¿ªÔÆÌåÓý

Spikes!


 

Dear Jamie,

Because there is a possible interaction between PHD2 and the Gemini doing the RA motor drive, I think you must eliminate the chance that PHD2 is generating the spikes.
I think you just use PHD2 to track a star (starting near equator and meridian intersection) and shut off the PHD2 autoguiding pulses.
This will show what the mount, tracking in sidereal only, is doing.

If that star tracking shows glitches, then they are for sure coming from your RA drive and nowhere else.

If the star tracking alone shows no glitches, but PHD2 autoguiding does show them, then something is wrong in PHD2 autoguide settings.?

Under the Brain icon, there is a checkbox for "Allow autoguide pulses"... just uncheck that box to shut off the autoguide pulses.?
you do that after you Calibrate, so PHD2 knows E, W, N , S directions.? And you can take your PHD2 log file then an insert it into the PE evaluation software like PECprep to get the PE and to look at the graph of the periods.

Very best, and good luck!
Michael


On Thu, Apr 1, 2021 at 10:55 PM Jamie Amendolagine <jamie.amendolagine@...> wrote:

I've been having what is becoming a frustrating experience with large RA spikes. I've lost track of how many nights that I've focused on just getting guiding under control, maybe 5 or so. I've switched my setup from my APO to my SCT for galaxy season, so there's a bunch of recent changes, but all of the equipment was used before and performed pretty well. This exact setup has performed performed around 1.1 to 0.7 RMS, and this was before multistar-phd2. Right now I'm having a hard time keeping it below 2.0 RMS. There are some huge spikes (10 .. 14 or so) happening somewhat regularly, but not on a precisely regular schedule. Here's some examples of the time gaps between the spikes: 197s, 247,? 412, 485. The setup is an EdgeHD8 with a 0.7 focal reducer and an OAG. 178MC camera for guiding. I'm thinking of switching to the guide scope just to see what I can find. If everything is great there, then maybe there's some sort of flexure in the OAG image train.? I've gone through things carefully, making sure my PA is good, balance is good, wires are not snagging etc. Help!!!

This is not the infamous 76 second spike is it???

It's such a strange spike, maybe there's something going weird with the camera? I'm going to try and guide with the main camera just to see what I see.?

Help!



--
Michael Herman
mobile: 408 421-1239
email: mherman346@...


 

¿ªÔÆÌåÓý

Hi!

A few details on these spikes:

Looking at Jamie's guidelog, it seems that for each spike in RA, there is a small pulse in RA in the opposite direction of the spike. I can not get my head around this - how does that resonate with the "stuttering"? (I mean maybe it does)

(As for the idea that PHD2 could have caused the spikes based on prediction: the pulse is very small, and the spike seems to me unproportional in relation to how the scope has reacted to other pulses. )

I my log, that I attach here as well, I have small RA pulses in the same direction as the spike. That made me think my case was too well balanced and "backlash" in RA, so the small pulse sent the scope a disproportionate distance. However, the scope was clearly east heavy.

So: is there a way to make sense of the small RA pulses at the start of each spike, in relation to the two (as I understand it) hypotheses about spikes: the "stuttering" caused by imbalance, and the "too well balanced"/RA "backlash"-issue? And/or are there more hypotheses about this on the mechanical side (for the moment not thinking of PHD2-caused issues)?

Best,

Magnus


Den 2021-04-08 kl. 12:13, skrev Michael Herman:

Dear Jamie,

Because there is a possible interaction between PHD2 and the Gemini doing the RA motor drive, I think you must eliminate the chance that PHD2 is generating the spikes.
I think you just use PHD2 to track a star (starting near equator and meridian intersection) and shut off the PHD2 autoguiding pulses.
This will show what the mount, tracking in sidereal only, is doing.

If that star tracking shows glitches, then they are for sure coming from your RA drive and nowhere else.

If the star tracking alone shows no glitches, but PHD2 autoguiding does show them, then something is wrong in PHD2 autoguide settings.?

Under the Brain icon, there is a checkbox for "Allow autoguide pulses"... just uncheck that box to shut off the autoguide pulses.?
you do that after you Calibrate, so PHD2 knows E, W, N , S directions.? And you can take your PHD2 log file then an insert it into the PE evaluation software like PECprep to get the PE and to look at the graph of the periods.

Very best, and good luck!
Michael

On Thu, Apr 1, 2021 at 10:55 PM Jamie Amendolagine <jamie.amendolagine@...> wrote:

I've been having what is becoming a frustrating experience with large RA spikes. I've lost track of how many nights that I've focused on just getting guiding under control, maybe 5 or so. I've switched my setup from my APO to my SCT for galaxy season, so there's a bunch of recent changes, but all of the equipment was used before and performed pretty well. This exact setup has performed performed around 1.1 to 0.7 RMS, and this was before multistar-phd2. Right now I'm having a hard time keeping it below 2.0 RMS. There are some huge spikes (10 .. 14 or so) happening somewhat regularly, but not on a precisely regular schedule. Here's some examples of the time gaps between the spikes: 197s, 247,? 412, 485. The setup is an EdgeHD8 with a 0.7 focal reducer and an OAG. 178MC camera for guiding. I'm thinking of switching to the guide scope just to see what I can find. If everything is great there, then maybe there's some sort of flexure in the OAG image train.? I've gone through things carefully, making sure my PA is good, balance is good, wires are not snagging etc. Help!!!

This is not the infamous 76 second spike is it???

It's such a strange spike, maybe there's something going weird with the camera? I'm going to try and guide with the main camera just to see what I see.?

Help!



--
Michael Herman
mobile: 408 421-1239
email: mherman346@...


 

Ok...you now feel the "stuttering" the same way I did.? You get it!
?
Some things to consider...experiments if you want to try them....
?
The rotation direction of the stuttering changed when I changed direction of the RA preload.? That is, when I East biased the RA with a cord hanging East, the stuttering was in the Eastward rotation, but sidereal tracking Westward direction was smooth. If I switched to a West side bias the vibration was in the sidereal direction ....so my conclusion was to East side bias, as everyone has also recommended.
It's exactly the same as what I see.? My imbalance was caused by the focuser plus DSLR and the SSAG.? When I turned the OTA in its rings to the other side the same thing happened just with the directions reversed.

Further experiments with Superlube and a CRC brake caliper grease showed different results. Lubing the worm and ring gear with Superlube was ineffective in treating the stuttering.? The vibration went away when I loaded in CRC Caliper grease...this has 3 lubricants to eliminate brake "chatter": MolyD, graphite, and teflon.? My chemical tests show no reaction with brass.? I suggest you get some and try it out.?
I'll see if I can find that.? I have no doubt that the lack of lubrication makes things worse.

Why would the ring gear chatter in one direction only when there is a axis torque force, than the other?? The answer has to do entirely with friction and whether you are pushing or pulling at an angle.
?
As an example you can try, if you take a piece of chalk, angle it like this ___/___ on a blackboard.? If you drag the chalk to the right there is no chatter.? If you push the chalk to the left it will chatter.? There is a critical angle of the chalk to cause chatter.? That angle depends on the coefficient of friction between the surfaces.??
?
What is the effect of the worm on the? ring gear teeth???
?
The effect of the hanging weight preload changes the side of the worm thread that interacts with the side of the ring gear tooth.? If the worm edge is dragging, there is no stuttering ...like the chalk in the blackboard example.? But if the worm edge is trying to push the ring gear, you get the chatter like the chalk being pushed against the blackboard.? ??
When the worm pushes the ring gear surface towards the imbalance side, the point of contact of the teeth experiences less pressure and it is easier for the ring NV gear to jump forward to mesh with the reverse side of the next tooth.? When it bumps into that it bounces back, and so on.

When the worm pushes the ring away from the imbalance there is much more counterpressure provided by the ring gear and this jumping does not happen.

True...but what about the possibility that the DEC axis stutters?? It is not moved if you have perfect Polar Alignment, but it gets moved by autoguiding.? ?It can be moved in either direction.? Can that possibly vibrate?
Out of the question.? My test was done with the motor coupling disengaged and the motors not running, DEC was doing nothing.? I turned the worm with my thumb.

Is there still a vibration from the moving worm to the (elastic plate) ring gear due to friction? That is, if the worm is binding to the flat plate ring gear due to friction, can the up or down rotation of the rotating worm lead to a vibration of the large diameter ring gear?? Remember the ring gear is larger than the thrust bearing under it.? You must consider the ring gear to be like the surface of a drum...it can vibrate it there is a rubbing (like the horsehairs of a bow on a violin string).??
Could be.? It could also be the counterweight bar that hits a resonance.? I will attach a laser pointer to various parts to find out if my spring loaded worm actually remove the backlash.? I can also do the same test, attach the pointer to the counterweight or the scope to see which part is vibrating most.

I know that users are not supposed to mess with the blocks as Brian suggested.? But why not if it is done with common sense.? There is only one DOF that can mess things up but if there is symmetry in the test that I did that can be ruled out.? To be precise, if we call the base plate on which the blocks sit horizontal and label the base tangential to the ring gear as the X direction, the radial direction towards the ring gear Y and the near block hole center the origin, then each block has 3 DOFs.? For instance the near block can move in X, Y and rotate in XY around the origin.? But the far block is aligned with the near block by the plate cover so we go from 6 DOFs to 3.? Assuming the worm meshes with the ring gear by the spring loading, we lose 1 more DOF so there are 2 left.? One of them is the worm axis angle with the ring gear.? Since the ring is circularly symmetric it should not matter what this angle is, if it is fixed.? The other DOF is a translation in the X direction that affects the angles that the teeth profile of the worm and ring gear engage at.? That type of musalignment would be bad, and it will be asymmetric between going towards one side or the other.? But since the stutter effect that I experienced is symmetric in both directions, that suggests that this misalignment is not the case.? In other words I don't believe that what I see is caused because I misaligned the blocks.

There's still plenty else that can go wrong.? BTW I put 2 Belleville washers under the hex screw of the near block that pull the block down flat against te plate so the motor can no longer wobble it.? It seems to be working.? The spring loading is very tricky altogether though.? Balance becomes much more critical.? I have seen that if the balance is not perfect, the ring gear can push the worm away with different force depending on where the OTA is pointing.? So there must be enough pressure to prevent this.? But the pressure also causes friction ghat is bad for imaging, so we want the least amount of pressure.? That can only be done with perfect balance.? If it gets too finicky I'll just revert to the old way and live with the backlash.?

Enough of this.? Work is calling, it will be Friday night when I can resume testing.


 

So, yes the spikes do happen, and look the same when guiding is disabled. The part that confuses me is why I have never seen these before. My setup has become very routine. I can't imagine my balance being much different than before. This spike is very persistent, and only disappeared after balancing to the east. I've always tried for perfect balance with this mount. Hmmm. I did only see the spike disappear for a period of about 20 min, so maybe it was just a Lucky 20 min, and it will come back on another east heavy run. I have one more log showing it dissapear when I balance east. I'll post it in a bit.?


 

¿ªÔÆÌåÓý

Hi!

Yes, please post the log. And: in which direction was the scope aimed when you did not have any spikes? And when they returned, where was it aimed?

Magnus


Den 2021-04-08 kl. 16:44, skrev Jamie Amendolagine:

So, yes the spikes do happen, and look the same when guiding is disabled. The part that confuses me is why I have never seen these before. My setup has become very routine. I can't imagine my balance being much different than before. This spike is very persistent, and only disappeared after balancing to the east. I've always tried for perfect balance with this mount. Hmmm. I did only see the spike disappear for a period of about 20 min, so maybe it was just a Lucky 20 min, and it will come back on another east heavy run. I have one more log showing it dissapear when I balance east. I'll post it in a bit.?


 
Edited

Here's the logs. So I found when the scope was pointed south to the equitorial line, where I did my calibration, there was no spikes. The spikes only showed up about a half hour after tracking, so pointing a bit west.?

This log shows guiding with a east bias to the balance. The spikes are gone. I posted the log with the guiding turned off, and the spikes showing earlier.?


 
Edited

Looking at my old run, I can see a 30min separation between the spikes, so the East bias may not have fixed it. It might just be luck that it did not show up before I turned in.?
This does not show 30min separation. I should look closer. These are shorter intervals, more like 3 to 5 min.?

https://photos.app.goo.gl/eEnXxydGK8E5Lyyu9


 

So a couple of things.

  1. So far it seems that east biasing fixed the issue. So since I was not having issues before, maybe the worm has come loose with spring? My current adjustment is also pretty loose, so maybe my worm is just a bit too loose, and if I tighten it up (not to much!) things may go back to normal.?
  2. I just noticed that the counterweight shaft was not tight. I don't know if it has anything to do with the spike, but it can't be good either way. I tightened it up, and we'll see what happens tonight. I'll try first with my best balance, and the counterweight shaft tight. Then with east bias, and finally with an adjustment of the worm making it a bit tighter, but with best balance.?
  3. Profit! :)


 

Very nice reports, Jamie!

Especially the good news that you are getting things working as you expected.

It's like a car...when it's working you enjoy the ride.? When the check engine light comes on....you have to check each item to see what's?going on.? "Tighten counterweight rod" was not on my checklist.? For sure if that's loose it could rattle.

The East side bias was something I did not need on my older direct drive G11 mounts.? But it helps a lot on the G11T, and I can see now that the design requires it. Only in the case of perfectly round gears and an unintentional imbalance and nearly zero worm friction can you get away without it.? That's too many lucky coincidences.? Murphy has a law that must be obeyed (used to be called Entropy).?

I got my G11T PE to 1.7 arcsec RMS.? I'm able autoguide 500 sec with round stars.? My stars are going out of focus as the night air temp changes and now I need to use the full capability of the motor focuser to temp compensate.? More to learn...? but the tracking us smooth and quiet and autoguiding is working fine.??

Stay well and have fun...but no sleep in the night ahead.

Michael




On Thu, Apr 8, 2021, 10:50 AM Jamie Amendolagine <jamie.amendolagine@...> wrote:
So a couple of things.

  1. So far it seems that east biasing fixed the issue. So since I was not having issues before, maybe the worm has come loose with spring? My current adjustment is also pretty loose, so maybe my worm is just a bit too loose, and if I tighten it up (not to much!) things may go back to normal.?
  2. I just noticed that the counterweight shaft was not tight. I don't know if it has anything to do with the spike, but it can't be good either way. I tightened it up, and we'll see what happens tonight. I'll try first with my best balance, and the counterweight shaft tight. Then with east bias, and finally with an adjustment of the worm making it a bit tighter, but with best balance.?
  3. Profit! :)


 
Edited

After over a week of scratching my head. Trying this and trying that, I'm back to imaging again. Target tonight is M53. I would go for a galaxy, but I'm one globular cluster short of getting my RASC Deep sky imaging certificate. Not that I need that for any reason, but it's just fun, and I able to get my kids excited about doing a wide-field certificate.?

I started out with my best balance, no east biasing, and saw some terrible guiding. I tightened up the RA spring, and the stop screw twice. Just a bit really, and my guiding is around 1 RMS, which should be good enough for this target. At some point when I want to tweak things further I'll return to east biasing, but for now I'm taking pictures!

Thanks everyone!!!!

Jamie


 

On Thu, Apr 8, 2021 at 07:24 AM, Henk Aling wrote:
Could be.? It could also be the counterweight bar that hits a resonance.? I will attach a laser pointer to various parts to find out if my spring loaded worm actually remove the backlash.? I can also do the same test, attach the pointer to the counterweight or the scope to see which part is vibrating most.

I can save you some time right now by telling you not to bother.? A few arc-seconds of vibration translate into a few x 10^-2 mm of movement on the laser pointer spot.


 

On Sun, Apr 11, 2021 at 09:57 PM, alan137 wrote:
On Thu, Apr 8, 2021 at 07:24 AM, Henk Aling wrote:
Could be.? It could also be the counterweight bar that hits a resonance.? I will attach a laser pointer to various parts to find out if my spring loaded worm actually remove the backlash.? I can also do the same test, attach the pointer to the counterweight or the scope to see which part is vibrating most.

I can save you some time right now by telling you not to bother.? A few arc-seconds of vibration translate into a few x 10^-2 mm of movement on the laser pointer spot.
Indeed.? It can be useful to test at slew speed to see how much everything wobbles with spring loaded worms, and at 8x I can set a 4 second bound on the 1x backlash, perhaps less.? Looking through an eyepiece at a remote hillside is more effective.

I'm getting 1.5" total RMS tonight with the 12", not hopeless but not what I hoped for.? I'm also testing the new 2600 and CC so I'm focusing on other things.