VintageBigBlue.org

 

Re: PEC training question


Sep 28, 2001

 


----------------------------

#6195 Sep 28, 2001

What is the recommended frequency of training the PEC on a G11 with

Gemini? I have been training it once a night for CCD work. I usually

remain on the same side of the meridian for the whole night.



It occurs to me that if I use goto or slewing, I'll wind up on a very

different portion of the gear than where the PEC was trained. Does

that mean I should re-train the PEC after each slew? I guess the

answer depends on where the periodic errors get generated, the gear

or the worm.



How often do you train your PEC?



TIA,



-Paul



----------------------------

#6214 Sep 29, 2001

I am assuming gemini has its own PEC and that you don't even use the

old G11 panel with gemini upgrade?



because, with base G11, the PEC locks out the slewing rates. If you

turn it off to slew, then you need to retrain it (bummer)



But IMO, if you slew to different part of sky, you should probably

retrain. Secondly, If you slew, seems to me you would want to. My

thinking may be all wet here, but PEC I believe is time based, and if

you slew your meshing the gears faster than sidereal so the "program"

would then be off slightly. The purpose of the PEC is to program out

the bumps in the road. If the "program" races ahead when slewing so

it is still slewing out the errors in gear, I suppose you would not

need to retrain. I am hoping that the program actually takes slewing

and drive correction into account though, because I am now thinking

that even over time since your still speeding up and slowing down

your drive with corrections, then you eventually would produce the

same upset in orientation of time versus position on ra gear.



Maybe somebody else that may really know something about it can chime

in.



Kevin

--- In Losmandy_users@y..., "Paul K" pkane@h...> wrote:

> What is the recommended frequency of training the PEC on a G11 with

> Gemini? I have been training it once a night for CCD work. I

usually

> remain on the same side of the meridian for the whole night.

>

> It occurs to me that if I use goto or slewing, I'll wind up on a

very

> different portion of the gear than where the PEC was trained. Does

> that mean I should re-train the PEC after each slew? I guess the

> answer depends on where the periodic errors get generated, the gear

> or the worm.

>

> How often do you train your PEC?

>

> TIA,

>

> -Paul



----------------------------

#32722 Apr 20, 2007

I tried training my G11 with my auto guider, and the results were

absolutely unusable!? I must be doing something wrong....

After the training session and with the PEC enabled, I put it on a star

with the autoguider working as well. The autoguider absolutely could

not keep up with where the PEC was trying to make the scope go. And

going WAY off target is where it wanted to go. It was as though there

were nothing, no PEC or autoguiding.

Here's how I set up the training session. On a clear reasonably steady

night, I picked a bright star about 55 deg above the horizon, one that

was plenty bright with 0.25 sec exposures. I set my autoguider (an SBIG

STV) to average 3 such exposures and make a correction every 0.75 sec.

A recording of the session showed a very smooth track, less than +/-

0.6 arc second avg deviation. However, as mentioned, the results

thereafter when trying to track with the PEC enabled, and either with

or with out autoguiding assistance was unusable. It moved with a

determined will of it's own to +/- 5 r 6 arc sec deviation. Nearly 10

times worse than a good non PEC assisted simple autoguiding session.

The reason I'd chosen the 0.75 effecive correction interval, was

because I'd read that the Gemini software corrects approximately every

0.5 second, so it should be able to easily handle my selected rate of

correction. Apparently not. So,... help! How do you guys train your PEC

with your autoguider? And can anyone tell me why this didn't work?



Many thanks in advance



Keith

willowberry.net/keithnk%5Fm42/



----------------------------

#32726 Apr 21, 2007

Keith, are you using Gemini? If so, what level/version? Anything

prior to Level 4 had a bug that made PEC unusable.



If you are not using Gemini (or using Level 4), how good is your

polar alignment? If not very good, the PEC will record a significant

drift that will be constantly applied to tracking. This can lead to

quick guide star run-aways.



When recording PEC, turn off DEC corrections -- these can cause

large errors in RA.



Try longer integration intervals with STV: too short an exposure

will be affected by seeing fluctuations, and you'll be recording

these into PEC. Try 2 to 5 seconds integrations. If this doesn't

correct PE enough and you are using Gemini, you can always record

another run, and then average it with the previous one to improve it.



Regards,



-Paul





--- In Losmandy_users@yahoogroups.com, "keithnkm42" keithnk@...>

wrote: >

> I tried training my G11 with my auto guider, and the results were

> absolutely unusable!? I must be doing something wrong....

> After the training session and with the PEC enabled, I put it on a

star > with the autoguider working as well. The autoguider absolutely

could > not keep up with where the PEC was trying to make the scope go.

And > going WAY off target is where it wanted to go. It was as though

there > were nothing, no PEC or autoguiding.

> Here's how I set up the training session. On a clear reasonably

steady > night, I picked a bright star about 55 deg above the horizon, one

that > was plenty bright with 0.25 sec exposures. I set my autoguider (an

SBIG > STV) to average 3 such exposures and make a correction every 0.75

sec. > A recording of the session showed a very smooth track, less than

+/- > 0.6 arc second avg deviation. However, as mentioned, the results

> thereafter when trying to track with the PEC enabled, and either

with > or with out autoguiding assistance was unusable. It moved with a

> determined will of it's own to +/- 5 r 6 arc sec deviation. Nearly

10 > times worse than a good non PEC assisted simple autoguiding

session. > The reason I'd chosen the 0.75 effecive correction interval, was

> because I'd read that the Gemini software corrects approximately

every > 0.5 second, so it should be able to easily handle my selected rate

of > correction. Apparently not. So,... help! How do you guys train

your PEC > with your autoguider? And can anyone tell me why this didn't work?

>

> Many thanks in advance

>

> Keith

> willowberry.net/keithnk%5Fm42/

>







----------------------------

#32732 Apr 20, 2007

Keith,



Your posting raises lots of questions, but perhaps the best I can tell you

is to look into a piece of software called PEMPro. It has a free trial

period of 30-60 days, I forget which, and will do an excellent job with your

Gemini/G11.



One thing to keep in mind with your approach is that the autoguider, by

definition, is always out of phase with the PE. It is a reactive process.

Also, depending on how you set up its software, it can be overaggressively

correcting, or under correcting. It reacts to seeing variations, especially

with the brief interval you have set up. If you have it correcting out the

entire variability, you're almost certainly chasing the seeing - after the

fact. So, when you have Gemini "play back" the PEC, it's almost certainly

both out of phase with the real PE, as well as not reflective of the real

PE, but rather dominated by seeing variations. Those issues and possibly

many others, would argue against your general approach.



The approach PEMPro uses, and you could use if you wished to write some

software, is to NOT attempt to correct for errors. Just let the mount do

its thing. But record the results. That will give you a PE curve with lots

of noise on top of it looking like grass against the real PE. Taking

multiple cycles of PE recording would give you a statistical means to smooth

the PE curve to discover its real nature, save it to Gemini, get it in phase

with the G11 worm, and enable PEC. That's all you'd have to do.



Or, just use a visual off-axis guider and make corrections by hand. The

human eye/brain has an incredible ability to know when it's observing seeing

variations and when a correction is needed. You'd probably do better than

trying to record an autoguide session, which has no intelligence built in to

distinguish PE variation from seeing variation.



Sorry this is long, but I'm trying to help, not criticize here and I'm a bit

short on time, typing like mad :-) Check out PEMPro. Even if you don't

bother to download it, it's excellent reading on the issues involved, etc.



Clear skies Keith,



Gary Ferdinand

Corvus Observatory

North Chatham, NY





> -----Original Message-----

> From: Losmandy_users@yahoogroups.com

> [mailto:Losmandy_users@yahoogroups.com]On Behalf Of keithnkm42

> Sent: Friday, April 20, 2007 5:04 PM

> To: Losmandy_users@yahoogroups.com

> Subject: [Losmandy_users] PEC training question

>

>

> I tried training my G11 with my auto guider, and the results were

> absolutely unusable!? I must be doing something wrong....

> After the training session and with the PEC enabled, I put it on a star

> with the autoguider working as well. The autoguider absolutely could

> not keep up with where the PEC was trying to make the scope go. And

> going WAY off target is where it wanted to go. It was as though there

> were nothing, no PEC or autoguiding.

> Here's how I set up the training session. On a clear reasonably steady

> night, I picked a bright star about 55 deg above the horizon, one that

> was plenty bright with 0.25 sec exposures. I set my autoguider (an SBIG

> STV) to average 3 such exposures and make a correction every 0.75 sec.

> A recording of the session showed a very smooth track, less than +/-

> 0.6 arc second avg deviation. However, as mentioned, the results

> thereafter when trying to track with the PEC enabled, and either with

> or with out autoguiding assistance was unusable. It moved with a

> determined will of it's own to +/- 5 r 6 arc sec deviation. Nearly 10

> times worse than a good non PEC assisted simple autoguiding session.

> The reason I'd chosen the 0.75 effecive correction interval, was

> because I'd read that the Gemini software corrects approximately every

> 0.5 second, so it should be able to easily handle my selected rate of

> correction. Apparently not. So,... help! How do you guys train your PEC

> with your autoguider? And can anyone tell me why this didn't work?

>

> Many thanks in advance

>

> Keith

> willowberry.net/keithnk%5Fm42/

>

>

>

>

> Yahoo! Groups Links

>

>

>

>



----------------------------

#32746 Apr 22, 2007

Hi Paul, I am using Gemini PEC of level 4. It came as a pkg with my

new mount. My polar alignment is not perfect, but fairly good. Stars

stay "on line" for the several minutes that the worm gear takes to

turn during drift alignment for both alt. and azm. I did not turn off

the DEC correction, so I will do that. I'll also try the longer

integrations you suggest. Thank you,



Keith

www.willowberry.net/keithnk_m42/

--- In Losmandy_users@yahoogroups.com, "Paul K" pkane2001@...> wrote:

>

> Keith, are you using Gemini? If so, what level/version? Anything

> prior to Level 4 had a bug that made PEC unusable.

>

> If you are not using Gemini (or using Level 4), how good is your

> polar alignment? If not very good, the PEC will record a

significant

> drift that will be constantly applied to tracking. This can lead to

> quick guide star run-aways.

>

> When recording PEC, turn off DEC corrections -- these can cause

> large errors in RA.

>

> Try longer integration intervals with STV: too short an exposure

> will be affected by seeing fluctuations, and you'll be recording

> these into PEC. Try 2 to 5 seconds integrations. If this doesn't

> correct PE enough and you are using Gemini, you can always record

> another run, and then average it with the previous one to improve

it.

>

> Regards,

>

> -Paul

>

>

> --- In Losmandy_users@yahoogroups.com, "keithnkm42" keithnk@>

> wrote:

> >

> > I tried training my G11 with my auto guider, and the results were

> > absolutely unusable!? I must be doing something wrong....

> > After the training session and with the PEC enabled, I put it on

a

> star

> > with the autoguider working as well. The autoguider absolutely

> could

> > not keep up with where the PEC was trying to make the scope go.

> And

> > going WAY off target is where it wanted to go. It was as though

> there

> > were nothing, no PEC or autoguiding.

> > Here's how I set up the training session. On a clear reasonably

> steady

> > night, I picked a bright star about 55 deg above the horizon, one

> that

> > was plenty bright with 0.25 sec exposures. I set my autoguider

(an

> SBIG

> > STV) to average 3 such exposures and make a correction every 0.75

> sec.

> > A recording of the session showed a very smooth track, less than

> +/-

> > 0.6 arc second avg deviation. However, as mentioned, the results

> > thereafter when trying to track with the PEC enabled, and either

> with

> > or with out autoguiding assistance was unusable. It moved with a

> > determined will of it's own to +/- 5 r 6 arc sec deviation.

Nearly

> 10

> > times worse than a good non PEC assisted simple autoguiding

> session.

> > The reason I'd chosen the 0.75 effecive correction interval, was

> > because I'd read that the Gemini software corrects approximately

> every

> > 0.5 second, so it should be able to easily handle my selected

rate

> of

> > correction. Apparently not. So,... help! How do you guys train

> your PEC

> > with your autoguider? And can anyone tell me why this didn't work?

> >

> > Many thanks in advance

> >

> > Keith

> > willowberry.net/keithnk%5Fm42/

> >

>







Contact Us
This Site's Privacy Policy
Google's privacy policies

S
e
n
i
o
r
T
u
b
e
.
o
r
g