Tracking error - Widefield dslr on Barn door tracker

[Jasp05 (Aaron)]

Hi guys,

Been researching on making Barn door trackers and people always mention "Tracking Error".

So a basic single arm barn door with curved rod is apparently only good for 15-20min before the tracking error is quite bad.

Does this mean you have to redo the polar alignment after 20 mins? or just reframe the shot in the camera due to it "drifting" through the frame.

And one last silly question. I have a barndoor tracker, and I use it with the hinge on the left hand side of the tripod. I have to use my tracker in reverse (close the doors together rather than drive them apart) to have it track. If I drive apart my star trailing gets worse..

All the other users I've seen on these forums and online, use there hinge on the left hand side but drive the doors apart and have no issues.

I only ask this as pulling the doors together introduces some other problems. (the gear on the rod wants to pull up away from the base, and almost disengages from the drive motor.


I know I could swap the hinge to the right hand side and drive the doors apart and have it work, but everything I see has the hinge on the left side. (feels weird looking down the hinge on the right side )

[Mickoid (Michael)]

Aaron, I've never used a barn door tracker so I can't really give you much help but I would imagine if you can get more than 5 mins accurate tracking with your mount, you're doing OK. Your PA would have to be pretty good for shots that long. Like you said, reverse the polarity on the motor so it rotates the other way and turn the barn door around so the hinges are on the right. Maybe the ones you've seen with the hinges on the left have been made for use north of the equator.

[JA]

Originally Posted by Jasp05

Hi guys,

Been researching on making Barn door trackers and people always mention "Tracking Error".

So a basic single arm barn door with curved rod is apparently only good for 15-20min before the tracking error is quite bad.

Hi J,

I suspect a lot of people wouldn't recommend more than a few minutes even on a commercial tracking mount, without auto-guiding. So I would suggest 15-20 minutes is quite good if you can get it from a barn door tracker, even with a curved threaded rod drive. Of course the visibility of the tracking error in terms of star trailing will be less evident with shorter focal lengths, so maybe that is where some of that generosity with the 15-20 minute exposure time comes from.

The tracking error becomes worse as the exposure time increases, since there are a few time dependent errors potentially at work, namely:

1. The Barn door tracker is not accurately set or manufactured to the true length of sidereal day as it was not set correctly or accurately enough in the first place or the variation in the length of the sidereal day (although tiny) is such that it no longer is close enough to that assumed in the design of the mount. Also possible is that the gear ratios are close enough, but not quite good enough, or similarly that the radius of curvature of the threaded rod is NQR, thread pitch NQR, hinge to drive shaft distance NQR .......

2. There is time dependent variation in the voltage supplied to the drive motor, as perhaps in the case of an unregulated power supply to a DC motor drive or slowly discharging battery.

In addition to these errors there will also be periodic errors in the drive, many of which relate to speed variation, backlash take-up ......

Originally Posted by Jasp05

Does this mean you have to redo the polar alignment after 20 mins? or just reframe the shot in the camera due to it "drifting" through the frame.

No, you shouldn't need to, unless you run out of drive rod. Just stop the exposure (or time it to a selected length) and the take another shot provided that the framing hasn't changed so drastically in your viewfinder that it can't be adjusted in post processing, either manually or say stacked automatically .

Originally Posted by Jasp05

And one last silly question. I have a barndoor tracker, and I use it with the hinge on the left hand side of the tripod. I have to use my tracker in reverse (close the doors together rather than drive them apart) to have it track. If I drive apart my star trailing gets worse..

Yes I can understand that. It relates to a slight difference in drive speed in both cases, whether pulling doors together (Left Hinge) or pushing them apart (Right Hinge). In one case: driving (pulling) the barn doors together, it appears that the barn door tracker's angular velocity more closely approximates the true siderial rate of rotation of stars around the South Celestial Pole. In the other case driving (pushing) the barn doors apart, not so much.... This is caused by a load variation between both cases which affects the drive speed slightly. Assuming of course the motor doesn't use speed regulation in the electronics.

Best
JA

[Jasp05 (Aaron)]

Just to clarify. I didnt mean that 15- 20mins was the exposure length. But i think they were more referring to thats the period of time where each of your exposures is still capturing the same frame each time. After that time fram you start to get drifting of the picture. Or at least that was how i was understanding it.

At the moment with my dodgy polar alignment (cant actually see the pole from my location) ive still been able to get 2 min subs at short focal lengths. I think 45sec to 1 min at 200mm is my best so far. (Which i think is ok for a first time barn door tracker).


And why would driving the doors apart not track the same as pulling together. I get that the load is kind of working against the push out. But surely once the motor is moving and up to speed it should be the same?? I am using an arduino for speed control of the stepper. Gears are one to one ratio at one rpm.

I would like to work out some error correction and program that in but i havent found a way to calculate my error as yet. (Any tips on this will be much appreciated).

Ive actually got a printable template with holes pre marked etc to help with accuracy if anyone would be interested. I just glued it to some wood and drilled with the markings.

[JA]

Originally Posted by Jasp05

Just to clarify. I didnt mean that 15- 20mins was the exposure length. But i think they were more referring to thats the period of time where each of your exposures is still capturing the same frame each time. After that time fram you start to get drifting of the picture. Or at least that was how i was understanding it.

At the moment with my dodgy polar alignment (cant actually see the pole from my location) ive still been able to get 2 min subs at short focal lengths. I think 45sec to 1 min at 200mm is my best so far. (Which i think is ok for a first time barn door tracker).

Hi J,

When you say "So a basic single arm barn door with curved rod is apparently only good for 15-20min before the tracking error is quite bad." (without any further qualification), it's fairly reasonable to conclude you meant a 15-20 min exposure , but if not, all the better, as shorter is obviously easier to handle with reduced error from time dependent effects.

Post a picture or two of your tracker, showing the whole unit, with camera mounted as you use it, showing the drive and also an example of a left hinge (door closing) astro image and a right hinge (door opening) astro image, both preferably of the same exposure length to better gauge the nature of the issue raised.

Originally Posted by Jasp05

And why would driving the doors apart not track the same as pulling together. I get that the load is kind of working against the push out. But surely once the motor is moving and up to speed it should be the same?? I am using an arduino for speed control of the stepper. Gears are one to one ratio at one rpm.

In one usage (closing doors/hinge left) the force from the weight of the camera & the moving part of the top plate acts in the same direction as the drive force provided by the motor via gears & the screw thread/nut. In the other usage (opening doors/hinge right) force from the weight of the camera & the moving part of the top plate acts against the drive force provided by the motor via gears & the screw thread/nut. This will increase the current demand from the motor and the nature of the motor's speed control may be affected by this. The power supply may sag slightly. The motor or power supply may be current limited or controlled, in addition to speed controlled or perhaps as a defacto to it. Who knows?

Do you have details of the nature of its speed control?

Whilst it would be better to solve this, you can correct for it instead....

Originally Posted by Jasp05

I would like to work out some error correction and program that in but i havent found a way to calculate my error as yet. (Any tips on this will be much appreciated).

Sure that's possible, but it would help to know what is the maximum period of time that your mount can track from the start to the end of the threaded rod and what camera/s you use.

Best
JA

[rcheshire (Rowland)]

Tangent arm mounts are a lot of fun and there are ways and means of tuning a double arm drive and getting consistent tracking performance. Here's a few ideas. Not too high tech.

[middy]

Originally Posted by Jasp05

Hi guys,

Been researching on making Barn door trackers and people always mention "Tracking Error".

So a basic single arm barn door with curved rod is apparently only good for 15-20min before the tracking error is quite bad.

Does this mean you have to redo the polar alignment after 20 mins? or just reframe the shot in the camera due to it "drifting" through the frame.

I started out with a home made barn door tracker (with a straight rod, not a curved rod). I would take many 30 sec exposures (the longest exposure time of my point and shoot camera). What you will find is that after 10 minutes or so the stars in the image will be elongated. This is because as the tracker is running its rotation rate is gradually slowing down due to tangential error. To fix it you have to wind the tracker back to the start position again and then carry on. No polar re-alignment necessary, only re-framing because the tracker has been reset.

However, using a curved rod should take care of the tangential error so it will keep tracking at the correct rate until you run out of rod.

Mine used a straight rod so I set it up to run slightly faster than the correct rate. That way it started out running too fast, after 10 minutes it would be rotating at the correct rate, after a further 10 minutes it would be too slow and need resetting. This effectively doubled the time I could image between resets.

cheers,
Andrew