Hi Stephan,
Good to hear from you again.
Bartels' system is quite impressive as well as being architecturally and electronically simple. I have seen implementations of it that are as good as the highest-end telescope control systems I've ever seen (which are, basically, PC-TCS on the Bok telescope and the 40" Yerkes telescope, the control system on the Apache Point 4m, and the control system on the Sloan Digital Sky Survey telescope). I've personally witnessed a bartels system slewing around a 36" Newtonian, and I've seen it control much smaller telescopes as well. It includes PEC programming, so the tracking can be very accurate, as well as mount error modeling (certain terms only, not as full-up a model as T-Point).
Disadvantages: It helps to know a lot about stepper motors and torque issues. If you get the right steppers and gearing, Bartels' system works incredibly well. If you goof, then it can be ugly. Also, a dedicated computer runs the system in real time, using MS-DOS, PC-DOS, or DR-DOS - not what I would consider robust, real-time operating systems, but that bias aside, I have to admit that it works very reliably if you dedicate the PC to the bartels software.
Most intriguingly, the Bartels system was originally designed for alt-az telescopes (hence large, easily mounted ones), and includes a derotator circuit. There is potential here for someone to build a 0.5 or 0.6 meter automated telescope for under, say, $6,000, and put it to some good use. About two years ago, I wrote a proposal to do just this. Naturally I didn't get the funds, but I wrote it mostly to make the point.
In any case, I'm sufficiently impressed that I am converting 701's 16" Meade to a Bartels control system (lightning zapped the Meade electronics to smithereens late last year). I've got the barest beginnings of a page about that here:
http://www.roboticobservatory.com/701/
You can see a bit of the mechanical work, etc. Since these images were taken, I've added 5:1 reduction on the motor drive shafts and fluid-filled flywheels (dampers, technically). This appears to be all it takes to increase our torque and quiet our motor vibration problems, so obviously I should have done it the first time. But this a bit of a learning experience.
>I have been downloading your ACP
>scripts;
The more I work with these scripts the less satisfied I am. They were written when I was still learning the VB ropes at a very basic level, and are full of all kinds of bad programming practices. I'm working on wholesale replacements (but don't hold your breath - I'll have them ready eventually, but probably not soon).
>The el-cheapo approach has quite a bit of relevance
>especially to young postdocs/faculty with big ideas
>but small budgets ... hence my interest in your
>remarks about Bartels.
I'm very much an advocate of the cheap approach on telescopes smaller than about 0.4 or 0.5 meters. I've now spent enough time in the field, working on various systems, that I think I've got a good basis when I say that I simply don't see the performance gain needed to justify the cost of a more "premium" system. I am currently using (for photometry) a 12" Meade that I acquired solely for software development. I'm looking forward to my next scope/camera in perhaps another year, and am genuinely torn between cloning the system that I've got, or going to something somewhat larger. I can do a lot of work that no-one else is doing with the 12"; doubling my observing capacity with an identical system also cuts down on custom programming issues that I might face, and makes general system maintenance simpler. So there are some arguments both ways.
My existing system cost well under $12,000, and I'm doing original and published/publishable scientific research with it, without ever attending the telescope. Its nearly a miracle. I can't believe more people aren't doing this.