Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Meade DS-2000 Azimuth Encoder/Motor Unit Fault

  • Please log in to reply
13 replies to this topic

#1 mdwsmith

mdwsmith

    Lift Off

  • -----
  • topic starter
  • Posts: 9
  • Joined: 10 Feb 2021

Posted 02 March 2021 - 12:14 PM

Hi Everyone.

I have a Meade DS-2090, which has proved impossible to align.


I have tracked to an issue with the telescopes Azimuth motor and/or encoder - it shows less than 360 degrees for one revolution actually 214 degrees, but its not consistent with that - when you return to the original position it does not show the start angle.  The Altitude axis does not show issues.

I have done this on a lead acid battery at 12.58V, as I've read that some power supplies give noise that can cause encoder problems ( https://www.cloudyni...tor-unit-fault/ )


To show the Alt/Azm on the Autostar, I press mode for >3s and scroll down to the Alt/Azm readout

Details


Azm Axis                        DS-2000 Azm readout

Start position   0h            0Deg00'00"

Half revolution 12h          106Deg42'02"  (should be 180degrees)

Full revolution 24h/0h      214Deg15"51" (should be 360/0degrees)


moving back to original position opposite direction

12h                               120Deg21'31"  (should be 180degrees)

0h (original position)     027Deg27'13"  (should be 0degrees)

I've calibrated the drives before the above trial.

I've had a go cleaning the Azm  encoder lens using dental floss, before this, but a fuller strip to allow a cloth/alcohol to clean may help. Also swapped around the HBX lead for the other way around.

I've also read that there has been a history of noise on the motors affecting the encoders - this dates from maybe circa 2000 ( http://www.data-plum...om/dsmotors.htm ).  This DS-2090 dates from a Costco sale circa 2010 - the firmware is the latest from 2007, and is the "Gen2" mount - silver outside on the arm, black on the inside.  I purchased it them and struggled to use goto then.  Life then happened, and its been stored until now when I have had time to fiddle!

Any ideas, suggestion including further debugging would be welcome.



#2 OzAndrewJ

OzAndrewJ

    Aurora

  • *****
  • Posts: 4,686
  • Joined: 30 Nov 2010

Posted 02 March 2021 - 03:55 PM

Gday mdwsmith

 

Sounds like you have done pretty much all the relevant tests.

If you have patched firmware, you can use my PEC editor

http://members.optus...om.au/johansea/

to read out the settings stored in EEProm

Two of these are the encoder LED settings ( that give the best readout )

What do they show as a value????

Other than that, you may need a cheap logic analyser to log

the encoder pulse train to see whats coming out over time.

Using a CRO is useless as the pulsetrain changes too dynamically to

allow you to see glitches.

 

I dont know what the DS encoder looks like, but if its the milky plastic type

what some people do is use a black marker pen to darken the fingers

 

Andrew Johansen Melbourne Australia



#3 mdwsmith

mdwsmith

    Lift Off

  • -----
  • topic starter
  • Posts: 9
  • Joined: 10 Feb 2021

Posted 03 March 2021 - 03:21 AM

Hi OzAndrew,

Thanks for your reply.  The Azm encoder is indeed a milky plastic type.  It looks like a gear but with gaps giving it a many many pointer star or petal type appearance.  I guess the presence/absence of plastic give the pulses to the IR detectors when it is moving.  Its milky to visible light, not sure what it is in the IR.  However, I do have a CD marker pen that could darken it, in visible at least.

 

My Autostar #497 firmware is 43Eg, the latest dating from 2007, it is not patched.  I like the idea of getting more information on the faults, so I'll do some reading up on patching.  I think my initial interest in diagnosing the fault(s) and fixing it rather than extra functionality.

 

I enjoy fixing things and this Meade system seems an elegant way to use cheap motors for goto, so I don't mind investigating this.  Thanks again.

Marcus


Edited by mdwsmith, 03 March 2021 - 03:23 AM.


#4 OzAndrewJ

OzAndrewJ

    Aurora

  • *****
  • Posts: 4,686
  • Joined: 30 Nov 2010

Posted 03 March 2021 - 03:38 AM

Gday Marcus

so I'll do some reading up on patching.  I think my initial interest in diagnosing the fault(s) and fixing it rather than extra functionality.

The patches are hosted on the StarPatch site

http://www.stargps.ca/downloads.htm

They are simple text files so open them in a test editor and you can read what each bit does.

Basically, the patches fix some bugs and add extra functionality.

One of the extra functions is to peek the memory, thus allowing a lot of "live" data to be investigated.

ie it allows more detailed investigation of whats going on in the Hbx memory

when trying to debug problems

 

Andrew Johansen Melbourne Australia



#5 mdwsmith

mdwsmith

    Lift Off

  • -----
  • topic starter
  • Posts: 9
  • Joined: 10 Feb 2021

Posted 05 March 2021 - 05:39 AM

Hi Andrew,
I have patched my firmware to 43GI using the latest patches on the startpatcher site.  I left off the GPS patches..
 
I struggled to find something like "encoder LED settings", read the manual and looked through the tabs - apologies for missing it.
 
Closest I got was from "list Data" tab - slected bits;
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
---- ---- 497 Classic Ver :43GI date : 03/04/2007 Run 05-Mar-2021 10:32:50
1 B600 HBx Contrast 06 6
2 B601 Alt Ratio 3F93D70A +1.15499997
3 B605 Az Ratio 3F572F67 +0.84056705
4 B609 Alt Train 06AF 1711
5 B60B Az Train 0A62 2658
6 B60D Az Motor Cal 95 149
7 B60E Dec Motor Cal 95

73 B747 PEC FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF
74 B7DD Cone Err Datum at Park FF 11111111
75 B7DE Az Cone Error at Park FFFFFFFF -1
76 B7E2 Alt Cone Error at Park FFFFFFFF -1
77 B7E6 CalSensor Alt 00000397 +00d15m19s
78 B7EA CalSensor Az 0000198E

Please let me know if you need anything else.

Thanks for this useful tool.

Best Regards
Marcus

#6 OzAndrewJ

OzAndrewJ

    Aurora

  • *****
  • Posts: 4,686
  • Joined: 30 Nov 2010

Posted 05 March 2021 - 06:14 AM

Gday Marcus

 

The two interesting bits are

6 B60D Az Motor Cal 95 149
7 B60E Dec Motor Cal 95

These indicate the LED calibration resulted in a mid range value

ie the calibration is probably OK.

Next step is to look at the encoder pulse train, and that needs a logic analyser.

 

Andrew Johansen Melbourne Australia



#7 mdwsmith

mdwsmith

    Lift Off

  • -----
  • topic starter
  • Posts: 9
  • Joined: 10 Feb 2021

Posted 06 March 2021 - 03:09 AM

Hi Andrew,

Thanks, did the usual forum/google searching on your answer.  You have recommended elsewhere a $10 logic analyzer using sigrok (would be best if I fitted in to what you're familiar with) would this be suitable and work with sigrok? (I'm not familiar with logic analyzers or sigrok, but have the time to learn and fiddle)

 

https://www.ebay.co....tUaAldCEALw_wcB

 

 

Will take 2-3 weeks to deliver.  In the UK even "UK sellers" are importing from China so delivery times are still 2-3+ weeks.....

 

For other cheap solutions I found this a good summary:

https://helpful.knob..._analyser_notes

 

For more limited solutions I have now; PC cards sound cards are frequency limited  - have a 10k linear potentiometer to scale the signals.  I also have a lolin32 ESP dev board  clone with the WROOM (not WROVER) ESP32 processor. There are a few projects that offer signal processing for this eg;

 

https://github.com/E...ll/esp32_sigrok

 

https://github.com/E...2_LogicAnalyzer

 

I also have a "Transistor Tester" clone (aimed at component testing) that offers square wave pulse generation and square wave frequency measurement on a ATmega.

 

I have access to  linux (mint) as well as windows.

 

Thanks

Marcus



#8 OzAndrewJ

OzAndrewJ

    Aurora

  • *****
  • Posts: 4,686
  • Joined: 30 Nov 2010

Posted 06 March 2021 - 05:00 AM

Gday Marcus

 

You have recommended elsewhere a $10 logic analyzer using sigrok (would be best if I fitted in to what you're familiar with) would this be suitable and work with sigrok?

The commercial unit you linked to looks like the unit i have

( They are all knock offs of the Salae units )

I know that unit works well within the Sigrok Pulseview app

and if you want to actually read and interpret the I2C comms,

my app can read the std pulseview files.

 

That said, Pretty much anything that can store and plot the pulse train will do to see

whats getting put out in the encoder pulsetrains ( and if there are glitches ).

 

It looks like some of the ESP stuff you linked to might do the same

stuff, ie Sigrok/Pulseview can read it, but thats outside my knowledge base.

Cant hurt to try it if you are happy with programming the ESP

If so, then that should also do well for std pulse train analysis

and Sigrok/Pulseview can be used to display/interpret it.

 

Andrew



#9 mdwsmith

mdwsmith

    Lift Off

  • -----
  • topic starter
  • Posts: 9
  • Joined: 10 Feb 2021

Posted 06 March 2021 - 04:14 PM

Hi Andrew,

I've ordered the Ebay unit I linked to, as its very cheap, your app and knowledge of this type of unit working with the Meade signals reduces the risks.

I may have a read/play with the ESP32 systems while I wait for the ebay unit to arrive in a few weeks.

When I have had a practice with either units (I have a signal generator)  I'll post again for any pointers in fault finding.


Thanks for your feedback.

Marcus



#10 mdwsmith

mdwsmith

    Lift Off

  • -----
  • topic starter
  • Posts: 9
  • Joined: 10 Feb 2021

Posted 17 March 2021 - 11:05 AM

Andrew,
Received the ebay 8 channel logic analyser I linked above quickly from China. Installed sigrok pulseview and logged a 2MHz squarewave signal (from my signal generator) at 48Mhz successfully - see attachment.

 

Would be grateful for some further reading, and/or hints for debugging the Meade encoder/motor system.

 

Its really cool what you can get for less than 10USD!

 

Thanks

 

Marcus

Attached Thumbnails

  • Capture.PNG

Edited by mdwsmith, 17 March 2021 - 11:08 AM.


#11 OzAndrewJ

OzAndrewJ

    Aurora

  • *****
  • Posts: 4,686
  • Joined: 30 Nov 2010

Posted 17 March 2021 - 04:10 PM

Gday Marcus

Fun little gizmos arent they grin.gif

My app can run in 2 modes for processing Meade data

 

Mode 1 does the comms and requires you to connect the probes as

D0- Motor Data

D1- RA/Az  clock

D2- DEC/Alt clock

D3- Aux Data

D4- Aux Clock

 

Mode 2 does the motor and encoders and requires the probes to be

D0- Motor PWM A  ***

D1- Motor PWM B

D2- Encoder A

D3- Encoder B

 

*** In my LPT port unit, simply using a 150k series resistance on the probe

allows my buffer chip to survive the +12V pulses.

I dont know if the USB chipset can survive even a mild over voltage so if

you do grab the PWM, its best to use a buffer or zener to drop it to 5v

 

I have found 1MHz logging works well and reduces file sizes, esp for long duration comms logging

 

Soooo, job 1 is make a breakout box for the Hbx so you can log the comms

( i just repurposed an old cat 5 cable by cut and resolder the right way round

and added a header strip for the tapping but you could also crimp your own.

Just remember, the Meade Hbx cable is directional when you add tappings like this

so it has to be fitted the correct way round )

 

For doing the PWM and Encoders, its a bit trickier for you as you have

a PCB based encoder, so need access for the tappings.

 

I suggest you first just use the analyser to log the encoder pulsetrain

and look for glitches

 

Andrew Johansen Melbourne Australia


Edited by OzAndrewJ, 17 March 2021 - 04:16 PM.


#12 mdwsmith

mdwsmith

    Lift Off

  • -----
  • topic starter
  • Posts: 9
  • Joined: 10 Feb 2021

Posted 17 March 2021 - 06:10 PM

Hi Andrew,

I hope you don't mind a few clarifying questions before I makeup my breakout box. I've number them 1-5 for clarity.

 

I found this of help for he HBX pinouts;

 

https://web.archive....rg/pinouts.html

 

 

Pinout "Autostart HBX port" on this page has almost all the pins you refer to for "mode 1" except for "motor data".

 

1) Is "motor data" A/A Data ie pins 5 and 7?  (not sure why they are duplicated)

 

 

Playing this back to check my understanding:

 

2) I note as you say the Meade HBX port is wired "straight through" unlike Ethernet cables - see Mead HBX cable on the web page above.

 

3) Also checking that the "mode one signals" are 0-12V not TTL?  If so I need to change the signal levels.  I have a Max3232 chip or I could use some 10K stereo liner pots I have... I see if I can find more data on the allowed voltage levels into my USB clone, as expected there was 0 documentation so google will be my friend.

 

4) The logic analyzer requires a ground so I'll connect that to the ground on the HBX cable.

 

 

For the breakout box, I have 2 straight through ether cable connectors (for joining cables)  I can use by taking apart one side each.  Unfortunately the T Ethernet connectors I have don't replicate all of the connections to each side.....

 

Thanks very much

 

Best regards

 

Marcus


Edited by mdwsmith, 17 March 2021 - 06:16 PM.


#13 OzAndrewJ

OzAndrewJ

    Aurora

  • *****
  • Posts: 4,686
  • Joined: 30 Nov 2010

Posted 17 March 2021 - 06:41 PM

Gday Marcus

I hope you don't mind a few clarifying questions

The more the better. If in doubt, ask.

 

1) Is "motor data" A/A Data ie pins 5 and 7?  (not sure why they are duplicated)

Yes.

Inside the mount, they are separated, but inside the handbox, they are shared, i guess to save pins on the PIC.

In the earlier 494s, the Aux and both motors shared the data line.

Just the way Meade did it.

 

I note as you say the Meade HBX port is wired "straight through" unlike Ethernet cables

Yep. Meade wired 1-8..8-1, which is why one of the diagnostic tests is to swap a handbox cable end for end.

If you make a breakout cable for tapping, this means you need to mark each end accordingly

such that you get the right signals ( ref one of my cables made from 8 core flat phone cable )

 

Also checking that the "mode one signals" are 0-12V not TTL?

Nope, In the 497s, the rs232 port runs at true rs232 levels

but the handbox clocks and data lines are all 5V TTL ( as is the encoder A/B )

The only "bitey" voltages are the motor PWM, ( which you dont really need to worry about yet )

 

The logic analyzer requires a ground so I'll connect that to the ground on the HBX cable.

Correct.
 

For the breakout box, I have 2 straight through ether cable connectors

Any RJ45 with 8 wires in it will do, just ensure when done,it is wired 1-8.. 8-1

Have attached the std piccies for the Hbx ports and the Mount port

 

Andrew Johansen Melbourne Australia

497 Splitter.jpg

Hbx Meade 497.JPG

497 Panel Front.jpg

 

 

 

 

 

 

 


Edited by OzAndrewJ, 17 March 2021 - 06:53 PM.


#14 OzAndrewJ

OzAndrewJ

    Aurora

  • *****
  • Posts: 4,686
  • Joined: 30 Nov 2010

Posted 17 March 2021 - 07:28 PM

And just for fun, have attached a sigrok file for what you should see if booting a 497 type mount.

ie its what you should see when you test your setup.

The Sigrok app saves data as *.sr files, but they are really just zips.

I renamed it to *.zip so it will load here, so after downloading it, rename it to *.sr

and you can import it to have a squizz.

Will provide a link to my app that can then decode it for you

but till then, it shows what you will be looking for.

 

Andrew Johansen Melbourne Australia

Attached File  Full497Boot.zip   216.38KB   1 downloads




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics