AA5 Guiding

[malc-c]

Just managed to pop out for an hour to test some guiding since putting the HEQ5 back to gears rather than pulley drive.

I tried AA5 demo and got the attached image of M45. This is a single 600s exposure at 800ISO (resized to upload and small adgustment in curves to cut out the bright background). I've also attached the data from the tracking graph and EQMOD. Could someone let me know if the trace on the dec axis is a balance issue or a setting in AA5.

Doing a similar guide with PHD and the DEC was very smooth but RA all over the show... :confused:

[malc-c]

Anyone want to chip in here ??

[Komet]

The settings you have in AA are a bit weird. Basically you haven't been guiding at all in DEC ....

Through the letters ....

A .... Once AA calibrates sometimes I find DEC will drift off. I get this after a meridian flip and have to check the INV box so AA guides the correct way. You'll see it if it happens - easy to see and easy to remedy. AA I believe doesn't actually know which way is up when it calibrates so in future once you start guiding just make sure that the INV box is checked if need be. More on this later ....

B .... This box is set to that any movement UNDER 0.1 of a pixel is ignored and no corrections are made. Try lowering that a bit to 0.05. The figure you will end up at (shown in the left of of the guiding spor graph) might be a lot higher than that in which case you could raise it a bit in future but at least give it a chance to guide out small errors.

D .... (Sorry its not alphabetical). This serves the same function as B.

C .... Where you have screwed everything up. The figure shown in this box represents the limit of guiding. In this case once the guide star moves 0.5 pixels away from its original position AA will not guide. This is very useful when clouds drift through and dim or obscure the guide star. It means that the mount won't go chasing something and it works pretty well. Unfortunately with the setting you have there it means that the guide star has moved over half a pixel from its original setting and so AA has stopped. That value really needs to be a lot bigger. I think the recommended value is 4 - I use 2.

Now I think your DEC guiding needs to be inverted because if you look at the DEC graph you will see that it has motored away until it has hit the +1 pixel line where because this is what you have set is the limit of guiding (the value in Box C) AA has stopped shoving in corrections. DEC has then just sat there. Look at E: you can see that DEC guiding has drifted away from the centre and then DEC is just a clump outside of the 1 pixel line. RA is fine.

So ... you need to Check the INV box on DEC guiding. (A)

Decrease the RA limit ( to 0.05 to see how fine you can get it.

Increase the Drift box © to say 2.

This should make the guiding run quite happily. You can then fine tune it in realtime in EQMod by using the Pulse Guide Width Monitor.

Please come back with your results

[malc-c]

Mick,

Thanks for taking the time to do such a detailed reply. I've no idea where those values came from and I was under the impression that the auto calibrate button would of sorted those out.. still I'll print off your recommendations and will give them a try next session. Hopefully we'll get a gap in the clouds soon.

[malc-c]

Mick,

Managed to spend an hour between the clouds to try your suggestions. Hopefully I've interpreted what you've said correctly? I've also tried various other settings, but things still don't seem right.

I did notice in the second capture that the blue lines in EQMOD, only the RA was scanning across. The DEC remains fixed at the point shown in the screen grab. The traces seem to suggest that DEC seems to either stop guiding, then the star must move past some threshold which caused it to jump back ??

At the time of the test the wind was gusting to 24mph, although the observatory does offer some protection, there was enough wind to move the camera strap on the dSLR.

[Komet]

Make the DE box (Don't guide when) = 0.5

Using the settings you have there it is not going to guide at all in DEC until the star moves 2 and then 4 pixels!

In EQMod what is the guide rate setting you have?

[malc-c]

Mick,

Thanks once again for the input. I see... it was me being a prat - I was reading that as the upper limit for RA !

The settings in EQMOD were default - haven't changed any of the pulse guiding settings in that app.

I'll make the changes the next time out. I will be converting the HEQ5 back over to belt drive soon, so will see if that also helps. I guess the idea is to get all the dots on the graph clumped as tight to the middle as possible. What would be the limit where images become affected by bad guiding when using a QHY5 camera... one ring, two ?

Edit:

Would these values be OK ?

[Komet]

Firstly - read the Manual

Values ..... Try these ......

Don't Guide When ......

Drift < 0.5 RA 0.5 DE

Drift > 2 RA or DE

Guide rate in EQMod .... try 0.8 or 0.9x

This should allow you plenty of fine tuning in the Pulse Guide Width Monitor. This is all up for experimentation relative to your setup mind. This is what I get after the brief calibration and a bit of tuning on the fly in EQMod

[malc-c]

Mick, who reads the manual when we have the SGL forum to hand

Thanks once again, I'll give that a whirl, and seeing your graphs has given me something to aim for....

One last thing, in EQmod, do you do the tweekaing with the DEC/RA width gain sliders or the pulse guiding settings (rates) ?

[chrisshillito]

Any "tweaking" whilst guiding is active should be done via the gain sliders - this will effectively shorten or lengthen the width of the guiding pulses as apposed to changing the rate applied during the pulse.

PHD Calibration should always be performed with the sliders at 100%. You do need to read the EQMOD guiding manual and/or watch the video (about 11 minutes in) to understand how to interpret the EQMOD guiding graph - it works on different principles to PHD's

Chris.

[chrisshillito]

Mick, who reads the manual when we have the SGL forum to hand

Clearly those who spend their lives answering your questions on the SGL Forum! :)

Chris.

[Psychobilly]

Clearly those who spend their lives answering your questions on the SGL Forum!

Chris.

Classic

[ollypenrice]

Clearly those who spend their lives answering your questions on the SGL Forum! :)

Chris.

Beat me to it.

I like the AA graphics because unlike the usual line graphs (which it does also provide) the target of circles with the record of guide hits is great for interpreting certain trends. The classic is the double set of blobs, where there are two distinct centres of guide hits with a gap between them. This is a sure sign of backlash and can usually be fixed by running a little heavy on the east. Although very accurate, my Takahashi mount is sensitive in this regard.

As well as reading the manual I'd advise you to look at the little window with all those parameters and make sure you think through what each one of them means. This will allow you to see what the mount is and isn't doing.

It can be worth looking at which way the Dec error is being corrected (Dec+ or Dec-) and disabling the direction that is not needed. Genuine Dec corrections will always go the same way, the way determined by your polar misalignment. By disabling the other direction you reduce spurious input from the guider.

In bad seeing go for longer subs and increase the 'Don't guide when < ' value to avoid chasing the seeing. Note that there is no need to guide at less than the pixel scale at which you are imaging. You are nudging the mount for no good reason.

You'll get there but it's best to get your head round what the dialogue boxes are doing.

Olly

[malc-c]

Clearly those who spend their lives answering your questions on the SGL Forum! :)

Chris.

Ouch !

OK now I've binned the smoldering jacket I'm going to RTFM and watch some video

[malc-c]

Yesterday I fitted the 15t motor pulley to the mount and this evening managed to grab an hour to clear the data in EQMod and re-sync on a few stars, and try out some settings in AA5 to check the guiding...The results are attached.

The cloud was rolling in hence the dips, and I love the way AA came back on track when the star re-appeared. Could still do with playing with the sliders in EQmod, but I'm quite happy with the clustering and the traces... Looks like I'll be ditching PHD in favor of AA for guiding

Thanks Mick, Olly and Chris for your help. For some reason the downloaded manual won't open, but the guide notes pdf covered the basics... I tend to use APT for my dSLR captures, and Sharpcap for web cam work, but might just invest in the retail copy of AA5. Just need to locate a downloadable copy of the manual to understand this better before outlaying the cash !

[malc-c]

Guys, I've read some notes but would like some suggestions as to what's happening with the guiding based on the attached graph.

I use the auto-calibration option and have EQMOD set to 100% on the sliders, but whilst the hits all tend to be within the 1st circle of the target, the RA does seem to be tracing out a sine-wave !

Mount is belt drive using 4:1 ratio

[Komet]

What are your settings?

Try reducing the PulseGuide Width monitor to 50% in RA and see what happens.

Remember to Clear the data you have on the screen there

[malc-c]

Hi Mick,

Settings that the auto calibration came up with are attached. I'll try dropping the slider to 50% as suggested

[malc-c]

Uhmmm, 50% seems to of made it worse

[malc-c]

I've played a bit more with the sliders - I can't work out if it's better to reduce the percentage or increase it to keep the hits a clustered as possible ?

[Komet]

I thought the RA mod would screw it. 8-)

You don't have much movement there at 0.8 pixel movement. What is your guide speed in EQMod? I would increase it so you get more movement. On my setup I use 0.8x and fine tune on the PulseWidth.

Remember to Recalibrate and set the PulseGuide Width to 100% before you do.

.... and uncheck that Bad Seeing Guide.

[malc-c]

I thought the RA mod would screw it. 8-)

LOL - !!

You don't have much movement there at 0.8 pixel movement. What is your guide speed in EQMod? I would increase it so you get more movement. On my setup I use 0.8x and fine tune on the PulseWidth.

Remember to Recalibrate and set the PulseGuide Width to 100% before you do.

.... and uncheck that Bad Seeing Guide.

The clouds have rolled in so the Observatory has been closed for the evening... but I think the guide speed was default x0.10 for both (but would need to check) I'll make the changes suggested next session, and thanks again for your continued assistance

[Komet]

If it is 0.1x then try it at 0.5x first.

[malc-c]

Well I managed a couple of hours to hone the guiding and so far managed my best AA track and also my best ever PHD trace

The problem I had with AA5 is that the demo times out on CCD capture after 30 mins, and having nebulosity2, can't warrant spending a further £150 on an application that I'll only use a fraction of. It's a shame the developers don't market the guiding part of the AA5 as a standalone application.

Anyway, here's the screen dumps

[Komet]

It's a shame the developers don't market the guiding part of the AA5 as a standalone application.

Email them and suggest it. Its a very good system.