Print

[pterodyne]

Tom,

Ive been experimenting with the Arduino version and a pair of 8825 drivers.  I have now ordered the official version from you.  It is destined for the USA, (understood that it won't be here by xmas).

Last night, I blew one of my stepper boards while adjusting the current limiter. This is on the bench only at this point as the Titan is installed in the observatory 100 miles away.   I'm wondering if I could send you my motor/gear specs and if you could tell me if you think it will work?

Losmandy Titan
Worm wheel 270 teeth both axes
Pulleys 40T on worm gear, and 10T on the stepper for 4:1 ratio
1.8 Degree Stepper NEMA 17 1.68A/Phase bipolar (full motor specs attached.)

Ive been putting into the configuration program:
1.8
4
270

and trying to get to 500 or more for max slew speed.

Im having a bit of a hard time between the relationship between the Motor voltage and the amp draw/vs speed.  I have some micro heatsinks for the driver chips as I read that 1.5A is the max per phase on the motor drivers without additional cooling.  That said, how do I figure out if the above will work without blowing my official unit when I get it?  Could the current be below 1.5 and still meet the above specs/speed requirement?  In the case of me blowing my stepper board, I was adjusting the current limiter, but was at a high voltage of 20V for the motor.  Trying to get EQMOD to not stop the motor at the high slew speed.  (Choosing a max slew speed the motor would spin up slowly, then stop)

[TCWORLD]

Ah. The higher the voltage, the lower the current the driver can take without overheating. You really ought to be using 12V not 20!

Those motors should operate reasonably well at ~1.4A. AstroEQ ideally would be run no more than 1.2A/phase, but you can push that to around 1.4A. The reason is the reverse polarity protection diode in the circuit starts to get hot at higher currents (there are 4 phases - 2 motors each with 2 coils), so the current through the diode is around 4 times higher - its not quite that simple as the driver boards convert voltage to current, so actually the input current will be lower than the coil current, hence you can push the limit.

To set the current limit, there is a small via on the driver boards near the IC (if you look on the back of the board there is a white ring drawn around it). If you measure the voltage between there and ground, whatever value you get, the current will be twice that. I.e. Ilimit = Vref * 2. So really you should adjust the potentiometer so that Vref is around 0.7V (no more than 0.72V).

If the current limit is set too low, the motors don't get enough power and so will skip steps at higher speeds. If the current is too high, the driver will overheat and either start skipping step or go into thermal shutdown.

[pterodyne]

I think I have it now..  I was thinking the same thing with a different result!  I was thinking the higher the voltage, the lower the current USED, and therefore more work from the motor while having a lower current through the driver.

do you think Ill be able to get speeds of 500 with those motors and the parameters used @12V without skipping? 

[TCWORLD]

They are decent motors, and the gear ratios aren't massively high, so its a fair bet you will get very decent speeds. On my EQ5, using some very tiny NEMA11 motors, I can reliably get speeds of 600x. A lot of it depends on weight.

[pterodyne]

I know you all are waiting to know the outcome of my project.  not really. 

I have completed my cables, and connections to the real manufactured AstroEQ, and have also made my NEMA 17 adapters to attach the motors to the losmandy titan worm blocks.  Now I'm just waiting on my fiberglass reinforced XL timing belts to arrive from Gates rubber (drop ship), ordered from Robot Marketplace.  Assuming they fit, the next step is to install on the actual worm blocks at the observatory site and test!  As it stands right now, I can run the motors up to 750X sidereal, not sure if it will still function when loaded.

Once I've had a chance to test Ill post results and pictures.  Im sure Im not the only one out there who has a hate hate relationship with the Gemini and Gemini II telescope controllers.

[pterodyne]

This is more of a mechanical question, but I received my XL timing belts yesterday.  All seems good, except now that I can actually spin the worm and the large pulley, I see that one of them has a bore hole that is slightly off center.  The whole assembly spins just fine, but it's creating a slight vibration in the entire worm block.

Anyone on here know if the tolerances will matter?  It's still a 10T xl pulley on the motor and a 40T pulley on the worm. 

[TCWORLD]

Hmm, you may find that the axis pulses as a result (an off centre pulley will act like a cam). But if you for example use guiding it shouldn't be too much of a problem. You can also try setting up PEC in EQMOD to try and counteract it.

[pterodyne]

other than the above post, I am ready to try to install.  Being a remote installation, Im a little worried about one aspect.

Tom, I noted previously that I can see the crosshairs in maximdl for the telescope oscillate back and forth a tiny bit when zoomed in.

you had indicated when I first mentioned the problem that I should increase the microstepping.  I cannot.

Here is my metrics for the losmandy titan:

motor angle 1.8
worm gear ration 270:1
motor gear ratio 4:1 (40T pulley on worm and 10T on stepper output shaft) 
microstepping 16

My recollection is that if I try to choose 32, I get an error in the configuration program, possibly about Ival 

at stepping of 16 I get around .375 arcsec per step.  I though that would be sufficient as it matches what an EQ6 does.  I don't want to install this if tracking or gotos are going to be a problem.

Tom, your thoughts?

[pterodyne]

ok.. doing some research here..

I still get the oscillation in MaximDL when the config program is using the following variables..

1.8
5.5:1
135:1
32 ustep

This gives a resolution of .273 steps per arcsec (the max I am able to get by fudging the numbers)

[pterodyne]

News! 
Sorry to have wasted any of your time Tom if youve been scratching your head..

I just tried the EQMOD simulator and it does the same thing so I, next installed EQMOD on an entirely different computer that had MaximDL but never anything related to AstroEQ on it, set it up to use the simulator, and it did the same thing.

So it probably is just a display glitch with MaximDL and EQMOD..  Lots of people are using EQMOD with MaximDL for serious astrophotography..

[TCWORLD]

The thing to remember is the position counters are in the AstroEQ controller. EQMOD has to read them back which it only does a couple of times a second, so it will appear to the computer as if it is jumping around a little bit while actually it is fine.

[pterodyne]

So I've ordered some new pulleys from a different supplier.  Hopefully they will be manufactured with greater precision.

Thanks for all the support on this Tom, I look forward to using it. 

[pterodyne]

So I finally installed the AstroEQ a couple of weeks ago.  Unfortunately the motors weren't powerful enough.  It would work fine, then randomly it would screech in one or the other axis when a slew was requested.  I tried loosening the worm gear mesh a little, but it only helped in a small way. 

In the end I think I'm going to try some NEMA 23 motors I have.  Ill have to get new belts and a different mounting scheme. 

[pterodyne]

Tom, Im assuming by limiting the current, I can still use these higher rated motors?

https://www.pololu.com/product/1478

They are 2.8A per phase @ 3.2V nema 23 @ 270 Oz-In

The motors from the above post were 1.68A per phase nema 17 @ 76 Oz-In

[TCWORLD]

You would have to limit the current to <1.5A at the most which would mean those motors would be running at half their rated current, so a massive reduction in its rated torque. If you want to get with much large motors like that, it would probably be best looking for alternate motor drivers - ones which have a step/direction type interface - and building a custom AstroEQ using those and an Arduino Mega.