Re: PHD2 versus ASCOM driver


FYI in recent releases, PHD can recognize if you are using encoders, and changes the default algorithm to lowpass2 for precisely the reasons Roland outlined

when you use the new profile wizard, it will calculate the correct step sizes, but it will not add time between your frames.

Wayne you might want to try bin 1 on your ultrastar. those are already pretty big pixels. Sounds like you have the time courtesy of the mount's accuracy that you can increase your exposure time if need be (and reduce time delay between exposures if need be)


On Mon, Mar 9, 2020 at 11:52 PM Wayne Hixson via Groups.Io <> wrote:
Thanks Roland. Of your suggestions, I’ll I’ve done so far is add a 1500 ms delay between 3 second exposures and getting the best guiding I’ve ever seen by far as I am live stacking M94 in SharpCap.  About 0.5” RMS over at least an hour with only occasional tiny corrections. 0.32” RA and 0.34” Dec.  OAG with an Ultrastar binned 2x2, 130GTX with field flattener, or 871 mm fl. Love it!

PS I used the Polemaster about 4 hours ago and just checked, the little green Pole square is still overlapping the red axis target circle by about 33%. 

Enjoy your time in Hawaii and I am looking forwards to using the keypad for modeling!


On Mar 9, 2020, at 10:21 PM, uncarollo2 <chris1011@...> via Groups.Io <> wrote:

We have another clear night expected tonight before the rains return. Any suggestions on these issues specifically or settings for PHD2 more generally?
hello Wayne,
One setting I recommend for Encoder Mounts is to use a 1.5 to 2 second delay time between guide exposures. Click on the little brain in PHD2 and bring up the Advanced Setting/Camera Settings. Set the Time Lapse to 1500 to 2000. Your guide exposures can be set to anything from under 1 sec to 5 sec or more. I usually use 2 sec exposures unless the seeing is jumping around, in which case a longer exposure is used.
I normally use 1x Sidereal guide rate, but you can use .5x also if the seeing is good. There is no need to go slower to .25x.
Calibration steps are 300msec at 1x sidereal, 600msec at .5x.
I use either Lowpass 2 or Hysteresis. In Lowpass2 my settings are normally Aggressiveness 50 to 80% depending on seeing. Higher if good, lower if poor seeing. Min Move depends on seeing as reported in Guiding Assistant. Usually set to 0.2 pixels for my setup which has 2 arc sec per pixel. Max Duration is 2500msec. Both axes are set the same, no Backlash Compensation is used (set it to 0).
I have used all the other algorithms with success. Settings vary by scope focal length and seeing conditions. In all cases use a 1500 to 2000 Time Lapse in order to let the mount settle between guide exposures and not try to chase the seeing. An encoder mount will reverse Dec very accurately for even 1/4 arc sec move commands, but not in zero time. When the mount gets a very tiny Dec move command to reverse it will so in a controlled manner at a controlled rate. It is not instant, so the Time Lapse lets the Dec axis get there and settle before getting another small move command.
In a non-encoder mount there is a delay time, but also a crossover hysteresis where it may take 2, 3 or more small move commands before the axis actually reverses. This is different in an encoder mount where only 1 command is needed for full reversal, or any move deemed necessary by the software. For instance if the guide star is 1 arc sec off the centerline and Aggress is set to 100%, then the software sends a -1 arc sec move command, and the mount will move the axis exactly 1 arc sec, no more, no less.
In RA it's always instant because the belt tension (or gear mesh) is always on in the direction of movement. It simply slows down or speeds up. In Dec it's also instant if moving in the same direction each time, but takes some 100 +milliseconds when reversing.
Reversing Dec axis is like taking a Tylenol pill for a headache. You take 1 pill and wait for the headache to go away, and it will after a bit of time. If you don't wait and think I took one pill, still have the headache and take another and another until the headache goes away, you have overdosed yourself. The first one + time lapse delay was actually best for your health.
So for Dec we don't call it backlash since there isn't any when reversing. In encoder terms it's called Settling Time. There is even settling time for RA during dither recovery where the mount is commanded to move many arc seconds. For instance, if the dither is 15 arc seconds and the guide rate is 0.5x, it will take the mount approximately 2 seconds to get back to the zero line (0.5x = 7.5 arc sec per sec). If you have a guide rate set for 0.5 seconds with no Time Lapse delay dialed in, the mount will get numerous 15 arc second move commands and will indeed overshoot at the end.
Hope this helps.
-----Original Message-----
From: Wayne Hixson via Groups.Io <>
To: main <>
Sent: Mon, Mar 9, 2020 9:05 am
Subject: [ap-gto] PHD2 versus ASCOM driver

Had a really good night with the Mach2 under clear but moonlit skies. PHD2 ran reasonable well with default settings, but  ran into a couple issues. First, it warned me I should disable Synchronous Pulseguide in the driver, I couldn’t find that setting and not sure I want to disable even if I could!

Second I had several times where I got a warning message that mount wasn’t responding to pulseguide commands and I had to disconnect/reconnect to resume.

We have another clear night expected tonight before the rains return. Any suggestions on these issues specifically or settings for PHD2 more generally?



Brian Valente

Join to automatically receive all group messages.