Re: Encoders in the Mach2 vs 1100


Roland Christen
 

The encoders are certified by the manufacturer. We also do extensive test of each mount using metrology to make sure that the mount performs up to our level. Takes a number of hours and is done on every mount.

Roland

-----Original Message-----
From: Seb@stro <sebastiendore1@...>
To: main@ap-gto.groups.io <main@ap-gto.groups.io>
Sent: Sat, Feb 27, 2021 2:09 pm
Subject: Re: [ap-gto] Encoders in the Mach2 vs 1100

All right, got it about the 24h period. I also know real life and ideal are two different worlds...

But then, as a manufacturer, how do you insure the mounts are accurate to that level before leaving the factory ? Is it based on some sort of empirical mathematics using the specs of the encoders, as well as mechanical, etc. or if you have an in-house standard testbench where every mount is tested for accuracy (or both maybe) ?

FYI, my questions are solely to satisfy my engineer's brain insatiable curiosity... Feel free to tell me, if you can't answer to protect some proprietary concept or something. I'll definitely respect that.

Regards,
Sébastien


De : main@ap-gto.groups.io <main@ap-gto.groups.io> de la part de Roland Christen via groups.io <chris1011@...>
Envoyé : 27 février 2021 13:41
À : main@ap-gto.groups.io <main@ap-gto.groups.io>
Objet : Re: [ap-gto] Encoders in the Mach2 vs 1100
 
Periodic error means that the error occurs periodically at some recurring rate. In the case of an encoder, the error being over a 24 hour period, you will never see it because the maximum tracking time will only ever be 12 hours or less (unless you wish to follow an object below the horizon). As such, then you will have only a partial period.

While the encoder can track accurately to the level in my previous post, the scope can never hope to match that due to a huge number of other factors:

1) atmospheric refraction which will vary between zero and as much as 200 arc sec per hour

2) telescope flexure which can be anything, but never zero. Typical flexure could be on the order of 1 to 10 arc sec per hour or more depending on instrument type.

3) polar misalignment which can be anything but never zero. It is impossible to get perfect polar alignment because of atmospheric refraction. You can get either alignment with the refracted pole or alignment with the actual pole. neither one of them will achieve zero tracking drift.

4) non-recurring movement of the scope/camera caused by various things like cable drag, shifting weight of attached peripherals, unbalanced loads, shifting mirrors, focuser shift, and many more.

So your answer in real life setups is unanswerable.

Rolando



-----Original Message-----
From: Seb@stro <sebastiendore1@...>
To: main@ap-gto.groups.io <main@ap-gto.groups.io>
Sent: Sat, Feb 27, 2021 12:22 pm
Subject: Re: [ap-gto] Encoders in the Mach2 vs 1100

 there is no real "periodic" error. The error tends to be a single smooth sine wave over a 24 hour period
Well, a sine wave is by nature periodic but I understand you're meaning that it's at a frequency irrelevant to a few minutes sub or a few seconds guide frame. Then over what period (1 hour, indefinitely, in-between ?) would the Mach2's "un-periodic" error stays under 0.25 arc-sec,  assuming a near-perfect PA, ideal sky conditions and for example a FL of 1400mm (1 a-s/pixel) setup ? Another way to put it would be to what "standard" conditions does the 0.25 a-s refers ?

My Mach2 being scheduled for shipment soon, I'm just trying to figure out what should reasonably be my expections in this regards. As well as trying to understand a bit more in-depth how the encoders works...  

BTW, thanks for the much-valued information you continuously share on this forum, Rolando. Definitely makes intermediate-newb like me want to know more about your products and keep learning on the matter.

Regards,
Sébastien


De : main@ap-gto.groups.io <main@ap-gto.groups.io> de la part de Roland Christen via groups.io <chris1011@...>
Envoyé : 26 février 2021 11:24
À : main@ap-gto.groups.io <main@ap-gto.groups.io>
Objet : Re: [ap-gto] Encoders in the Mach2 vs 1100
 
The error in an encoder mount is not periodic, so there is no real "periodic" error. The error tends to be a single smooth sine wave over a 24 hour period.

Rolando



-----Original Message-----
From: Seb@stro <sebastiendore1@...>
To: main@ap-gto.groups.io <main@ap-gto.groups.io>
Sent: Fri, Feb 26, 2021 8:22 am
Subject: Re: [ap-gto] Encoders in the Mach2 vs 1100

Thanks for the explanations Rolando. That clears it up a bit. Indeed resolution and accuracy are two different beasts...

What I was missing is the fact that the accuracy spec is for an entire revolution of the ring. I also had a somewhat hard time wrapping my head around the fact that the pitch of the ticks (or barcodes) is the same for any ring size (at 30 microns) and that its not giving any benefits in actual resolution (ticks per arc-sec). But reading further, I understood that the output resolution is not solely dependant on the mechanical aspect, but also on the encoding protocol used. 

So my understanding at this point is that even if the pitch is the same for every size of ring, hence giving more "barcodes" on larger rings, the readhead is nonetheless outputting a value that is limited by the characteristics of the serial protocol used (26 bits for the Mach 2, as you stated). 

I also understand that these figures of 0.16 and 0.12 arc-sec accuracy are theoretical and achievable only under hypothetical ideal conditions. 

From all this, should I also understand that the Mach2's stated native periodic error of 0.25 arc-sec (peak-to-peak) is also a per-hour figure or am I still missing some parts of the puzzle ? (Is periodic error the same as accuracy in this context with encoders ?)

Sébastien


De : main@ap-gto.groups.io <main@ap-gto.groups.io> de la part de Roland Christen via groups.io <chris1011@...>
Envoyé : 24 février 2021 19:41
À : main@ap-gto.groups.io <main@ap-gto.groups.io>
Objet : Re: [ap-gto] Encoders in the Mach2 vs 1100
 

I would have thought that a bigger ring gives more space to mark a higher number of "ticks", hence giving higher resolution and so accuracy...
Resolution is not the same as accuracy.

There are 2^26 individual addresses (67,108,864) that can be accessed in the RESA encoder, whether it is 75mm or 100mm. Therefore the resolution is the same for both rings.

The stated accuracy is a measure of how accurate the encoder is over a 24 hour period of revolution (360 degree total angle of rotation). For the 75mm ring, accuracy would be approximately +-3.82arcsec/24hr or 0.16 arc sec per hour of tracking. For the 100mm ring it would be 0.12 arc sec. per hour.

In reality one can never get the ring to have zero runout, so the practical accuracy will come in at perhaps +-0.5 arc sec per hour, more or less. So whether you have a 100mm ring or a 75mm ring, it makes no practical difference. Star motion due to atmospheric refraction will be an order of magnitude higher if you are anywhere but at the exact zenith.

The way the RESA works is not like any other encoder system. It does not read "ticks" the way an ordinary encoder does. It reads a barcode that is imprinted on the steel ring. The barcode is read over a fairly large circumference angle. The readhead is a miniature camera, not a photodiode that registers black and white tick marks. It's really quite revolutionary how it works, and it does work splendidly for telescope mounts.

Rolando




-----Original Message-----
From: Seb@stro <sebastiendore1@...>
To: main@ap-gto.groups.io <main@ap-gto.groups.io>
Sent: Wed, Feb 24, 2021 4:40 pm
Subject: Re: [ap-gto] Encoders in the Mach2 vs 1100

Hello Roland,

Your post made me take a quick look at Renishaw's spec for the Resolute extended temp encoder and I found two interesting observations (not related to the low temp version) which made me realize I'm probably missing something in my understanding of how the encoders actually works...

First, when we look at the following table, the "system accuracy" is increasing with the diameter of the ring (kind of opposite of what you stated earlier), which made sense to me since I would have thought that a bigger ring gives more space to mark a higher number of "ticks", hence giving higher resolution and so accuracy... So I assume the "system accuracy" of the encoder (which is defined as graduation + SDE by Renishaw) doesn't directly translate into the "tracking accuracy" of the mount.



Second observation, still looking at that table, the order of magnitude of that system accuracy seems to be more than ten-fold lower in comparison to the spec'ed tracking accuracy of the mount (+/- 3.82 arc-sec "system accuracy" for the 75 mm ring vs +/- 0.25 arc-sec "tracking accuracy" of the Mach2).

To explain these differences, my guess would be that some (gear/pulley) ratio somewhere does indeed make the tracking accuracy similar throughout the mount models and while at the same time increasing it by a factor of about 10X relative to the Renishaw's specs, but I wondered if there was more to it...

Am I lost in space ? 

Regards,
Sébastien


De : main@ap-gto.groups.io <main@ap-gto.groups.io> de la part de Roland Christen via groups.io <chris1011@...>
Envoyé : 24 février 2021 11:48
À : main@ap-gto.groups.io <main@ap-gto.groups.io>
Objet : [ap-gto] Encoders in the Mach2 vs 1100
 
Hello Astronuts,

To clear up any confusion about mount encoders, both Mach2 and 1100/1600 use the same Renishaw RESA high resolution encoders. The main difference is that the ring diameter of the 1100/1600 mounts is 100mm, the Mach2 uses a 75mm ring. Resolution and accuracy is the same for all. The readheads are the same RESA readheads, except that they are matched to their respective diameters, so that the 75mm readheads cannot be used on the 100mm rings and vice versa.

Roland Christen
Astro-Physics Inc.

--
Roland Christen
Astro-Physics

--
Roland Christen
Astro-Physics

--
Roland Christen
Astro-Physics

--
Roland Christen
Astro-Physics

--
Roland Christen
Astro-Physics

Join main@ap-gto.groups.io to automatically receive all group messages.