Back to forums

PHD2 Guiding calibration numbers

Open PHD Guiding Project PHD2 Sky-Watcher EQ6-R Pro Acquisition Guiding Deep Sky
17 replies•691 views
RICHARD BREKNE avatar
RICHARD BREKNE avatar
Been reading some interesting stuff on ZWO's help forums about guiding for AM5… In particular the justification for using 0.5 sec. One thing that interested me is that guiding at 0.5 of sidereal means introducing 7.5 arc seconds of movement for every second of guiding pulse.  I immediately wondered if for a EQ6R a whole second pulse should ever be needed ? Seems like a lot of correction movement.  I could limit the max pulse length for either/both axis…. but was wondering what folks experience in this with the EQ6R is.  In the AM5 discussion on ZWO's page it is recommended by one highly respected fellow to limit max pulse to tops 200 ms. Default on PHD2 is 2500… big difference.
Thoughts ?
Helpful Engaging
Tony Gondola avatar
Tony Gondola avatar
Two different mount technologies here. Harmonic drive mounts need very frequent guiding corrections to work well, on the order of 0.5" as you mentioned. The EQ6r is a totally different beast. On my EQ6R is use 2.5 sec. guiding exposures and get 0.25" RMS on nights with decent seeing.
Helpful Concise
RICHARD BREKNE avatar
RICHARD BREKNE avatar
Tony Gondola:
Two different mount technologies here. Harmonic drive mounts need very frequent guiding corrections to work well, on the order of 0.5" as you mentioned. The EQ6r is a totally different beast. On my EQ6R is use 2.5 sec. guiding exposures and get 0.25" RMS on nights with decent seeing.

Not really what I was asking about. I was asking about the max pulse time.  I am aware of the differences between strain waves and EQ6 type mounts with regard to guide camera exposure times.
Tony Gondola avatar
Tony Gondola avatar
RICHARD BREKNE:
Tony Gondola:
Two different mount technologies here. Harmonic drive mounts need very frequent guiding corrections to work well, on the order of 0.5" as you mentioned. The EQ6r is a totally different beast. On my EQ6R is use 2.5 sec. guiding exposures and get 0.25" RMS on nights with decent seeing.

Not really what I was asking about. I was asking about the max pulse time.  I am aware of the differences between strain waves and EQ6 type mounts with regard to guide camera exposure times.

My pulse max lengths that result in the above numbers with the EQ6R are the defaults, 2500ms.
astroian avatar
astroian avatar
I am probably misunderstanding something here. Am I correct in thinking that .5s that you refer to is the time to capture data and therefore the rate at which corrections are made? In which case, surely the AM5 can track better than 7.5” over a 0.5s interval?

Cheers,
Ian
Well Written Respectful Engaging
Rainer Ots avatar
Rainer Ots avatar
Let's go to the official documentation of Max Duration ( https://openphdguiding.org/man/Advanced_settings.htm#Algorithms_Tab ) which in my opinion explains it very well:
You might reduce this below the default value if you want to avoid chasing a large deflection that could be caused by a spurious event (e.g. wind gust, hot pixel, etc.) .'  However, there is rarely any reason for changing it if you have protected yourself against hot pixels (Minimum star HFD).

If you have such problems - reduce it. If not - no need to change it.
Guillermo (Guy) Yanez avatar
Guillermo (Guy) Yanez avatar
RICHARD BREKNE:
Been reading some interesting stuff on ZWO's help forums about guiding for AM5... In particular the justification for using 0.5 sec. One thing that interested me is that guiding at 0.5 of sidereal means introducing 7.5 arc seconds of movement for every second of guiding pulse.  I immediately wondered if for a EQ6R a whole second pulse should ever be needed ? Seems like a lot of correction movement.  I could limit the max pulse length for either/both axis.... but was wondering what folks experience in this with the EQ6R is.  In the AM5 discussion on ZWO's page it is recommended by one highly respected fellow to limit max pulse to tops 200 ms. Default on PHD2 is 2500... big difference.
Thoughts ?

*This is an interesting topic. The harmonic drive will require very frequent and large pulses which in turn means, you do not want to keep the pulse for too long, so you can get it to settle before a new pulse is launched and avoid overcorrections. In the EQ mounts, you rather guide at much longer exposures (say, 3 seconds) and the pulses are way smaller, so you can keep the pulse going for much longer. In my case, I have a Losmandy GM8 mount and can go well beyond 4000ms (I do not have a belt drive as with the EQ6 but I guess it may have a similar behavior). BTW I guess this is why most will say that guiding is mandatory for a Harmonic mount but you may get decent unguided images with an equatorial mount (as long as you keep the subs reasonably short).
Helpful Insightful Respectful Engaging
RICHARD BREKNE avatar
RICHARD BREKNE avatar
I am probably misunderstanding something here. Am I correct in thinking that .5s that you refer to is the time to capture data and therefore the rate at which corrections are made? In which case, surely the AM5 can track better than 7.5” over a 0.5s interval?

Cheers,
Ian

The 0.5/7.5 relates to  the % of sidereal tracking speed that guiding is set to. Sidereal is 15 " sec so 0.5 x 15 is 7.5 arc".  All this relates to just how much of a max guide pulse you need for a given mount.  In strain wave mounts you need only a very short pulse because you dont need to move a whole 7.5 arc " every 0.5 seconds (which is how often you typically guide expose for).  If you set max pulse to 1000 ms thats a whole second, so you would be moving the mount a whole 7.5 arc sec... so with these strain wave mounts this max pulse gets typically set to a much lower time interval... say 200 ms so the actual movement of the mount is reduced.
My question goes to how much actual guide movement does a typical EQ6 actually need...and then what max guide pulse would correspond to that.
RICHARD BREKNE avatar
RICHARD BREKNE avatar
Guillermo (Guy) Yanez:
RICHARD BREKNE:
Been reading some interesting stuff on ZWO's help forums about guiding for AM5... In particular the justification for using 0.5 sec. One thing that interested me is that guiding at 0.5 of sidereal means introducing 7.5 arc seconds of movement for every second of guiding pulse.  I immediately wondered if for a EQ6R a whole second pulse should ever be needed ? Seems like a lot of correction movement.  I could limit the max pulse length for either/both axis.... but was wondering what folks experience in this with the EQ6R is.  In the AM5 discussion on ZWO's page it is recommended by one highly respected fellow to limit max pulse to tops 200 ms. Default on PHD2 is 2500... big difference.
Thoughts ?

*This is an interesting topic. The harmonic drive will require very frequent and large pulses which in turn means, you do not want to keep the pulse for too long, so you can get it to settle before a new pulse is launched and avoid overcorrections. In the EQ mounts, you rather guide at much longer exposures (say, 3 seconds) and the pulses are way smaller, so you can keep the pulse going for much longer. In my case, I have a Losmandy GM8 mount and can go well beyond 4000ms (I do not have a belt drive as with the EQ6 but I guess it may have a similar behavior). BTW I guess this is why most will say that guiding is mandatory for a Harmonic mount but you may get decent unguided images with an equatorial mount (as long as you keep the subs reasonably short).

Hmm... my understanding is that your guide speed setting determines the amount of movement the mount can move per whole second of pulse.  A 0.5 guide speed setting will then result in 7.5 arc " of movement.  (0.5 x sidereal rate, which is roughly 15, so 0.5 x 15 = 7.5) .  With 4000 ms you would get as much as 30 arc " of correction movement.   Seems like an awful lot yes ??.  Ok... if you get a wind gust or something you might get out close to that much for a couple of seconds... but the mount will come back on its own when the gust passes... at least this is the reasoning I am reading from this guru fellow on the zwo user forum site... Chen.  He has some very convincing arguments and maths to back up his thoughts.
Helpful
Brian Puhl avatar
Brian Puhl avatar
There's no real reason to change max guide pulse.    If for some reason you need more movement, up your guide speeds.  (PHD is actually pretty good at adapting to this without recalibration)   Currently I run 0.5x on RA and 0.9x on dec for my EQ6R Pro.   PHD doesn't like this when it calibrates because it thinks both should be the same speed, but its of zero consequence.  There are reasons specific to this mount for why I run those guide speeds, but I wont dive into that here.   I run typically 1 second updates, but if seeing gets bad, I'll slow it down to 2 seconds, rarely higher, any worse than that, I'll just park the scope because I don't want that quality of data.  EQ6 needs mount PEC to be trained if you want to guide at those slower speeds, IMO. 

If the wind is pushing you off by more than your image scale, I'd probably say you should park the scope instead of trying to image through it.   The data will be horrible.   If it's just the occasional gust, then I would not change your settings to accommodate it.   Cull the sub and move on.   As for 'correcting faster', this is also a reason why I choose faster updates.    If you run 3 second updates, you get one, potentially major correction every 3 seconds.   With 1-2 second, I can get alot more, smaller corrections in time.    Now keep in mind, these are all still fully dependent on seeing conditions.    If you have wind, AND bad seeing, I'd park the scope.

Just to echo previous comments, comparing anything to the harmonics is apples and oranges.  Their periodic error is so high that they cannot perform without guiding at all, save for extremely wide imaging scales.

Helpful Insightful
RICHARD BREKNE avatar
RICHARD BREKNE avatar
Have you read this from the ZWO user forum ? Would be interested to get your response to his reasoning.  I understand that strain mounts and EQ6 like mounts are not comparable in most regards related to this... but the general idea that a max guide pulse limit with regards to how much any given mount needs to move to deal with corrections seems sound to me.  Check it out and let me know what you think of this guys reasoning.


https://bbs.zwoastro.com/d/15711-very-large-backlash-in-new-am5/13
apennine104 avatar
apennine104 avatar
As an AM5 owner, I have read lots of suggestions to lower the pulse duration to no greater than the exposure length. For example, when guiding with 0.5s exposures, change the durations to 500ms. At least with my AM5, I noticed no difference from the default 2500ms durations. What has worked best for me is 1s exposures and the default 2500ms setting. When I review the logs, I rarely see that PHD2 commanded a move longer than 1000ms anyway, even though it could go to 2500ms. PHD2 choses the duration it think will work best within the given range.

But, all those tweaks are on account of strain wave mounts more rapid error development. On traditional non strain wave mounts like the EQ6R, I believe longer exposures and default settings work best. I have a CEM40, and what works best for me is running the guiding assistant for ~10min on target, and than accepting it's exposure and MnMo settings it recommends. I have tried numerous times changing things to what I think will work, and it's just never better than the guiding assistant long term.
Helpful Insightful Respectful Engaging
Rainer Ots avatar
Rainer Ots avatar
RICHARD BREKNE:
Have you read this from the ZWO user forum ?

I read it a bit (most seems somewhat irrelevant to the discussion here). I understand that he thinks much smallers values "should still be ample" and is planning to write a totally new guiding system working differently than Asiair guiding / PHD. As I understand one of his motivations was that in PHD the max duration also controls pulses used for other purposes than guiding and he wants to spread out the pulse evenly throughout the exposure length.
This is not doable in PHD currently, that is why he wants to write something new.
So unless this system is actually ready (post is from 2 years ago so maybe?) and you are using it - you can't achieve what he is advocating.
Helpful
Brian Puhl avatar
Brian Puhl avatar
RICHARD BREKNE:
Have you read this from the ZWO user forum ?

https://bbs.zwoastro.com/d/15711-very-large-backlash-in-new-am5/13



I've read it, and it honestly doesn't apply here.  You can see my guide pulses in the screen shot I posted above.   Rarely do I ever see more than 100ms pulses at 0.5x rate.    Again, if you're trying to counter for wind, you're kinda wasting your time.  Especially in broadband, the second the guide star is pushed off, the damage to the sub has been done.    When you adjust things like aggression, PredictiveWT, ReactiveWt, you're asking PHD to only apply a percentage of what it thinks is proper.    This is producing the same kind of effect you're getting at by limiting to a percentage of the guide pulse, however PHD is still allowed to make those huge jumps, should it need to.     If you're attempting to make larger corrections to overcome backlash, I would simply keep your guide rate a bit higher, and PHD will figure out your backlash for you.

At the end of the day, your guiding is only as good as your seeing, and there are plenty of nights where I've had great guiding, but seeing is still poor.   I wouldn't overthink it.   If you're having issues guiding, myself and others here I'm sure are willing to help.
Helpful Insightful
Guillermo (Guy) Yanez avatar
Guillermo (Guy) Yanez avatar
RICHARD BREKNE:
Guillermo (Guy) Yanez:
RICHARD BREKNE:
Been reading some interesting stuff on ZWO's help forums about guiding for AM5... In particular the justification for using 0.5 sec. One thing that interested me is that guiding at 0.5 of sidereal means introducing 7.5 arc seconds of movement for every second of guiding pulse.  I immediately wondered if for a EQ6R a whole second pulse should ever be needed ? Seems like a lot of correction movement.  I could limit the max pulse length for either/both axis.... but was wondering what folks experience in this with the EQ6R is.  In the AM5 discussion on ZWO's page it is recommended by one highly respected fellow to limit max pulse to tops 200 ms. Default on PHD2 is 2500... big difference.
Thoughts ?

*This is an interesting topic. The harmonic drive will require very frequent and large pulses which in turn means, you do not want to keep the pulse for too long, so you can get it to settle before a new pulse is launched and avoid overcorrections. In the EQ mounts, you rather guide at much longer exposures (say, 3 seconds) and the pulses are way smaller, so you can keep the pulse going for much longer. In my case, I have a Losmandy GM8 mount and can go well beyond 4000ms (I do not have a belt drive as with the EQ6 but I guess it may have a similar behavior). BTW I guess this is why most will say that guiding is mandatory for a Harmonic mount but you may get decent unguided images with an equatorial mount (as long as you keep the subs reasonably short).

Hmm... my understanding is that your guide speed setting determines the amount of movement the mount can move per whole second of pulse.  A 0.5 guide speed setting will then result in 7.5 arc " of movement.  (0.5 x sidereal rate, which is roughly 15, so 0.5 x 15 = 7.5) .  With 4000 ms you would get as much as 30 arc " of correction movement.   Seems like an awful lot yes ??.  Ok... if you get a wind gust or something you might get out close to that much for a couple of seconds... but the mount will come back on its own when the gust passes... at least this is the reasoning I am reading from this guru fellow on the zwo user forum site... Chen.  He has some very convincing arguments and maths to back up his thoughts.

 

Your reasoning makes sense, but in certain cases, equatorial mounts may exhibit weird spikes due to poor mechanical engineering or imperfections in the gears. This is usually not a problem with harmonic drives., and hence, why the settings may differ between the two types of mounts (i.e., keeping pulses short in harmonic drives and longer in equatorial mounts)

To my understanding, the max pulse is—well—a maximum, acting as a cushion in case you experience those crazy spikes. The only way to recenter the object in the image is to maintain the guide pulse for an extended period, provided you do not exceed the exposure time for the next guide image. You may never need such an extended guide pulse, but it is there just in case. In my experience, I have encountered these issues with Celestron AVX and EQ5 mounts. Everything looks fine most of the time, and then all of a sudden, your target goes on a runaway path with no apparent external cause (such as wind).

If wind gusts interfere, then yes, you may get unnecessarily long guide pulses that will completely disrupt your guiding. In those situations, I would simply revert to a more conservative setting.
Helpful
RICHARD BREKNE avatar
RICHARD BREKNE avatar
As an AM5 owner, I have read lots of suggestions to lower the pulse duration to no greater than the exposure length. For example, when guiding with 0.5s exposures, change the durations to 500ms. At least with my AM5, I noticed no difference from the default 2500ms durations. What has worked best for me is 1s exposures and the default 2500ms setting. When I review the logs, I rarely see that PHD2 commanded a move longer than 1000ms anyway, even though it could go to 2500ms. PHD2 choses the duration it think will work best within the given range.

But, all those tweaks are on account of strain wave mounts more rapid error development. On traditional non strain wave mounts like the EQ6R, I believe longer exposures and default settings work best. I have a CEM40, and what works best for me is running the guiding assistant for ~10min on target, and than accepting it's exposure and MnMo settings it recommends. I have tried numerous times changing things to what I think will work, and it's just never better than the guiding assistant long term.

My experience so far as well.. Reasuring to hear others are not finding the 0.5 exposure to be the best. Thanks for the input.
RICHARD BREKNE avatar
RICHARD BREKNE avatar
Brian Puhl:
RICHARD BREKNE:
Have you read this from the ZWO user forum ?

https://bbs.zwoastro.com/d/15711-very-large-backlash-in-new-am5/13





At the end of the day, your guiding is only as good as your seeing, and there are plenty of nights where I've had great guiding, but seeing is still poor.   I wouldn't overthink it.   If you're having issues guiding, myself and others here I'm sure are willing to help.

One of my biggest problems is knowing when I am actually chasing stars in bad seeing, and when I am not.  I have this idea that worst case in clearish skies I should be able to keep at least around 1.0 rms. Helps to have two mounts going so I can see if guiding is similar or not.  But I remain unsure of just exactly what to do. Also,  I dont really understand the guide log stuff so much and cant utilize them to best extent. Then, there is that  about the only thing I see of the parameters in phd2 that obviously makes a big immediate change in guiding is the exposure speeds.  I cant see that changing max pulse makes any real difference unless you get so low that PHD2 cant make big enough changes at all. Here is where I have a bit of a problem accepting Chens explaination.... Max guide pulse doesnt mean that PHD2 will send a max pulse, it just means it can if something requires it. He seems to not think there is ever any reason to go more then a very short pulse, so it should be limited.
Thats the other thing... knowing what to believe from who... a fact of life I suppose one simply has to deal with.  I spent 3 nights so far chasing this 0.5 sek guide camera exposure thing and never got an improvement or even equally good guiding then when I use 1 sec.... and I am completely unsure of whether its something I am doing wrong in PHD2 or whether Chens advice is just bogus,  or perhaps even in my exact AM5 case the mount simply responds better with 1 sec.
Engaging
RICHARD BREKNE avatar
RICHARD BREKNE avatar
Was out trying this last night  and not having much success. Found that so far defaults 1.0 sec calibration worked best.  Ran into the following from Pegasus that gives a detailed recommenation of phd2 settings for calibration and guiding exactly for strain wave mounts.

https://pegasusastro.com/nyx-101-guiding-recommendations/

Had a go at Predictive PEC as per their instructions and found it promising. But seeing was not good at all and regardless what I tried there was a good deal of Texas throughout the night.  One thing I noticed, for those of you who use Predictive pec, is that if leaving the auto period update check box on, my initial 288 value for my newer AM5 steadliy got reduced over say 20 minutes of guiding all the way down to about 75, and then started to rise slowly again before I decided to stop and read some more before continuing with that algorithm. Dont really know how to interpret this behaviour yet... seemed like PHD2 was getting confused with my mounts particular period.