Jump to content

Narrowband

EQ6 belt drive modification


George

Recommended Posts

Sounds good... and thanks you.

At least I'm hoping to use this as a comparison for when I eventually sort out a suitable belt drive system for the HEQ5. If I choose the same star I can then have a direct comparison.

Link to comment
Share on other sites

Hi Malcolm,

I wouldn't worry too much that looks quite typical to me and shows the effects of the non harmonic standard gearing. I'll write up a short analysis report to explain exactly how to process and interpret the data within PECPrep.

Chris.

I look forward to that Chris, it's alright getting data (no chance here tonight its blowing a gale) but the knack is deciphering it :(

Im a bit baffled by the worm harmonics.....8 of them :) if you could explain what each of the harmonics refers too it would be easier to look for the problems that relate to them.

Edited by George
Link to comment
Share on other sites

Malcolm,

Please find attached a pdf detailing how I processed your PHD data in PECPrep. As I suspected from your screenshot the stepper gear and transfer gear are making quite significant contributions to your overall PE and this is why your data appears quite non-periodic in nature (the least common multiple of the stepper gear, transfer gear and worm periods ensure that the no repeatable pattern within the period of one complete axis rotation.

The gear mesh jitter is still present although is of significantly lower magnitude than the signals. I know that some folks have reduced this by polishing their gears. Given that your guiding resolution when using the ST80 and QHY5 is around 2.7 arc secs per pixel (I think I have that right) and this jitter is coming in at only around 1.1 arcsec peak to peak it may be that most of it falls within PHD's minimum movement threshold and this would explain why it was still present in the guided log you took.

Chris.

Edited by chrisshillito
Link to comment
Share on other sites

Chris, thanks for that. I'll go through it in detail later as I'm off to work in a few moments. One thing I noticed is that the trend value through the 200P was zero, where as the ST80 gave a value suggesting a miss-polar alignment. Am I right in thinking that due to longer focal length the tolerance would be tighter on the 200 than the ST80, and thus would be more sensitive.

Could it be that the ST80 isn't mounted 100% true in all axis to the 200P's optical axis ? - If the 200P is out of alignment, in which direction is it out (i.e alt what set of bolts need adjusting)

Thanks for your help on this Chris

Link to comment
Share on other sites

While I'm no expert, it certainly seems to me from what you say Malcolm, that the ST80 is miss-aligned while the 200P is correctly polar aligned. So the ST80 is not exactly aligned with the 200P. I guess getting the two scopes exactly parallel would be quite difficult.

Edited by Gina
Link to comment
Share on other sites

AFAIK there's no need for Critical alignment between the Guide scope and the imaging scope.. infact you might have to deliberately move the guidescope to find a guidestar - although with a decent sensitive guidecam you very rarely need to...

What is important is that there should be no relative movement between the two during a run.....

Peter...

Link to comment
Share on other sites

One thing I noticed is that the trend value through the 200P was zero, where as the ST80 gave a value suggesting a miss-polar alignment. Am I right in thinking that due to longer focal length the tolerance would be tighter on the 200 than the ST80, and thus would be more sensitive.

Could it be that the ST80 isn't mounted 100% true in all axis to the 200P's optical axis ? - If the 200P is out of alignment, in which direction is it out (i.e alt what set of bolts need adjusting)

Sorry I messed up and turned accidentally turned off the linear regression checkbox for the 200P data - that's why the trend gradient shows as 0. Attached is an updated pdf - the gradient was actually 0.0181 arcsecs per sec. Its difficult to say for sure why the two gradients should differ.

One thing that is noticable is that the raw data (grey curve in PECrep for the 200P does show a steady gradient where as in the ST80 it starts off roughly flat and only after a cycle and a half doe it move upwards at a steady rate. No dead what that means though.

I suppose it could be an optical axis alignment error but then you also have to bear in mind that drift rates due to polar misalignment are not constant and vary depending on where in the sky you are pointing. So over the two hours of capture time you might expect a continuous change in gradient.

Also it may be that what we're seeing as a linear drift over the capture period is actually fractions of long period periodic signal such as could be contributed by the roller bearings supporting the axis.

Another factor may be the accuracy of PHD's calibration in determining the camera angle. I guess the greater the pixel movement produced during calibration the more accurate the angle measurement will be but then again if its blowing a gale you might find that the higher resolution of the 200P is more susceptible to calibration errors due to a wind gust. If you analyse the DEC data for the 200P you will see that there are the characteristics peaks in there associated with the Worm transfer and stepper gears albeit with small amplitudes. As the DEC drive was not moving this can only mean that PHD hasn't quite got its angle calculation right and that a small component of RA movement is being mapped into the DEC data. This hasn't happened for the ST80 data.

Chris.

malcolm_pecprep.pdf

Link to comment
Share on other sites

Hi George,

There are some configurations changes you will need to do to for PECPrep to properly handle your modded mount.

First open up the marks.def text file in notepad (this is in the PECPrep install directory) and add the following to the bottom.

BeltDrive2.png

Use tabs as separators. This file determines what periods are significant for your mount - you can add more entries to the list if you discover new characteristics you want to quickly identify.

With PECPrep running go into the Mount menu and select edit types - a new window will appear. Select one of the "undefined" slots and fill it the data in as shown below.

BeltDrive1.png

Then hit the Apply Changes and then when you close the window your modded mount should appear in the PECPrep mounts menu.

Chris.

Link to comment
Share on other sites

Im a bit baffled by the worm harmonics.....8 of them :) if you could explain what each of the harmonics refers too it would be easier to look for the problems that relate to them.

Hi George,

I guess the easiest way to view harmonics is to think of them as echos of a fundamental frequency. In the real world the error signals coming from rotating elements such as pulley as gears rarely produce perfect sinuosids - slight imbalances or eccentricities will produce signals harmonic with the fundamental and that are modulated by it. The fundamental will always be the most significant signal but adding in lower amplitude harmonics can significantly change the basic shape of the overall signal. You can see this effect by playing with PECPrep's simulator. For instance if you add only odd harmoincs (i.e fundametal divided by 3, 5, 7 etc.) you'll find your fundametal sine wave is quickly turned into a square wave! Mechanically the presence of odd harmonics might therefore indicate that there is some restriction of movement in a rotating part.

Anyhow, the point is that the FFT spectrum analyser will break down the input signal into combinations of sine waves. Because the worm fundamental generally by the highest amplitude signal present its harmonics are also likely to show up in the spectrum as significant peaks and that is why I gave the option to quickly identify up to the 8th harmonic.

How you treat the harmonics is up to you (vibration analysis is a rather involved subject!). If your doing comparative tests then a reduction in harmonic magnitudes would represent a mechanical improvement to that particular component - that may be all you need to know. Alternatively you may want reconstruct the error signal of a particular mechanical component by using the simulator set to use the phase magnitude and amplitude of each fundamental/harmonic associated with that component.

I do hope I haven't added to your bafflement!

Chris

Link to comment
Share on other sites

Thanks Chris...

Is the 0.018 arc seconds a lot in the scheme of things ? - I polar aligned my scope using the built in alignment part of EQMOD, and then used the dSLR drift method (produces V trails if out of alignment) to try and dial it in.

@gina and nick - glad you guys like this... all these numbers and nice graphs look good, I'm failing to grasp this and make sense of it all.

Link to comment
Share on other sites

One of the first things you learn is that a square wave is made up of loads of sine waves. You can make various shaped waves by adding together different frequency waves. I.e. You can make a triangular wave.

All of this is exactly related to this but instead of generating all the sin waves and adding them together our mechanical system is introducing its own repeating error...each of these repeating errors is like my individual sine waves on the electronic version.

These repeating errors all add together to create our PE or repeating total error.

All the maths does is take the error curve you produce from tracking the star movement and basically calculate each of the repeating errors (the frequency of these) and their sizes (magnitudes).

Then someone like Chris who actually knows what he's talking about, unlike me, can tell us which thingy is related to what mechanical wotsit.

This and induction motor theory are two bits of engineering maths that i thought was...elegant.

Link to comment
Share on other sites

Thanks Chris...

Is the 0.018 arc seconds a lot in the scheme of things ? - I polar aligned my scope using the built in alignment part of EQMOD, and then used the dSLR drift method (produces V trails if out of alignment) to try and dial it in.

@gina and nick - glad you guys like this... all these numbers and nice graphs look good, I'm failing to grasp this and make sense of it all.

To me a 0.018 arcsecs per sec drift isn't anything to worried about. Your guider is going to correct for it anyhow and there is no guarantee that it is all down to polar alignment. PECPrep really doesn't tell you if something is bad or not - it simply presents graphs and statistics based upon the data you give it. Ultimately its up to you to decide what, if anything, is having a detrimental effect on the Astronomy you want to do.

Chris.

Link to comment
Share on other sites

Chris I want to thank you for taking the time todo this :) , hopefully when the weather cooperates I can grab that important data (to me at least) and find out how much improvement I have gained. But TBH I'm more than happy to have turned a dog mount into something that has no more backlash and tracks better than I could have imagined.

Hopefully at some point belt modded EQ6/HEQ5 users will reach critical mass when they realise that a simple modification can overcome one of the mounts most fundamental shortcomings.

I'm pretty sure a belt driven mount would have been cheaper to manufacture than a geared mount god knows what Synta was thinking.

Thanks again from a EQ6 user for EQmod and all the secondary programs that run with it :(

Hi George,

There are some configurations changes you will need to do to for PECPrep to properly handle your modded mount.

First open up the marks.def text file in notepad (this is in the PECPrep install directory) and add the following to the bottom.

BeltDrive2.png

Use tabs as separators. This file determines what periods are significant for your mount - you can add more entries to the list if you discover new characteristics you want to quickly identify.

With PECPrep running go into the Mount menu and select edit types - a new window will appear. Select one of the "undefined" slots and fill it the data in as shown below.

BeltDrive1.png

Then hit the Apply Changes and then when you close the window your modded mount should appear in the PECPrep mounts menu.

Chris.

Link to comment
Share on other sites

Well, my machinist has worked his magic and i have everything back and ready. Now where is that extra small allen key...er....that really small easy to lose Allen key....oh darn it.

And none of the shops have one that small..

I hate mechanical thingies...

Link to comment
Share on other sites

Well, my machinist has worked his magic and i have everything back and ready. Now where is that extra small allen key...er....that really small easy to lose Allen key....oh darn it.

And none of the shops have one that small..

I hate mechanical thingies...

I couldn't find mine either but I have one of those micro screwdriver sets and found a flat bladed one I could jam in :D

Link to comment
Share on other sites

Well, my machinist has worked his magic and i have everything back and ready. Now where is that extra small allen key...er....that really small easy to lose Allen key....oh darn it.

And none of the shops have one that small..

I hate mechanical thingies...

LOL - try RC model shops.

Link to comment
Share on other sites

I couldn't find mine either but I have one of those micro screwdriver sets and found a flat bladed one I could jam in :D

Aahhh now thats a whole new set of tools i can try....after dinner..

...and if that fails "to the model shops!". If they have them over here.

Link to comment
Share on other sites

Well at last we had a clear night and it was time to play :)

Spent an hour tweaking the polar alignment with EQalign and eventually got the green & red circles merged as one even after 20 minutes of tracking. Ive marked the pier head and mount up now with some tippex so in the future if I have to remove the mount from the pier it will be a doddle to get everything up and running quicker.

I then parked the scope up and resynced the steppers and created a new point list, after the first 3 stars I got every slew in the centre of the crosshairs on the QHY8. Slewing accuracy has definitely improved with the belts so a big plus there for the conversion.

post-13224-133877685368_thumb.jpg

post-13224-133877685374_thumb.jpg

Now it was time for the unguided PHD data :icon_salut:

Fired PHD up disabled guide output and slewed to my chosen star, I wanted something bright that would show thru some high level cloud at the time.

Guide scope ST80 FL400mm

Guidecam QHY5 Pixel size 5.2x5.2

Object name: Enif

Magnitude: 2.38

Equatorial: RA: 21h 44m 47.291s Dec: +09°56'02.723"(current)

Equatorial 2000: RA: 21h 44m 11.158s Dec: +09°52'30.041"

Horizon: Azim: 227°03'59.883" Alt: +37°50'48.433"

Bit bright for guiding really but I didn't want any cloud losing me any data :) Must say even with perfect polar alignment (well as perfect as I could get) after the first few minutes PHD's guide graph started some horrible mountaineering and I couldn't watch and left it todo its thing, came back an hour or so later & the target star was still more or less centred in PHD's bullseye.....here's my PECprep graph.

post-13224-133877685361_thumb.jpg

Looks pretty horrible unguided :)

post-13224-133877685351_thumb.jpg

I edited my pecprep ini and added the mechanical modifications as Chris suggested, on the above graph they are ticked to highlight their contribution which to my untrained eye looks negligible.

The biggest culprit still appears to be the worm which I checked for runout while everything was dismantled, runout was 0.001mm which in engineering terms is as perfect as you will get.

Ive attached my unguided log perhaps Chris can comment :D

PHD_log_28Oct1_testdata.txt

Link to comment
Share on other sites

I then parked the scope up and resynced the steppers and created a new point list, after the first 3 stars I got every slew in the centre of the crosshairs on the QHY8. Slewing accuracy has definitely improved with the belts so a big plus there for the conversion.

Nice report George, and for the above reason alone it makes sense to do the belt conversion IMO

Link to comment
Share on other sites

It'll be interesting to see what Chris says.

I have ordered a set of 30 Allen Keys so one of them should allow me to put it all back together.

I have put the pully on the worm gear but its really jammed on. I understand that the thermal coefficient of expansion for Aluminium is larger than steel so should i be able to get it off after sticking it in the oven for a bit and smacking it with a rubber mallet?

Its on perfectly for trying everything out i am just preparing myself for if i need to put the gears back on.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.