - I should have been clearer, the subscription for Maxim and TSX is if you want updates and bug fixes (not upgrades) after a year.  Software Bisque's fee is about 20% of the full price. Both continue to function if you choose not to. In a year TSX has a lot more updates than Maxim. You can read that both ways but I noticed several well documented bugs that were not fixed in Maxim DL 5.

[buzz]

This is EXACTLY what I needed to know.  I'm afraid I need to scratch 10Micron off the list then.  I am going to take another look at the MX+ as I already am very familiar with The Sky X Pro.  In fact Bisque is doing a workshop at my clubs dark sky site soon.  Maybe there is still a spot open..

I would not discount 10 Micron. The issues that I related to from 2013/14 are now history. The distributer did take my unit back and the firmware has moved on since. Operationally, those with difficulties similar to mine have resolved their issues to the extent that my colleague, who had to wait several months for a GM2000 to be replaced has just bought an additional GM3000 for his observatory. He wouldn't sink that kind of money into something if it did not perform. 

I'm happy with the MX and today, I would be happy with a 10Micron too, especially if I had a permanent setup which could make the most of its accuracy.

[Physopto]

Just a point about the Software Bisque software costs and Maxim DL.

I bought a Paramount ME some years ago and it came with The SkyX etc. I checked up on their website as there was some mention of a reduction in the price of the "Universal Subscription" for newly purchased mount owners. The price was discounted from $1200 to $700. This gave access to nearly all their software bar a couple of items. If you pay $200 each year after that you get all updates and upgrades to the software. I have done that now for several years. You also get the daily updates/upgrades as they become available.

I also have Maxim DL and only upgrade as and when I think it worth while. I don't think Version 6 was worth it, as it said it was fully "multithreaded", it is not! It still stops tracking whilst downloading images. There is no way to stop that happening. The reason for this is stated as that the ccd chip is flushed to clear it of image retention. So the flush may affect the tracking ability. If so they should have allowed the imager to decide if that was needed.

Derek

[wolfman_4_ever]

6.09 is so buggy (the version that allows you to use a check mark to allow camera and filter wheel operation to be in seperate threads) that when you enable multi-threading, half the program does not work or crashes.

[buzz]

6.09 is so buggy (the version that allows you to use a check mark to allow camera and filter wheel operation to be in seperate threads) that when you enable multi-threading, half the program does not work or crashes.

I switched to SGP a year ago, at the same time I switched mounts. SGP is a breath of fresh air and although not perfect, it is very capable and fantastic value. The guys are responsive to user input. They implemented features that I suggested in 24 hours! They are closely coupled with the PHD2 guys and the two programs work very well together. The weakest part of the imaging chain is focusing, but there again, if Maxim's had been that good, Focusmax would not have had much following.