RE: tait

[Tom Droege <tdroege2@earthlink.net>]

At 01:01 PM 3/19/02 -0700, you wrote:

>Hey Michael,  I'm going to appologise in advance for being my usual dense
>self.  I'm CC'ing Tom as he will be interested, and he might answer a
>question or two also.


>1) What is the step count between limits for ra/dec/focus drives?  I
>calculated 1076 from the 70 degree range and .065 step size for dec.  Just
>read limit.c, and it seems now that we're at 160 degrees (+60 -90) for 2308
>steps.  Just out of curiosity, why limit the ranges (+50 -20)?

There is a conversion between steps and angle that I will have to look up 
for the RA drive.  It is somewhere around 3 steps per pixel.  So roughly 
7/3 arc seconds per step.  I can get you an exact number when I can find 
the right log book.

But you don't need to know it unless you plan to single step the RA 
drive.  This can be done, but in general I don't know why you would want to 
do it.

The Decl. drive has a conversion factor.  I think it is 90 steps per 
degree.  But it is going to change when I get the new motors.  So just put 
n steps per degree into the code and be prepared to assign n some new value.

For the RA drive you do things by time.  When you turn it on it goes 15 
degrees per hour.  You can fast forward and fast reverse it.  Then it goes 
at 32x the sidereal rate.  You can change this fast rate in powers of 2 but 
no faster than 32x.  i.e 16x, 8x ... are possible.  Note that there  is a 
dac that provides a +/- 2% trim on the sidereal rate.  So part of every 
location is a dac value for dac 12 that sets this correction.  There is a 
10 turn pot that gets you into the ballpark.  Don't touch this unless you 
have a period meter to test it or you have a fun time trying to get it back 
in range.  Note it is roughly at 0.14 seconds per step.  With this and 
knowing that the radius arm is 12" and the screw thread is 24 threads per 
inch will get you steps to angle conversion.  Then a few pictures at known 
times will allow exact correction.  Note that this is one of the things to 
be built into the control program.  A way to take a few images and make 
adjustments to the RA drive trim to get it right.  I have been patiently 
waiting for this code.  Also needed to be built into the code is a focus 
scheme.

There is only one limit on each drive.  The Declination drive is set to 
find the limit moving towards the horizon from above the horizon.  The 
limit sense goes from 0 to 1 when at the limit.  The declination motion can 
go from the southern horizon to somewhat past the pole.  For the 
Declination drive, the best thing to do is to move it away from the limit 
constantly testing the limit switch until it goes to 0 state.  Then move it 
towards the limit switch (slowly) testing after each step until the limit 
goes to !.  If you don't care for an exact position you can move to the 
limit at full speed if you test fast enough.

There is only one limit on the RA drive.  But it can be moved in 1/12 
radian steps.  You can actually move through the limit switch.  So if when 
you turn on the motor decides to be in a state where it is moving towards 
the limit switch and you are in the limit switch and are not testing it, 
you can go right through the limit switch and grind the motor when it 
reaches the physical limit.  Further, if you (without looking at the 
telescope) try to home the RA, it is going in the wrong direction and will 
just never get there.

So the first thing you want to do when turning on power is to stop all the 
motors.  Just look at the init in the QBasic code.

I figure this is self correcting.  After you replace the chamois in the RA 
drive a few times you will figure out how to make sure this does not happen.

Note that there are many more channels available for adding limit 
switches.  They are cheap.  So if you all want to add more protection, it 
can easily be done.  Well, you may have to drill more mounting holes in the 
RA drive and figure out how to glue a limit switch in place for other 
things, but I just got tired of worrying with it.  There will be plenty of 
opportunity for you all to improve your Mark IVs.


>2) I notice that in radec.h, the home position is defined as -17.5 degrees,
>yet the software limits are at +-16 degrees?

I have not measured it.  I think the home position can best be determined 
by an experiment measuring a star field and knowing the time and how long 
it has been since the telescope was at the home position.

Note, in general, if you want to know where things are you want to take out 
the backlash.  So you always want to perform the same sequence of moves to 
the limit switch so the backlash is always in the same direction.



>3) Are there limit switches on both sides of travel for the mechanisms
>(ra,dec,focus)?  Seems not from limit.c.  If I do a <ra tolimit> (and is
>this -17.5 or -16 degrees), and then track siderealy, what happens when I
>reach the end of travel?  Is this a very bad thing to do and I must prevent
>this from happening?  Same for focus.  Seems like dec has a single limit
>switch tripped at both extreme's of travel.

A well adjusted chamois RA drive will overpower the drive motor.  So it 
will not be damaged (usually) if you hang at one limit or the other.  There 
is only a limit switch at the most east position.  When you reach the end 
of the travel, the mechanism just hangs with the motor trying to drive it 
further.  The only really bad thing about this (if you don't wreck the 
chamois drive) is that it may not have enough torque to get back off the 
end of travel.  So when you try to reverse it, it may just sit 
there.   Then you have to go to the telescope and unjam it by hand.

When I am doing a focus run, I always move fully into the limit switch and 
then out to the test position.  The trombone is not as rigid as I would 
like and this gives you a better chance to get a good focus.  Sigh!  I 
wished I had use 1/2" shafting instead of the 3/8".


>4) Can the scope be damaged by attempting to go beyond limit switches,
>either by direct command (dec/focus) or by letting it track too long (ra)?
>If so, I'm thinking of a simple watchdog program to keep track of the
>functionality of TAIT, and/or maybe adding a spawned task to TAIT to keep
>track of ra position when tracking sideral rate or moving.

The focus motors have different problems.  There are stops on the focus 
trombone so the camera cannot fall off the end on to the floor.  (probably 
damaging the TEC in the process).  At the close in position, there is a 
limit switch and if you try to move too far past the limit, it just jams 
and stalls the motor with mostly no damage.  If you don't have the hard 
stops set correctly on the trombone, the drive screw could pull out of the 
motor.  Then you would have to go to the telescope and command movement 
toward the limit switch and push the drive screw into the motor to get it 
started.


>5) Why no ra abs/rel movement?  Is there a repeatability problem?  Backlash?

Yes, there is backlash.  I don't know how much.  Just begging to be 
measured by some simple experiments with star fields.

OK, possibly Michael can stuff this into the "service notes" so I can avoid 
answering these questions yet again.

Tom Droege



>Thanks for being patient,
>Rob