Another question about guiding...

I just want to put this guiding question out there and see what everyone thinks.

Back in the day when guiding was done manually (oh boy) it was normal to have a large and very precise worm and wheel on the RA and not much on DEC, maybe a tangent arm with a fine manual adjustment, or, if you were lucky, a dc motor. With a well aligned mount all of the action was in RA with only an occasional touch up on the DEC. These days it seems very different. When I look at published guiding graphs, it seems that the system is working just as hard in DEC as it is in RA and that's confusing to me. If you are polar aligned to within 1 arc/min of the pole (something that's very possible to do with modern software assistance)  and pointed at a star at a declination of 0 (worst case) the guide star will drift in DEC at a rate of approximately 0.3 arc/sec per. min. yet the guider still bangs away at it with almost as many corrections as in RA. I don't understand why DEC isn't just a slow drift with corrections every min. or so. Is the software chasing the seeing (I know small aperture guide scopes make that more likely) or am I missing something? Wouldn't be the first time ;)

Generally I think you are correct. If you're polar aligned reasonably well, the Dec shouldn't drift that much. My guider barely corrects for Dec on my setup, almost all of the corrections are done in RA.

Dec is going to work harder (well ra too) , if your mount isn’t perfectly level.  When my dec starts working hard with a good polar alignment, I check level and it’s usually off.  Don’t trust the little circular bubbles on the mount/tripod.

I’m not sure I understand…

let’s remove variables that we can’t control, like seeing and p.e. 

assuming perfect level, and perfect polar alignment, a mount only needs to move in RA. DEC won’t need to move at all. 

if the mount is canted in a random direction, movement in RA only won’t follow a star. The star will drift in both RA and DEC.  Guiding RA will result in a DEC drift.  All in all, both axis have to work harder when you add in other factors. 

Am I misunderstanding something?  

¯\_(ツ)_/¯

Ahh yes i get it now. Thick skull. I guess my dec drift was coincidental.

Scott Badger:
he was getting rotationally miss-aligned diffraction spikes that he attributed to the mount not being level. I don't recall if all the subs were taken while not level, or if some were taken while level and others while not level.

The orientation of diffraction spikes is a different issue than polar alignment. The position of diffraction spikes is determined by the orientation of the obstructing vanes with respect to the star field. If the plane of the bottom of the mount is changed in orientation between sessions, and an alt az adjustment is then used to correct for the consequent polar mis alignment, then yes, you will end up rotating the spikes with respect to the star field, for the same reason as tracking a star using an alt az mount will see a change in the orientation of diffraction spikes. It isn't whether the mount is level or not, it is that the base plane has changed. Leveling the mount is simply a way to try to get the base repeatably parallel between sessions.

Sean Mc:
...

Yes... you are assuming your mount and OTA are mechanically rigid and perfect at the scale of arc seconds or less - I guarantee neither are that perfect, not even from Astrophysics.  There are several phenomena that are relevant and your guider measures the cumulative sum of them all, in both axes.

If you can make your scope vibrate a little by nudging it, it means it is also flexing under its own weight and by more than just a few arc seconds. Now consider how much it moves if you apply a lateral force equal to say the weight of everything above the dovetail clamp. 

Bear in mind that over the length of the axes between bearings, of say 200mm, a movement of 1 micron subtends 1 arc second. A little wiggle of 60 microns is a whole second of arc... A movement of 1mm at the foot of a tripod translates to a substantial fraction of a degree.

1. Flexure within your OTA, mount, tripod and even the ground. Everything sags under its own weight and the weight of whatever its supporting and amateur-grade stuff is very lightly built to reduce weight and cost. It starts with the topped of the OTA, the cells the optics are mounted in, the tube itself, the way its attached to the dec axis, then the RA axis, the way the gears are mounted (often a big source of flexure there), the tripod, and even the ground its sitting on.

The downside is flexure is variable as the rig moves around the sky, and significant at the scale of arc-seconds.

With almost all German equatorial mounts there is measurable flexure of the declination axis as it rotates, and in particular the connection between the dec axis and the dovetail clamp is a weak point (that flexes too) and in addition, the way most of you mount your scopes on top isn't all that rigid.

2. Non-perpendicularity between the axis of the mount, and between the optical axis and the dec axis. And worse, as per above the the connection between the dec axis and the dovetail clamp is a weak point (that flexes too), unless you're using fork mount.

3. Bearings. It's not uncommon to see a periodic error in declination which correlates to the RA axis moving up and down a little as it rotates over the balls or rollers in its bearings because - guess what - despite how insanely hard the bearings are, the bearing races, balls and rollers are also measurably elastic at this scale.

Conversely, professional observatory instruments were designed by mechanical engineers who understood this stuff and even use fine-element-analysis to model the flexure and minimize its effects. No wonder those mounts and scopes are massive.

But that doesn't happen for amateur gear.

4. Lastly, atmospheric refraction and seeing.

On my 10" with the guider working at 3,000m focal length, the effects of seeing are quite obvious and in poor conditions the guider will "chase the seeing", ie it attempts to respond in both axes to the random movements of the guide star. 

While the little rigs with tiny guiders often report they're guiding at insanely small RMS levels like 0.1 arcsec while mine is managing 0.6 arcsec, I dont quite believe the small ones; all it means is the guidescope and sensor are so small that they simply can't detect the seeing, or errors smaller than 1 arcsec.

Scott Badger:

Alex Nicholas:
If your mount is perfectly polar aligned, it can be mounted sideways.... How level the baseplate is makes NO difference to polar alignment based drift, or guiding accuracy... The only thing that matters is that the centre of rotation of the RA axis is pointing directly at the pole... that is the be all and end all of polar alignment, and polar alignment based drift....

That's my intuitive understanding as well, but in one of John Hayes' presentations, and maybe this is something different, he was getting rotationally miss-aligned diffraction spikes that he attributed to the mount not being level. I don't recall if all the subs were taken while not level, or if some were taken while level and others while not level.

Cheers,
Scott

Yeah, misaligned diffraction spikes could be caused by having different leveling on multi night imaging sessions, but if the baseplate is tilted 2° forward and 2° to the right, as long as its the same every night, and the polar alignment is done to the same accuracy, it  will not affect anything.

I intentionally have my tripod level off (leaning forwards) as my latitude is nearly at the lower limit of what my mount will allow, so by having the tripod tilted forward a few degrees, it provides me a few extra degrees of altitude adjustment when I'm aligning.

 Having a level tripod makes polar alignment easier, as Azimuth adjustments with only alter azimuth misalignment, and altitude adjustments will only alter the altitude misalignment... if your tripod is tilted in any direction, adjusting the altitude will likely alter your azimuth alignment and vice versa... If you use an incremental adjustment process, it won't affect you much at all, but if you try to get your Azimuth locked in perfectly, then adjust your Altitude, you'll throw off your azimuth adjustment when you make altitude adjustments.. That's really the long and short of why people need to level their tripod.

Think of it as if you were hunting, as long as your target is in the crosshairs, it really doesn't matter if you're laying on a perfectly flat section of ground, or rotated by some arbitrary amount, as long as you align the target in the crosshairs, you're aligned... Same with the mount, as long as the RA axis of rotation is aligned with the earths axis of rotation, it will be just fine.

MaksPower:

Sean Mc:
...

Yes... you are assuming your mount and OTA are mechanically rigid and perfect at the scale of arc seconds or less - I guarantee neither are that perfect, not even from Astrophysics.  There are several phenomena that are relevant and your guider measures the cumulative sum of them all, in both axes.

If you can make your scope vibrate a little by nudging it, it means it is also flexing under its own weight and by more than just a few arc seconds. Now consider how much it moves if you apply a lateral force equal to say the weight of everything above the dovetail clamp. 

Bear in mind that over the length of the axes between bearings, of say 200mm, a movement of 1 micron subtends 1 arc second. A little wiggle of 60 microns is a whole second of arc... A movement of 1mm at the foot of a tripod translates to a substantial fraction of a degree.

1. Flexure within your OTA, mount, tripod and even the ground. Everything sags under its own weight and the weight of whatever its supporting and amateur-grade stuff is very lightly built to reduce weight and cost. It starts with the topped of the OTA, the cells the optics are mounted in, the tube itself, the way its attached to the dec axis, then the RA axis, the way the gears are mounted (often a big source of flexure there), the tripod, and even the ground its sitting on.

The downside is flexure is variable as the rig moves around the sky, and significant at the scale of arc-seconds.

With almost all German equatorial mounts there is measurable flexure of the declination axis as it rotates, and in particular the connection between the dec axis and the dovetail clamp is a weak point (that flexes too) and in addition, the way most of you mount your scopes on top isn't all that rigid.

2. Non-perpendicularity between the axis of the mount, and between the optical axis and the dec axis. And worse, as per above the the connection between the dec axis and the dovetail clamp is a weak point (that flexes too), unless you're using fork mount.

3. Bearings. It's not uncommon to see a periodic error in declination which correlates to the RA axis moving up and down a little as it rotates over the balls or rollers in its bearings because - guess what - despite how insanely hard the bearings are, the bearing races, balls and rollers are also measurably elastic at this scale.

Conversely, professional observatory instruments were designed by mechanical engineers who understood this stuff and even use fine-element-analysis to model the flexure and minimize its effects. No wonder those mounts and scopes are massive.

But that doesn't happen for amateur gear.

4. Lastly, atmospheric refraction and seeing.

On my 10" with the guider working at 3,000m focal length, the effects of seeing are quite obvious and in poor conditions the guider will "chase the seeing", ie it attempts to respond in both axes to the random movements of the guide star. 

While the little rigs with tiny guiders often report they're guiding at insanely small RMS levels like 0.1 arcsec while mine is managing 0.6 arcsec, I dont quite believe the small ones; all it means is the guidescope and sensor are so small that they simply can't detect the seeing, or errors smaller than 1 arcsec.

Maks.  What an excellent, comprehensive response. Thank you!

Dave Rust:
Here's an example of such a device designed for small scopes like ours.

*I used to have AO on 4 or 5 of my old SBIG cameras back in the day. The results were brilliant.. AO would run corrections via its refractive element 10 times per second, and once that refractive element found itself at 80% of its tip/tilt limits, it would send an autoguide correction to the mount.... Back then I had a 11" SCT on an EQ6R, imaging at 2800mm focal length and the AO unit made that EQ6R seem like a paramount ME...  though, as you say, only when I could find a guide star at 2800mm focal length and F/10...

Rüdiger Patommel:
The alignment of the guide camera is also important. I always make sure that the pixels of this camera are precisely aligned with the rectation.

I was under the impression that calibration assistant accounted for this?

Arun H:

TiffsAndAstro:
I was under the impression that calibration assistant accounted for this?

Yes, it does. The rotation of the guide camera does not matter as long as you run the calibration assistant after a rotation. The purpose of calibration is to determine how specific motions in DEC and RA translate to motions of the guide star on your sensor, which includes determining the angle with respect to some reference on the sensor.

Incidentally, it is highly recommended to run calibration assistant and listen to it when it asks you to slew to a certain location for the calibration. This is because, close to DEC=0, calibration will be more accurate as the same change in angle will result in a larger shift of the guide star in pixels on your camera.

just double checking, sorry. Didn't want to go and try and line both cameras up