¿ªÔÆÌåÓý

ctrl + shift + ? for shortcuts
© 2025 Groups.io
  1. Gemini_users
  2. Topics

Catching up: Losmandy/ Gemini I and PEC

 

I'm catching up with twenty years of experience regarding PEC and Gemini I.? I've read Losmandy/ Gemini I doesn't play well with PEC because of the non-integer gear reduction in the stock Losmandy gear boxes.? Did that used to be true?? Is it still true?? Do the new gear boxes Losmandy sells have integer reductions, or are the McLennen gear boxes still useful mods?? Does PEMPro deal with non-integer reductions smoothly?? Does Gemini I support PEC programming using a 2x worm cycle, or is that for Gemini II only?
?
I'd rather not train my mount (go-to MI-250 w/ Gemini I) the old fashioned way, so I'm interested in changes and advances which have happened while I've been using eye pieces.
 

Hi John,
?
>>? I've read Losmandy/ Gemini I doesn't play well with PEC because of the non-integer gear reduction in the stock Losmandy gear boxes.??
?
That is incorrect. There's not even a remote possibility that gearboxes could produce a non-integer reduction. Here are some specs for gears and their ratios:
?
As far as I know, MI-250 gearboxes were similar, except driving in the reversed direction, so Gemini gear reduction setting was negative for these mounts.
?
I'm guessing that the non-integer reduction in this case really refers to the error introduced by poorly adjusted worm blocks, usually not seated straight. Same page by Mark Crossley addresses that, as well:
?
?
The error that this type of mis-adjustment produces often manifests itself as a 76-78 sec periodic error that can't be corrected by Gemini PEC since it doesn't repeat with every worm cycle (hence a non-integer ratio). This only becomes important for PEC if your mount really has this error (by far not everyone has it) and if you can't adjust it out by properly aligning the worm blocks. I'm not sure how the worm is held on an MI-250, but I assume similar problems could occur if orthogonality is not adjusted properly. Easy enough to check with a PEMPro PE recording.
?
>> Does Gemini I support PEC programming using a 2x worm cycle, or is that for Gemini II only?
?
That feature was added in Gemini-II. If you're still thinking about the problem where your PEC playback phase shifted between sessions, the worm-block error can't do that. I've used PEMPro and PEC on Gemini-1 for many years with good results. The only thing you may need to check is that PEC programming with PEMPro doesn't introduce a linear drift -- that was something that could happen with PEMPro + Gemini-1. But that also wouldn't account for phase shift between sessions.
?
Regards,
?
? -Paul
?
On Mon, Aug 19, 2024 at 06:49 AM, macdonjh wrote:

I'm catching up with twenty years of experience regarding PEC and Gemini I.? I've read Losmandy/ Gemini I doesn't play well with PEC because of the non-integer gear reduction in the stock Losmandy gear boxes.? Did that used to be true?? Is it still true?? Do the new gear boxes Losmandy sells have integer reductions, or are the McLennen gear boxes still useful mods?? Does PEMPro deal with non-integer reductions smoothly?? Does Gemini I support PEC programming using a 2x worm cycle, or is that for Gemini II only?
?
I'd rather not train my mount (go-to MI-250 w/ Gemini I) the old fashioned way, so I'm interested in changes and advances which have happened while I've been using eye pieces.
?
?
 

The reason you have a PEC sensor and robust software that looks across multiple periods is so you don't have to have an exact integer (as opposed to rational) number in a gearbox reduction.
?
Not to be too pedantic about it but I think you're both misusing the term "integer".
?
You can easily get a non-integer ratio: Take a 10 tooth gear and mesh it with a 7 tooth one. That produces a 10/7=1+3/7 = 1.428571 and so on ratio.
10/7 is a rational number, not an integer.
?
In this hypothetical example an algorithm that looks across 7 periods would pick up the subharmonic caused by the remainder of 3.
?
A 10 and 5 combo would produce an integer ratio: 10/5 = 2
?
And yes, proper alignment of the two-part worm blocks in older mounts removes a lot of the problem, which was not caused by gearing but by misalignment of the two separate worm blocks working against the balls in the bearings of the blocks. I, like others before me, clamp a piece of known flat angle iron to the blocks so they form a temporary one-block mounting while doing worm adjustments. After tightening the two blocks (but not overtightened - the metal is thin at the edges) in place remove the angle iron.
?
Best regards,
Mark Christensen
 

At least to my thinking about PEC, integral with the worm cycle is what's significant, not the actual gear ratio reduction. Any periodic error that repeats on a period that isn't evenly divisible into the worm cycle is non-integer and can't be fixed with a one-period PEC recording.
?
On Mon, Aug 19, 2024 at 10:59 AM, Mark Christensen wrote:

The reason you have a PEC sensor and robust software that looks across multiple periods is so you don't have to have an exact integer (as opposed to rational) number in a gearbox reduction.
?
Not to be too pedantic about it but I think you're both misusing the term "integer".
?
You can easily get a non-integer ratio: Take a 10 tooth gear and mesh it with a 7 tooth one. That produces a 10/7=1+3/7 = 1.428571 and so on ratio.
10/7 is a rational number, not an integer.
?
In this hypothetical example an algorithm that looks across 7 periods would pick up the subharmonic caused by the remainder of 3.
?
A 10 and 5 combo would produce an integer ratio: 10/5 = 2
?
And yes, proper alignment of the two-part worm blocks in older mounts removes a lot of the problem, which was not caused by gearing but by misalignment of the two separate worm blocks working against the balls in the bearings of the blocks. I, like others before me, clamp a piece of known flat angle iron to the blocks so they form a temporary one-block mounting while doing worm adjustments. After tightening the two blocks (but not overtightened - the metal is thin at the edges) in place remove the angle iron.
?
Best regards,
Mark Christensen
 

Thank you for your reply, Paul.? Also, I appreciate the link to Mark Crossley's webpage.
?
For the "non-integer gear reduction" I referred to, I meant the non-integer reductions between the individual gears, as Mark Crossley mentions in his webpage. As I understand it, the MI-250 was built with Losmandy motors and gear boxes.? I know my MI-250 has a current Losmandy motor for RA, and the gear boxes look very Losmandy-like.
?
I've read various accounts of misaligned worm blocks in Losmandy mounts built prior to the the OPW (my G11 is from 1999 and has separate worm blocks).? The worm bearing housing in my MI-250 is "in between" the original Losmandy design and the OPW design: the in-board and out-board beaning housings are bolted to a common piece of stock.? Variation in alignment of the worm bearings when everything is disassembled and reassembled should be minimal.? However, there is the possibility the three pieces aren't machined exactly square and parallel, thus introducing misalignment and the "pinch" Mark Crossley refers to.? Perhaps another similarity between the G11 and MI-250.? I understand your comment about PEC not being able to correct for the "76 second error".
I also remember all the discussions about the "76 second error", that was long before I was interested in photography and if my G11 has it, it didn't bother me at the eye piece.
?
Thank you for confirming the "2x worm cycle" is a feature in Gemini II.??
?
I understand that worm phase shifting between sessions is a separate issue from gear boxes, worm bearing block alignment and linear drift.
?
Since you bring up linear drift: I've seen a few posts about that.? None of the posts I've read mentioned if the drift manifested itself over several sub-frames or during a single frame; and in either case, how long it took for the drift to become noticeable.? Does the linear drift manifest itself as oblong stars, or does it take long enough it manifests itself as the target object slowly drifting off-center?? Something else?
?
 

I've posted a couple of photos of my MI-250 worm assembly, in case anyone is interested.

IMG_0911.JPG
IMG_0913.JPG

 

Hi John,
?
>> For the "non-integer gear reduction" I referred to, I meant the non-integer reductions between the individual gears, as Mark Crossley mentions in his webpage. As I understand it, the MI-250 was built with Losmandy motors and gear boxes.? I know my MI-250 has a current Losmandy motor for RA, and the gear boxes look very Losmandy-like.
?
As I said, the non-integer gear reduction has no direct effect on PEC. What does have an effect is that the error has a period that is an integer multiple of the worm cycle.
?
Your MI-250 appears to have a fully aligned, one piece worm assembly. It's unlikely that you'll experience the 76sec error that is reported on G11 mounts.
?
Before you start worrying 76s error or other uncorrectable ones, the simple test is to record 30-60 minutes of unguided star positions using short exposures with PEMPro or PHD2. That can be analyzed in software, so you can easily see which errors are most dominant and which can or cannot be corrected with Gemini PEC.
?
>> Since you bring up linear drift: I've seen a few posts about that.? None of the posts I've read mentioned if the drift manifested itself over several sub-frames or during a single frame; and in either case, how long it took for the drift to become noticeable.? Does the linear drift manifest itself as oblong stars, or does it take long enough it manifests itself as the target object slowly drifting off-center?? Something else?
?
The drift was not huge. It could be measured as the change in star position over a period of multiple worm cycles. Software tools make it easy to determine the amount of drift. From my experience with Gemini-1 and PEMPro, the drift was sometimes introduced after programming the curve and turning on PEC. It didn't happen every time, and Ray provided an option to remove the drift in PEMPro many years ago. This drift doesn't seem to occur with Gemini-II and PEMPro, in my experience.
?
Regards,
?
? ?-Paul
?
On Mon, Aug 19, 2024 at 10:27 PM, macdonjh wrote:

Thank you for your reply, Paul.? Also, I appreciate the link to Mark Crossley's webpage.
?
For the "non-integer gear reduction" I referred to, I meant the non-integer reductions between the individual gears, as Mark Crossley mentions in his webpage. As I understand it, the MI-250 was built with Losmandy motors and gear boxes.? I know my MI-250 has a current Losmandy motor for RA, and the gear boxes look very Losmandy-like.
?
I've read various accounts of misaligned worm blocks in Losmandy mounts built prior to the the OPW (my G11 is from 1999 and has separate worm blocks).? The worm bearing housing in my MI-250 is "in between" the original Losmandy design and the OPW design: the in-board and out-board beaning housings are bolted to a common piece of stock.? Variation in alignment of the worm bearings when everything is disassembled and reassembled should be minimal.? However, there is the possibility the three pieces aren't machined exactly square and parallel, thus introducing misalignment and the "pinch" Mark Crossley refers to.? Perhaps another similarity between the G11 and MI-250.? I understand your comment about PEC not being able to correct for the "76 second error".
I also remember all the discussions about the "76 second error", that was long before I was interested in photography and if my G11 has it, it didn't bother me at the eye piece.
?
Thank you for confirming the "2x worm cycle" is a feature in Gemini II.??
?
I understand that worm phase shifting between sessions is a separate issue from gear boxes, worm bearing block alignment and linear drift.
?
Since you bring up linear drift: I've seen a few posts about that.? None of the posts I've read mentioned if the drift manifested itself over several sub-frames or during a single frame; and in either case, how long it took for the drift to become noticeable.? Does the linear drift manifest itself as oblong stars, or does it take long enough it manifests itself as the target object slowly drifting off-center?? Something else?
?

1 - 7 of 7