Guiding help please

[Leebert]

Hi All,

I've been attempting guided photography and have been trying to perfect my set up. Recently, a friend lent me a Meade DSI for use as a guide camera as I was having trouble finding stars with my SPC900. Last night, I put the scope in the garden, did a careful polar alignment, slewed around to roughly Cygnus (using cdc) and started phd calibrating. It calibrated okay (but it took ages - 10 to 15 mins).

I then took some subs starting with 5 minutes but they were probably worse than unguided! Any pointers to what I may be doing wrong would be greatly appreciated.

The set up is as follows:

HEQ5 pro;

Skywatcher 200p;

Meade DSI I guide camera mounted on 200p finderscope with homemade adapter.

The sky was completely clear when the images were taken and wind calm.

Many thanks,

Lee

[Leebert]

Apologies. Laptop desktop image of spider cropped out.

[malc-c]

Guiding is something of a dark art... one day it goes really well and the next it's all over the place. I'm no expert but the problem could be related to the PHD settings. The RA aggressiveness seems low, I would suggest starting that at the default of 100% and possibly increase it to no more than 110%. Also make the minimum motion smaller. For my ST80 I have it set at 0.15 when using a QHY5, but for your camera you may need to experiment with the value so that calibration takes less time.

Hope that helps

[Ant]

There could be some flex between finder scope and main scope.

I would change the guide exposure to 2 seconds, this helps the signal of the guide star and also helps counter atmospherics.

Also what I do is allow PHD to colibrate and start guiding, then going into EQMod and adjust the ASCOM PULSE GUIDE SETTINGS for both ra and dec down to about 0.6.

Could also be the scope settling down, often the first few minutes after starting the guiding is poor, then it improves markedly.

Cheers

Ant

[terryf]

I think the probem is in your calibration step size. PHD uses this to take a step during calibration (turn motors off or speed them up depending on direction) and if it does not see a big enough step in terms of pixels on the guide cam chip, then it can struggle with calibration. If that setting was OK previously, it may be that the new guide cam has bigger pixels and so needs a longer step during calibration. You should also check your RA and DEC rates if you are using EQMod as they should be set to 1.0 during calibration if I recall correctly otherwise they'll scale back the steps even further I think.

Terry

[chrisshillito]

If calibration is completing then the calibration step setting is fine. Increasing the calibration step will simply result in a fewer number of longer steps being used. The time taken for calibration to complete is dependent on the guide rate being used, but you shouldn't be choosing a guide rate just to get quick calibration. What rates are you currently using? Most folks seem to settle for around 0.5 x sidereal.

There is no reuirement to set the guide rates to 1.0 for calibration. I beleive what Terry was refering to is the gain sliders that are part of the EQASCOM pulse guide monitor rather than the guide rates themselves.

Another thing to check is that the DSI is mounted squarely on the mount so that a movement in RA corresponds to a movement in the X direction of the image. Whilst the webcam had square pixels the DSI may not and this can mess with PHD's measurement of angles - get ir orthogonally aligned and this won't be an issue.

Chris.

[riklaunim]

Max RA and DEC durations are to high. Quite likely PHD is overshooting the guide point back and forth by using to long guide commands. Set 150-200 ms.

[Leebert]

Thank you very much for the comments and pointers. I will try these once the skies are clear again.

The DSI camera has rectangular pixels of a ratio of 1.28:1. The camera was square to the rear of the finderscope but it was approx 30-45 degrees rotated with respect to RA and Dec.

The EQMOD ASCOM Pulseguide settings are 0.10 for both. I have attached a screenshot of the settings in case there are any other settings that need tweaking. To my knowledge, they are all default.

Many thanks,

Lee

[Ant]

The only thing that looks wrong to me I've circled in green

At the start (for calibration in PDH) these I would have all the way to the right (0.9). Once calibrated and guiding I'd lower to around 0.6.

Cheers

Ant

[chrisshillito]

I would agree that 0.1 is usually too low for a guide rate. Set it to 0.5 and see how it goes - your calibration will certainly speed up by a factor of 5.

I can't agree with Ant's suggestion of using different rates for calibration and active guiding. The whole purpose of calibration is for PHD to learn how the mount responds to a known set of pulses. By doing this it knows how long a correction pulse to send for a given error. If you calibrate at a high rate and the change to a lower rate when guiding then the corrections PHD issues will not produce the intended amount of corrective movement.

Chris.

[Ant]

I was always told that this helps stop the back and forth that you can get by sending the full correction. The RA aggressiveness which is always what I thought would do that, never seemed to make any difference for me.

If there isn't any advantage to making an adjustment - I have to wonder why they are adjustable.

As always the advice that I give is what works for me, I'm fairly happy with what I do as I can have round stars in 40 minute subs. Which hopefully I'll be able to do again if it ever stops raining...

Cheers

Ant

[chrisshillito]

From a general contol theory point of view a critically tuned reponse (one that reaches a steady state in the shortest time) will produce overshoot/undershoot that rapidly diminish in amplitude. I guess that by lowering the guide rate after calibration you are tricking PHD into a more overdamped response than is its default. If that is what gives you the best results then fair enough.

The EQASCOM guide rates are changeable so that you can set the responsiveness of the system. If you have a periodic error, or linear drift error that is moving at a rate quicker than your chosen guide rate then the mount is never going to catch up. Conversly if the guide rate is too fast then small errors can only be corrected by using very small pulse durations which may not be implemented accurately.

Within the eqascom pulseguide monitor area are 'gain' sliders that there for the purpose of modifying the 'strength' of guiding input. These sliders do not affect the guide rate but rather act to increase/decrease the guiding pulse duration that is applied.

Chris.

[terryf]

Chris is absolutely correct, I was referring to the gain sliders, as 15-20 mins for the calibration step seemed very long - where I live, that's enough for a clear sky to turn into horizon-to-horizon cloud!

I'm not sure that the rectangularity of the pixells is an issue though as PHD works out an angle so that it can make adjustments to the RA and DEC steps, but that angle is then applied to any movement of the centroid on the basis that the pixels are square, so I believe that this would be self-compensating - the angle will be wrong, but the corrections should be calculated correctly I think. Clearly it makes more sense to ensure that the guide camera is lined up with the mount, but I wouldn't get too fussy about it.

Terry

If calibration is completing then the calibration step setting is fine. Increasing the calibration step will simply result in a fewer number of longer steps being used. The time taken for calibration to complete is dependent on the guide rate being used, but you shouldn't be choosing a guide rate just to get quick calibration. What rates are you currently using? Most folks seem to settle for around 0.5 x sidereal.

There is no reuirement to set the guide rates to 1.0 for calibration. I beleive what Terry was refering to is the gain sliders that are part of the EQASCOM pulse guide monitor rather than the guide rates themselves.

Another thing to check is that the DSI is mounted squarely on the mount so that a movement in RA corresponds to a movement in the X direction of the image. Whilst the webcam had square pixels the DSI may not and this can mess with PHD's measurement of angles - get ir orthogonally aligned and this won't be an issue.

Chris.

[x6gas]

Max RA and DEC durations are to high. Quite likely PHD is overshooting the guide point back and forth by using to long guide commands. Set 150-200 ms.

Hmmmm. Interesting.

I struggled with PHD not being able to make large enough adjustments (I guess because my mount was poorly aligned in the first place) and so upped the duration of RA to 1000ms and Dec to 2000ms. If PHD is going well then it is making smaller adjustments then it doesn't need to apply such long corrections and I manage to control overshooting with the aggressiveness and hysteresis settings.

I did spend pretty much a whole session finding settings that work for me, but what I ended up with seems to work for me. I've seen much flatter graphs that I get but do manage sub pixel guiding...

FWIW, my settings:

RA aggress. 60

RA hyst. 20

Max RA 1000ms

Search region 20

Min motion 0.15

Calibration step 800 ms

Time lapse 0

Dec guide mode: auto

Dec algorithm: resist switch

Dec slope weight 5.00

Max Dec 2000

Star mass 0.5

Noise reduction: none

Dither scale 1.00

[Leebert]

Thanks very much for the replies and for the list of settings. I have a lot to try out now on the next clear night.

I managed to pick up a mono QHY5 with an ST4 cable so will be giving 'on camera' settings in PHD a try. I'm guessing that would take the EQMOD settings out of the equation too?

Many thanks,

Lee

[chrisshillito]

I managed to pick up a mono QHY5 with an ST4 cable so will be giving 'on camera' settings in PHD a try. I'm guessing that would take the EQMOD settings out of the equation too?

Switch to ST4 based guiding doesn't really affect the number of key parameters you must set. You still have to set the guide rates that will be applied during the guide pulses, although with ST4 there are fewer rates to choose from. What you will loose is the ability to monitor the guiding peformance from within EQASCOM and the ability to dynamically adjust the strength of guiding if you see something you don't like. You also need the extra cable.

Chris.

[JacobvonChorus]

I don't have any suggestions to improve guiding but I have to say that your guiding graph still blows mine out of the water! At 500mm, my many guiding errors don't show up (thank god). I will try to follow the advice on this thread though. Good luck.

Jacob

[Leebert]

I thought I'd update the thread. I used the QHY5 for the first time as a guide camera last night with a fresh install of PHD on the laptop. To cut a long story short, everything just worked. Even though I can't be sure which element was quite right, I now know that I can easily get 2 min subs with phd default settings which is a result for me.

Thank you all for your comments and suggestions