EQ6Pro Belt Mod notes

[chrisshillito]

Here are some notes/comments regarding my own belt mod conversion. These note complement the following threads:

http://stargazerslounge.com/diy-astronomer/159157-eq6-belt-drive-modification.html

and

http://stargazerslounge.com/diy-astronomer/160466-heq5-experiments-belt-drive.html

• The two part construction of the motionco 12 tooth gear means that when bored out to 5mm there is very little metal left on the inner sleeve that joins the hub to the pulley itself. Its probably worthwhile ordering a spare 12 tooth gear just in case anything goes wrong in the boring process.
  

• Order spare belts – if you don't get the tension right the belts will slip and quickly the teeth will wear.
  

• I lengthened the flat on the worm shaft that accepts the pulley grub screw. This gives a little more flexibility in positioning the pulley and allows you to get it away from the worm bearing.
  

• Extend the slots on the motor mounting plates. On the DEC motor plate I filed back one of the motor support pillars just to get a little more play when setting the tension.
  

• Position the pulley on the motor shaft so that the distance from the top of the pulley flange to the motor plate is 9mm. Note that this means the top of the motor shaft will be recessed inside the pulley.
  

• With the worm carrier removed fit the motors and check the motor pulley position by looking through the slot through which the belt will pass.
  

• With the motors removed, fit the worm carrier and make sure the belt has plenty of clearance passing through the mount body. I widened the casing slot slightly and rounded off all the edges - may not be necessary but I'm just cautious.
  

• I used epoxy to to secure the pulley to the motor – with extra applied to the top of the flange as it isn't being held on by much metal (and I don't want it coming loose and falling into the worm compartment). I know some folks will be concerned that doing this makes the mod irreversible but even if the epoxy can subsequently be weakened by heat it really shouldn't be too hard to locate some suitable replacement 200 step bipolar stepper motors.
  

• Make yourself a wire hook to help fit the belts on the motor pulleys – once you get the knack its quite easy.
  

• When putting the motors in use a strongly magnetised screwdriver or a dab of grease to keep the washers in place in until you've got the bolt safely located. You can bet that if a washer does drop off it will head straight into one of the worm drives or down onto the polar scope (I had both before I learned my lesson!)
  

• To adjust the belt tension I made up the little device shown in the accompanying attachments. It simply sits on the top corner of the motor and the set screw is tightened against the casing side wall using an allen key. This pushes the motor inwards tensioning the belt. When your happy you just tighten the mounting plate bolts and slacken set screw to remove the tensioner – patent pending, every belt kit should have one!
  

• Run the motors at a fast speed for a few minutes then check inside the motor housing for any signs of belt dust – if there is remove the motors and inspect the belts (you might as well remove the DEC axis totally as you'll find it hard get the belts back on otherwise.
  

The belt mod is rather fiddly. The belts are not easily accessible and its hard to monitor what is going on with them once the mount is reassembled. Changing the belts will require complete disassembly. Be prepared to disassemble the DEC axis if you remove the motors – it hard to get the belts back on with the DEC axis in place. Should I ever have the mount in pieces again I may well drill and taping some holes in the rear of the motor housing to make the process of belt engagement and inspection easier.

I would recommend limit protection to be enabled as there is the possibility that the belt might start to slip with resultant damage to the belt teeth if movement is obstructed. Although the belts are cheap the hassle involved in belt replacement (particularly RA axis) means you really want to take all steps to avoid this if possible. Cable snags also present a situation where damage to the belt might occur.

All in all I'm happy with my belt mod. The only real concern in recommending the belt mod to others is the marginal nature of the machining required on those 12 tooth pulleys. An alternative solution worth of consideration would be to source a pair of 0.9 degree/step bipolar stepper motors and fit 24 tooth motionco pulleys instead (these are of a solid construction).

Chris.

[malc-c]

Chris,

Nice write up. I agree 100% about the small motor pulley not having much meat left on it when bored out. At least on the HEQ5 both the motors and the drive axis are exposed once the cover is off , allowing quick replacement of the belts and or pulleys should they need it.

On the subject of belts, Belingonline supply 80t, 82t, 85t, 87t, 88t and 90t which should allow for adequate tension. I have a couple of 82t belts on order which I'm hoping will fit the bill on the HEQ5... I'll up date my thread later after they've been delivered.

[malc-c]

An alternative solution worth of consideration would be to source a pair of 0.9 degree/step bipolar stepper motors and fit 24 tooth motionco pulleys instead (these are of a solid construction).

Chris.

Chris,

Googling "0.9 degree/step bipolar stepper motors" resulted in loads of hits, and varying prices

Is is there any other info on the specification for the existing motors around ?

[George]

Great writeup Chris

All in all I'm happy with my belt mod. The only real concern in recommending the belt mod to others is the marginal nature of the machining required on those 12 tooth pulleys. An alternative solution worth of consideration would be to source a pair of 0.9 degree/step bipolar stepper motors and fit 24 tooth motionco pulleys instead (these are of a solid construction).

Chris.

Im just waiting on some pricing on all steel 12 tooth pulleys ready bored out to 5mm, problem is minimum order of 50 for these custom made.

Ive bought myself a small lathe and Im toying with the idea of offering everything needed in a ready to fit kit if I can get the minimum of 24 forward orders to make it worthwhile getting the custom 12 tooth gears made.

[malc-c]

Great writeup Chris

Im just waiting on some pricing on all steel 12 tooth pulleys ready bored out to 5mm, problem is minimum order of 50 for these custom made.

Ive bought myself a small lathe and Im toying with the idea of offering everything needed in a ready to fit kit if I can get the minimum of 24 forward orders to make it worthwhile getting the custom 12 tooth gears made.

PM sent

[chrisshillito]

Attached are my initial PE recordings for the belt drive.

As you can see from the frequency spectrum there are only really two signals present (worm/worm pulley and stepper pulleys) and the response is nicely repeatable. There is no evidence of any 10S jitter from the belt meshing which is rather pleasing because with gears I always used to have a 10s gear mesh signal present. The overall Peak to Peak amplitude is very similar to what I used to get prior to the belt mod so I assume that the worm is generating the majority of this rather than the pulley drive itself. Perhaps with a bit more tweaking tweaking this can be reduced a bit more (I may try swapping the worms next time I have the mount in bits)

I've still to do a logged run with PEC active but this signal should correct nicely.

Chris.

[MG1]

That a really nice repeatable wave there!

Should correct well I'm guessing!

[chrisshillito]

Chris,

Googling "0.9 degree/step bipolar stepper motors" resulted in loads of hits, and varying prices

Is is there any other info on the specification for the existing motors around ?

Hi Malcolm,

Sorry I haven't researched it yet at all, it was just something that occurred to me whilst I was writing the post above.

I know they will be 12V and bipolar but I'll have to measure the resistance of the coils of the current motors and the dimensions.

I expect 0.9 degree/step will be quite expensive but in themselves they offer a performance upgrade - more torque and smoother microstepping. Being able to fit a larger radius gear on the stepper end with double the number of teeth gripping the belt must also make for a more robust drive.

Chris.

[malc-c]

Hi Malcolm,

Sorry I haven't researched it yet at all, it was just something that occurred to me whilst I was writing the post above.

I know they will be 12V and bipolar but I'll have to measure the resistance of the coils of the current motors and the dimensions.

I expect 0.9 degree/step will be quite expensive but in themselves they offer a performance upgrade - more torque and smoother microstepping. Being able to fit a larger radius gear on the stepper end with double the number of teeth gripping the belt must also make for a more robust drive.

Chris.

Sounds like the ideal upgrade..I guess you would ideally match the current motors rather than those with more torque as I would assume there is a risk to the controller board if a high torque motor dew too much current from the controller ?

[Psychobilly]

Thanks for the info and all those involved in the original threads...

Generally the NEMA Frame size is the Edge length in 1/10" ... 23 and 24 are both the same...

NEMA 8 0.8" square

NEMA 11 1.1" square

NEMA 14 1.4" square

NEMA 15 1.5" square

NEMA 17 1.7" square

NEMA 23 2.3" square

NEMA 24 2.3" square

NEMA 34 3.4" square

NEMA 43 4.2" square

I may have a go at the HEQ-5 Syntrek sometime after all it only cost me a Fiver ....

Peter...

[MG1]

For those following the thread who like me have/had no idea what frame size means, here's some reading: http://www.motorsanddrives.com/cowern/motorterms2.html

Every day's a school day on SGL!

[chrisshillito]

I know they will be 12V and bipolar but I'll have to measure the resistance of the coils of the current motors and the dimensions.

Scrub that comment about being 12V -I just noticed the driver chips have their own switched mode power supply - so I need to check. There is a guy on the EQ6 Y! group who has already fitted 0.9 degree motors in a 3:1 belt mode (giving effectively 6:1 over the original design). I'll ask him.

Chris.

[chrisshillito]

Sounds like the ideal upgrade..I guess you would ideally match the current motors rather than those with more torque as I would assume there is a risk to the controller board if a high torque motor dew too much current from the controller ?

You'd match with the current motor spec for coil resistance, voltage and dimension. The torque is a function of motor design but if all else is equal (including torque itself) having a half the step angle means that the maximum torque will be applied every 0.9 degrees of ration instead of every 1.8 degrees.

Don't get your hopes up though - I really don't know if appropriate motors are available!

Chris.

[MG1]

Out of interest would this require a custom ASCOM diver? I know that EQmod has the facility to input custom profiles, but what about other software...Stellarium, CdC, phd?

Or would the idea be that the pulley ratios be such that the actual output resolution is the same?

Would this effectively stretch the PE graph? Or since the main contributor is the worm, have little real effect?

[chrisshillito]

Out of interest would this require a custom ASCOM diver? I know that EQmod has the facility to input custom profiles, but what about other software...Stellarium, CdC, phd?

Or would the idea be that the pulley ratios be such that the actual output resolution is the same?

Would this effectively stretch the PE graph? Or since the main contributor is the worm, have little real effect?

EQMOD is the custom ASCOM driver. At the ASCOM levels there are no, steppers, microsteps, worms or pulleys - it operates at a much higher level. So gotos and position reads are in terms of RA, DEC, Alt, AZ. ASCOM Pulse guiding simply works by specifying a direction (N, S, E, W-) and a duration. It is the driver that maps all this to whatever hardware is present.

The PE graph is not streatched in any way. We still have to track at sidereal! The rotation period of the worm therefore does not change at all wen performing these belt mods - yes the gearing in the worm drive changes if you go to a 4:1 belt mode but the driver compensates according by instructing the mount controller to step a little faster than it would normally do.

Chris.

[Reggie]

Scrub that comment about being 12V -I just noticed the driver chips have their own switched mode power supply - so I need to check. There is a guy on the EQ6 Y! group who has already fitted 0.9 degree motors in a 3:1 belt mode (giving effectively 6:1 over the original design). I'll ask him.

Chris.

Any idea what the motor driver chips are Chris? They might well be running at 12v with a motor that is specced as an example to 6v, it's perfectly acceptable to do this if you current limit the power getting to the motor in some way depending on the driver, for instance, the easy driver stepper driver board for the SGL observatory automation focuser has it's own current limit adjustment pot which is a function of the 'chopper' motor driver IC, just make sure that you are supplying the correct current and as much or more voltage than you need and everything is fine. All that happens is you increase the speed that the coils energise.

If anyone has their mount apart with the motor driver board out, I'd be most grateful if someone could take some hi-res images of the front and back of the board.

[chrisshillito]

Any idea what the motor driver chips are Chris?.

For each axis the board provides microstepping via two A3959 Full bridge PWM motor driver chips.

Chris.

[chrisshillito]

Had this response back from Jeff Machesky on the EQ6 Y! group

The stepper motors I'm using are 17Y9304S-LW4 from Anaheim Automation. I had to special order them at the time, but they might have them in stock now. They were a bit pricey compared to 1.8 deg/step motors, but not too bad. Very good quality motors. I did end up reducing the resistance on the current sense resistors a little bit to get the little extra current these want. They worked without that mod, but didn't have the full torque. Simple mod however. The IC's still run fairly cool.

17Y9 - High Torque Stepper Motor

Chris.

[Reggie]

The stepper is a 2.8v 1.19a per coil, unless there is a 3a or bigger v-reg set lower than 12v in there then it seems to suggest that it's being run at > 2.8v (the rating for the stepper) I could be wrong. Getting the current correct as he has done is key for getting the torque and speed consistent with the ratings of the stepper.

Those are beefy little chips 3amp 50v max output I believe, I just wonder why they went for 2 seperate ones instead of one of the integrated stepper chopper driver chips that allegro do? I wonder if IC temperature was a consideration or cost. I wonder if the design has changed at all over the years?

[chrisshillito]

A LT1170 switching regulator provides the load supply to the A3959s.

I would imagine cost is the driving factor for the design. As I understand it each PWM drives a separate coil. The PWM drivers are out of phase so as the current to one motor coil increases the current to the other coil decreases thereby implementing the microstepping.

Chris.

[Reggie]

That ties in quite nicely with the anecdotal stuff I'd heard about it needing 2-3amps when doing a full speed slew that uses both axis.

[George]

Had this response back from Jeff Machesky on the EQ6 Y! group

The stepper motors I'm using are 17Y9304S-LW4 from Anaheim Automation. I had to special order them at the time, but they might have them in stock now. They were a bit pricey compared to 1.8 deg/step motors, but not too bad. Very good quality motors. I did end up reducing the resistance on the current sense resistors a little bit to get the little extra current these want. They worked without that mod, but didn't have the full torque. Simple mod however. The IC's still run fairly cool.

17Y9 - High Torque Stepper Motor

Chris.

17Y9304S-LW4 by ANAHEIM AUTOMATION - Buy or Repair at PLCCenter - PLCCenter.co.uk

Estimated Retail Price - £37.56

[malc-c]

At those prices it does make you wonder why Synta don't use these motors with a standard interger (ie 4:1 or 5:1) ratio with belt drives as standard. I'm not saying the HEQ5 or EQ6 aren't good mounts, but given the volume they must sell world wide, using this sort of precision drive train wouldn't add a lot (if any) to the cost of production.

[chrisshillito]

I would assume the design is such that it meets a particular target cost and that can be reliably produced to particular minimum standard. There may be little advantage in fitting higher precision motors given production tolerances elsewhere. We as individuals can supertune our mounts but that isn't going to happen on the production line. Also a belt is a more fragile component than a brass gear. A loosely meshed gear won't loose its teeth - all you get is backlash. A loosely tensioned belt will slip and the belt teeth will soon get damaged.

I've always assumed there must be some very good design decisions behind why Synta chose such odd ratios - I'm just not smart enough to figure it out. The best I can come up with is that not all their design goals are focused soley on maximizing astronomical performance and that the non integer ratios are chosen specifically to ensure an even wear on all components.

Chris.