Guiding - Disturbance in RA

[Guest]

I have an HEQ5 with the rowan belt mod which I installed myself. I'm just getting into guiding but I may have a problem with the RA.

The PHD2 guiding works ok but I noticed on the graph that every now and then there'd be  big disturbance in RA which PHD would compensate for. I assumed this was due to a gust of wind because I have a big scope which catches the wind. Last night I decided to investigate it so I took off the scope and mounted the guide scope directly on the mount. I still got the disturbances in RA but I found that they come very regularly every 90 seconds.  Also the shape of the disturbance on the PHD2 graph is exactly the same. RA suddenly dives down almost off the chart. Then PHD2 rapidly compensates, RA overshoots the other side but is then brought back down and then its ok again for another 90 seconds. The disturbance from start to finish takes about 10 seconds and looks like one cycle of a sine wave. I have PHD2 on 1 second cycle.

So I guess this might mean there's a fault in the drive chain somewhere and I should be able to work it out from the 90 second cycle. I wondered if anyone could give me a pointer.

Cheers

Steve

 

[malc-c]

Hard to say what could be the cause.  When you say 90 seconds, I presume that is when tracking?  - do you hear any "tick" form the mount when this happens? - If you do does the ticking speed up when the mount is slewing ?

You could remove the belt kit and check both pulleys and the belt for any defects, and if you have concerns return it for a replacement, but these are CNC machined and I've never heard of any complaints before.

Another thing to check is the free-play of the worm, both end play and tightness against the main gear.

If you have the guiding logs, try uploading them so we can see what's going on

[Guest]

Hi Malcolm,  I've just been looking at the guide logs for last nights work. The graph clearly shows the disturbances in RA which are actually 120 seconds apart and not 90 seconds. This corresponds exactly with the rotation period of the motor shaft with the 9 tooth belt wheel on it.  I've taken the side off and I can't see anything obviously wrong. I thought the belt was a little tight so I've slackened it off a bit.  Take a look at the guiding log at 23:20:34.

I'm going back to take the belt off and see if there's anything.

The PHD2 log viewer is pretty cool.

Steve

 

PHD2_GuideLog_2021-08-11_091104.txt

[michael8554]

Hi Steve

There are other examples of the RA problem.

Here's part of your Calibration:

Note how the RA steps (blue) are bunched together in pairs, instead of being evenly spaced like Dec.

So there might be slight stiction that is taking two pulses to overcome.

From other posts I've read it's likely your 120 sec problem is something in the belt drive.

Try shifting the belt a couple of teeth on the cogs.

Try swapping RA and Dec belts if they're the same.

Look for signs of trying to jump while slowly slewing.

Michael

Edited August 12, 2021 by michael8554

[Dinglem]

It maybe binding, when you installed the Rowan belt mod you would have adjusted the backlash, you could run the mount for several minutes with no load and listen to the motors, if they change pitch there's likely to be binding in the gears.

[Guest]

59 minutes ago, Dinglem said:

It maybe binding, when you installed the Rowan belt mod you would have adjusted the backlash, you could run the mount for several minutes with no load and listen to the motors, if they change pitch there's likely to be binding in the gears.

Any binding would be on the worm and worm wheel would it? Since there are no gears in the drive chain now with the belt mod.

 

[Dinglem]

Yes in the worm gears.

[malc-c]

Based on what's been said, if the period correlated to one revolution of the motor pulley, then it can't be an issue with the belt as any defect in the belt would present a longer period.  One thing to consider is when a belt drive is fitted you are placing tension (compression) between the motor pulley and worm, whereas the natural tendency when the stock gears are fitted is the opposite, as the gears are pushing out against each other.  This could mean that there is stiction between the worm and main gear.  It might be worth adjusting the mesh and backlash of the RA axis to see if that resolves the issue.

With the belt removed, does the motor pulley run true, or is there any wobble, or eccentricity ? 

The issue is very regular as you say, at 120 seconds 

[Guest]

Thanks Malcolm,

I've made those adjustments and I'm hoping to try it out tonight.

Steve

 

[newbie alert]

The reason that you're getting the same outcome is because you're using the same system... the guidecam/ scope is the same..

Don't sound like the worm to me as that will be far more than 120secs to do a full revolution .. I'd look at the pulley and check it's not oscillating,  so run the gears on full and see it's not got a wobble... working on a 4-1 reduction sounds like it could be around that timeframe

 

[malc-c]

22 minutes ago, newbie alert said:

 

Don't sound like the worm to me as that will be far more than 120secs to do a full revolution .. I'd look at the pulley and check it's not oscillating,  so run the gears on full and see it's not got a wobble... working on a 4-1 reduction sounds like it could be around that timeframe

 

I have to agree that it's too frequent to be on the main gear, but it could be worm, but more likely the motor pulley.  However the Rowan doesn't use a custom 4:1 ratio, it uses the stock 47:9 ratio.   

The RA gear ratio of an HEQ5 is 705:1, so if I've got the math right, and bear with me here, 24 hours is 86,400 seconds (24*60*60).  86,400 divided by 705 to get the period for a single revolution of the RA motor results in 122.5 seconds (rounded).  This is very close to the 120s frequency of the spikes in the tracking graph.  So either the pulley has a fault or there is something going on with the stepper motor

[markse68]

is the motor pulley a loose fit on the motor shaft?

Mark

[Guest]

On 12/08/2021 at 16:26, michael8554 said:

Hi Steve

There are other examples of the RA problem.

Here's part of your Calibration:

 

Note how the RA steps (blue) are bunched together in pairs, instead of being evenly spaced like Dec.

So there might be slight stiction that is taking two pulses to overcome.

From other posts I've read it's likely your 120 sec problem is something in the belt drive.

Try shifting the belt a couple of teeth on the cogs.

Try swapping RA and Dec belts if they're the same.

Look for signs of trying to jump while slowly slewing.

Michael

Hi Michael,

Last night was the first clear night I had to test this again.

I checked that the 9 tooth motor cog on RA was tight. It seemed tight but I tweaked it a bit more.

I ran it with the cover off at high and low speeds. I couldn't see any wobble on any of the wheels. I didn't see any belt jumping or stalling. Basically it looked ok. It seemed to be running smoothly with no change in the sound. In the end I reinstalled the belt drive mod on RA checking carefully at each stage.  The belt looks ok, no malformed teeth or uneveness. I did notice that the belt was running slightly off centre. The motor cog has a shoulder at the end and the belt was running up against that. So I adjusted that so it was clear of the shoulder.

I ran it all again last night but exactly the same problem. Also on the calibration I had exactly the same paired points on RA.

So I'm thinking I'll take out the belt mod on RA and put the gears back.  If you have any other ideas I'd be very glad to hear them.

If there is stiction or binding am I right in saying that the only place that could occur is at the worm and worm wheel?  I did slacken off the backlash in RA in case I'd over done it. My understanding is that backlash in RA is not important in guiding because it's always running the same direction. I can now feel a very slight backlash in RA. 

Cheers

Steve

 

[newbie alert]

1 hour ago, woodblock said:

Hi Michael,

Last night was the first clear night I had to test this again.

I checked that the 9 tooth motor cog on RA was tight. It seemed tight but I tweaked it a bit more.

I ran it with the cover off at high and low speeds. I couldn't see any wobble on any of the wheels. I didn't see any belt jumping or stalling. Basically it looked ok. It seemed to be running smoothly with no change in the sound. In the end I reinstalled the belt drive mod on RA checking carefully at each stage.  The belt looks ok, no malformed teeth or uneveness. I did notice that the belt was running slightly off centre. The motor cog has a shoulder at the end and the belt was running up against that. So I adjusted that so it was clear of the shoulder.

I ran it all again last night but exactly the same problem. Also on the calibration I had exactly the same paired points on RA.

So I'm thinking I'll take out the belt mod on RA and put the gears back.  If you have any other ideas I'd be very glad to hear them.

If there is stiction or binding am I right in saying that the only place that could occur is at the worm and worm wheel?  I did slacken off the backlash in RA in case I'd over done it. My understanding is that backlash in RA is not important in guiding because it's always running the same direction. I can now feel a very slight backlash in RA. 

Cheers

Steve

 

Depends if you want to rely on phd to guide it out, which looks like it's doing.. 

I still think every 2 mins is too frequent to be the worm

Looking at the graph above it looks like the spikes at getting larger and larger the more time goes on

Edited August 29, 2021 by newbie alert
Added info

[Guest]

On 12/08/2021 at 16:26, michael8554 said:

Hi Steve

There are other examples of the RA problem.

Here's part of your Calibration:

Note how the RA steps (blue) are bunched together in pairs, instead of being evenly spaced like Dec.

So there might be slight stiction that is taking two pulses to overcome.

From other posts I've read it's likely your 120 sec problem is something in the belt drive.

Try shifting the belt a couple of teeth on the cogs.

Try swapping RA and Dec belts if they're the same.

Look for signs of trying to jump while slowly slewing.

Michael

Hi Michael, just to let you know that after stripping the mount and replacing all the bearings this problem has gone and it's looking a lot better. I didn't find anything obviously wrong and the original bearings seemed ok but there must have been something I suppose.  The 120sec disturbance has gone as well. This is what the calibration looks like now.