Upgrading 1984 Wells Index 820 with Allin1DC

All things related to Centroid Oak, Allin1DC, MPU11 and Legacy products

Moderator: cnckeith

Chevy427z
Posts: 166
Joined: Mon Apr 12, 2021 9:39 am
Acorn CNC Controller: No
Allin1DC CNC Controller: Yes
Oak CNC controller: No
CNC Control System Serial Number: A900999
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No
Location: Clinton, NC, USA
Contact:

Re: Upgrading 1984 Wells Index 820 with Allin1DC

Post by Chevy427z »

I double checked the current draw with the red lead in the right hole.

The zero is with the ALLINONE powered off, then 125.1 mA with it powered up.

I had only one DB9 plugged in at a time. X was 125.1 mA, Y was 121.1 - 152.1, and Z was 114.1 - 145.0

Then I retested the X encoder alone referring back to polar's TTL output encoder video.

A and B channels were fine at 4.01. A/ and B/ were very different from earlier findings, running .3 - .5 mV. Can

that be right? In your video you talk about zero to 1 volt.

While I have the positive lead of my meter on A (and B), and I rotate the shaft, I get the expected rise and fall

of voltage. While connected to A/ (and B/) that low voltage (.3 - .5) does not change.

Am I testing wrong (again) or is that encoder shot? That was X only, no time for all three tonight.
Attachments
20210517_174158.jpg
20210517_173947.jpg
20210517_172023.jpg
20210517_170640.jpg
M400
A900999
CNC12
SYSTEM ID - 0008DC111213-0301212560
polaraligned
Posts: 231
Joined: Mon Feb 07, 2011 8:03 pm
Acorn CNC Controller: No
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: 1030090103
DC3IOB: Yes
CNC12: Yes
CNC11: Yes
CPU10 or CPU7: No
Location: Montvale, NJ

Re: Upgrading 1984 Wells Index 820 with Allin1DC

Post by polaraligned »

Chevy427z wrote: Mon May 17, 2021 6:13 pm I double checked the current draw with the red lead in the right hole.

The zero is with the ALLINONE powered off, then 125.1 mA with it powered up.

I had only one DB9 plugged in at a time. X was 125.1 mA, Y was 121.1 - 152.1, and Z was 114.1 - 145.0

Then I retested the X encoder alone referring back to polar's TTL output encoder video.

A and B channels were fine at 4.01. A/ and B/ were very different from earlier findings, running .3 - .5 mV. Can

that be right? In your video you talk about zero to 1 volt.

While I have the positive lead of my meter on A (and B), and I rotate the shaft, I get the expected rise and fall

of voltage. While connected to A/ (and B/) that low voltage (.3 - .5) does not change.

Am I testing wrong (again) or is that encoder shot? That was X only, no time for all three tonight.
Current draw on the 3 encoders is fine.

The A and the B channels can either be 3 - 5 volts, or 0 - 1 volt depending on the position of the shaft. This voltage is measured to the power supply common (negative). As you rotate the shaft, as shown in the video, it will change and may land on either the higher voltage or the lower one depending on where the shaft stops. Bottom line is that the change occurs 1000 times for each shaft revolution, so the shaft is very sensitive- but I think you understand this and experienced it testing the encoder.

The /A and the /B should act EXACTLY like the A and B channel when measured to common. If you can't get any change on them (while turning) when measured to common, there is a problem.

Test the other ones and see if you have the same result. I have a hard time believing they all failed at the same time. I have repaired these before, matter of fact, just repaired one on a 80's HP spectrum analyzer.
martyscncgarage
Posts: 9912
Joined: Tue Mar 28, 2017 12:01 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: Yes
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: Yes
CPU10 or CPU7: Yes
Location: Mesa, AZ

Re: Upgrading 1984 Wells Index 820 with Allin1DC

Post by martyscncgarage »

Chevy427Z,
Sounds like you should contact Polaraligned and see what he would charge you to properly test your encoders.

You have spent so much time on these, I would have just replaced them all by now with new ones.
My .02 (I am less tolerant of beating my head against the wall so much ;) )

Marty
Reminder, for support please follow this post: viewtopic.php?f=20&t=383
We can't "SEE" what you see...
Mesa, AZ
Chevy427z
Posts: 166
Joined: Mon Apr 12, 2021 9:39 am
Acorn CNC Controller: No
Allin1DC CNC Controller: Yes
Oak CNC controller: No
CNC Control System Serial Number: A900999
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No
Location: Clinton, NC, USA
Contact:

Re: Upgrading 1984 Wells Index 820 with Allin1DC

Post by Chevy427z »

martyscncgarage wrote: Tue May 18, 2021 12:30 pm Chevy427Z,
Sounds like you should contact Polaraligned and see what he would charge you to properly test your encoders.

You have spent so much time on these, I would have just replaced them all by now with new ones.
My .02 (I am less tolerant of beating my head against the wall so much ;) )

Marty
Definitely good advice! As I mentioned before, tight budget, I enjoy the learning process (as long as the teacher doesn't get fed up), and while I would love to get this thing up and running, I am in no particular rush.

I hope to have some time this weekend (between house/yard work and shopping for a new dishwasher for my wife) to be able to sit down and retest my encoders. If that fails, I will either contact polar as you suggest and/or just buy new ones.

I'm unsure of the ethics of buying the cheaper ones people have pointed out to me over the Centroid units because as I understand it, this a Centorid owned/run bulletin board and I don't want to piss off those who are helping me. I'd be hesitant to but the cheap ones, too, not fully understanding exactly what I need. While the Centroid units are more expensive, there's less chance of me screwing something up.

Stay tuned! LOL

Mark
M400
A900999
CNC12
SYSTEM ID - 0008DC111213-0301212560
martyscncgarage
Posts: 9912
Joined: Tue Mar 28, 2017 12:01 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: Yes
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: Yes
CPU10 or CPU7: Yes
Location: Mesa, AZ

Re: Upgrading 1984 Wells Index 820 with Allin1DC

Post by martyscncgarage »

You are not going to upset Centroid. Though if you are going to by their encoders, you might as well buy their encoder cables all made up as well.
Reminder, for support please follow this post: viewtopic.php?f=20&t=383
We can't "SEE" what you see...
Mesa, AZ
Chevy427z
Posts: 166
Joined: Mon Apr 12, 2021 9:39 am
Acorn CNC Controller: No
Allin1DC CNC Controller: Yes
Oak CNC controller: No
CNC Control System Serial Number: A900999
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No
Location: Clinton, NC, USA
Contact:

Re: Upgrading 1984 Wells Index 820 with Allin1DC

Post by Chevy427z »

Just to make sure that I am understanding....

There is no power to the servo motors, correct?

There is no mention of it until the installation manual starts talking about the electrical cabinet layout, (which I haven't gotten to, just reading about) and no one has mentioned it yet.

Just covering my bases.

Mark
M400
A900999
CNC12
SYSTEM ID - 0008DC111213-0301212560
martyscncgarage
Posts: 9912
Joined: Tue Mar 28, 2017 12:01 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: Yes
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: Yes
CPU10 or CPU7: Yes
Location: Mesa, AZ

Re: Upgrading 1984 Wells Index 820 with Allin1DC

Post by martyscncgarage »

Chevy427z wrote: Thu May 20, 2021 3:22 pm Just to make sure that I am understanding....

There is no power to the servo motors, correct?

There is no mention of it until the installation manual starts talking about the electrical cabinet layout, (which I haven't gotten to, just reading about) and no one has mentioned it yet.

Just covering my bases.

Mark
No power to the servo motors, would fault anyway because of your encoder issues.
You have to resolve the encoder problems first.
Reminder, for support please follow this post: viewtopic.php?f=20&t=383
We can't "SEE" what you see...
Mesa, AZ
polaraligned
Posts: 231
Joined: Mon Feb 07, 2011 8:03 pm
Acorn CNC Controller: No
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: 1030090103
DC3IOB: Yes
CNC12: Yes
CNC11: Yes
CPU10 or CPU7: No
Location: Montvale, NJ

Re: Upgrading 1984 Wells Index 820 with Allin1DC

Post by polaraligned »

Chevy427z wrote: Wed May 19, 2021 3:50 pm If that fails, I will either contact polar as you suggest and/or just buy new ones.

Mark
If you want to mail them to me, I would be happy to put them thru the paces and make sure they work. No charge. You would just have to cover postage. I totally understand not wanting to buy new encoders. I am using my vintage early-90's BEI encoders and they work fine. There is a quality difference between a $100 encoder and one for $275, but it probably will not ever be a factor in an occasionally used hobby mill. The cheap ones may use printed plastic discs instead of metal or glass. ZS Instruments makes some really quality encoders with glass discs, but they are in the $275 range. Accuracy and quality come at a cost.

Scott
polaraligned
Posts: 231
Joined: Mon Feb 07, 2011 8:03 pm
Acorn CNC Controller: No
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: 1030090103
DC3IOB: Yes
CNC12: Yes
CNC11: Yes
CPU10 or CPU7: No
Location: Montvale, NJ

Re: Upgrading 1984 Wells Index 820 with Allin1DC

Post by polaraligned »

So Chevy427z sent me his encoders for testing. What I found on bench test was that one had a failed A channel and another one had a bad Z channel. Funny thing is that all the inverted channels (A/, B/ and Z/) were not working on all 3 units. So I opened them up to find that the inverted channels were all tied to ground! Crazy to have the terminals for them on the back and not have them functioning. So these were wired as quadrature encoders and not differential quadrature encoders. The key word being differential. The "not" channels are an inversion of the regular channels and used to help with noise reduction. They have nothing to do with the encoder counting. Centroid requires encoders to be differential for that noise immunity.

Image

Scratching my head because this connection is definitely factory, I looked at the datasheet again and sure enough on the last page it states that they are all grounded. Makes no sense to just bring them out as grounds and leave the labels in place.

Image

The logic is clearly there for the inverted channels so it was just a matter of finding where to connect them. In pic below I removed the jumper and ran wires to the proper spot on circuit board.

Image

As for the bad A and Z outputs, I replaced the output chip and that solved the problem.

Image

I did screw one up and could kick myself in the ass. When I did this Saturday, my wife was just home from surgery on Friday and friends arrived upstairs so I was rushing to get the one I was working on back together. I did not clip the leads after I soldered in the new chip and when I tightened the board down, the leads hit the glass and broke it. Grrr....

Anyway, I made it better by Machining out the old housing to fit a 1000 line BEI encoder that is surplus to me. I epoxied it in. Reason I did this is so the encoder will mount same as the old one did. Chevy427z will just have to connect the encoder leads to the pigtail.

Image
Image

I connected the 2 repaired original ones to my machine to make sure they were working ok. Results are in the video below:

martyscncgarage
Posts: 9912
Joined: Tue Mar 28, 2017 12:01 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: Yes
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: Yes
CPU10 or CPU7: Yes
Location: Mesa, AZ

Re: Upgrading 1984 Wells Index 820 with Allin1DC

Post by martyscncgarage »

Polaraligned you are a saint to take a look!
Chevy427z...hope it all works out after all that. You are indebted to him if it does!
:D
Reminder, for support please follow this post: viewtopic.php?f=20&t=383
We can't "SEE" what you see...
Mesa, AZ
Post Reply