How to check Motor Encoder 2

[Bucknut]

I am looking for a method to determine if the encoder mounted to the motor in a production situation is good or not,where 1 minute downtime = $10,000. These encoders and not field changeable. The motor manfacturer is Yaskawa. I have 2 types
1) Encoder Type "Absolute"
2) Voltage "+5vdc"
3) Communication "Serial"
4) Pulses per Rev "1024,2048,8192" "12&15 Bit"
5) Position Count "Signed"
6) Discriminate over multiple rotations

The 2nd type is the same except:
3) Communication "Analog

 

[jOmega]

Why not ask Yaskawa ?
Would you think that because "they" made it; they are in the best position to give you a definitive answer ?
hmmmmmmm.

 

[Bucknut]

I hope you would not think that asking the motor manufacturer was not my first step before trying to get help from people I think might have came accross this same senario in other mfg facilities !!!!!!!! They only repair the motors in Japan and do not make the encoder itself and will not tell me who does, so if you have viable infromation please reply !!!!

 

[AT1000]

Hello
I understand what problems you are having as I work in a servo motor repair shop and get all makes of motor. Siemens, Fanuc, Indramat,Control Techniques to name just a few and by far the worst for information is Yaskawa because they simply do not give anything but new or repair in Germany. I have in the past sent the encoder to a manufacturer who will also struggle for a test as without their diagnostics it's a losing battle as all checks out o/k you give it back to the customer and you guessed it won't work. Put in a new motor and away you go. Some Yaskawa have incremental encoders which are straight forward but the absolute ones also carry commutation channels for the motor alignment so don't move it. I am in England and so must send for repair th Germany. Inverters are o/k but the servos are quite different.Could also be a multi-turn so even has more info.Wish I could more help.Yaskawa strikes fear into any servo shop. Believe me.
Good luck
AT1000

 

[Bucknut]

Thank you AT1000! I have located 2 devices via google search.
1) Is a RIM Tach M100 Encoder Tester that will work for my abso enc with A,B and Z signal from

http://www.northstar-tec.com

2) Is a USB1 Encoder Data Acquisition device from USDigital.
I am hoping that this will work for my serial data pulse train encoders but I'm not sure that it will.

If anyone has used these units before please let me know if they work.Northstar is local so next week they are going to give me a demo.

.....And your right the motors come back from repair some 3-6 months later saying No Problem, but we all know in the industry thats what the problem was, because we change the Servo Pack (Drive) first, check cables/change cables and then last Change the MOTOR/Encoder because with Yaskawa I cannot just change out the encoder seperatly.

 

[automatic2]

Do you have a spare encoder?. Strip it down and draw up a schematic. Jump on with a meter or scope and check your channels. No big deal. If your channel info is getting back to the resolver, check your error plus sign signal to your power driver. Also, make sure your drive amp circuit is properly setup. No offense but the biggest problem for encoded DC is a motor repair shop. When you start having problems with your drive, clean the comm. and reseat the brushes.

 

[Bucknut]

Thank you automatic2, Yes I do have a spare encoder to work with. In a test bench senario I think we can do well with a scope, but what I'm really after is a quick method (must be under 5 minutes)for any level of maintenance tech. can check without engineering support.We have 3 shifts of maintenace and engineering only works first shift.I was thinking some small hand held battery powered device with Green flicker LED for communication O.K. and a Red LED for N/G. I just thought something like that would already be on a shelf. I can design my own, but we are nearing our Winter Shutdown in which we will be installing new equipment so I can't start on this until Feb.Also we only run Brushless stlye motors.

 

[AeroPD805]

Hi,
I used to service & repair Yaskawa servo motors and though I'd offer my
.05 worth.
I found a long time ago Yask offers 0 support because they'd rather sell
you a new replacement. Typically there encoders require special tooling to
remove and/or install.
If you have an absolute positioning multi-channel endcoder it's very
important that the encoder disc is indexed to the shaft before
reomoval!
Or in some cases the disc is indexed to the commutator and will need to be
indexed electrically.
Also, Yaskawa likes to thread-lock all the fasteners for reliability, but this also
makes the motor hard to service. Apply heat to the fastener(s) to access
the motor.
For testing the encoder I used to use a cable adapter and an O-Scope with
a shaft indexing strobe. The cable adapter typically required current limiting
resistance for the led and resitive pull-ups for each channel.
Hope this helps...

 

[Bucknut]

Thank you AeroPD805 for your input. The encoders in which you decribe are exactly the type we have Approx 3,000 of them in my department only, not to mention 4 othr departments and 10 Plants in North America.
I will try to use our Fluke ScopeMeter (because of portability) and I did not know about the resistive pull-ups for channels, and all I have to do is add an external 5vdc supply. Correct ?

 

[AeroPD805]

BuckNut,
Yes, on the 5v supply.
re: resistors; depending on the model encoder you have, the resistors
for the channels and the led may already be included on the encoder
pcb, but don't assume... I'd typically start with 4.7k using a decade
box for the channels and then adjust the resistance to get a clean
TTL signal. Once the best resistance is found, find a standard resistor
that's close and build that into you cable adapter.
Just curious how many channels do these encoders have, and is there
an index pulse?? Once up on the scope, the index pulse should line
up with one of the channels.
I've seen absolute encoders up to 9 channels!
Good Luck...

 

[Bucknut]

AeroPD805,
I do not think that I have any resistors on the encoder itself, I believe all of that is done back at the servo pack (Driver) for the tranfers and lifts and inside the robot controllers for the robots. I will have to destruct a motor/encoder this week to build my cable adapters.
From my cable drawings I have 4 different configurations,depending on motor size and application.
1) A,B,C and S Channel
2) A,B,C and Z Channel
3) RX+,RX-,A,B,Z Channel
I believe the 3rd channel is the index channel. The Yaskawa manual references C,Z,and S as the indexer depending on what page you are on.So in Ex:1,2 I think Z,S is not used.
4) Data+ and Data- only
Examples 3,4 are the ones I am leary about. Can I hook the scope to data+ and data- as 1 channel ? If so I do not know what I should expect.
My Fluke ScopeMeter only has 2 channels, so do you think if I check A and B only and they are 90 deg out the encoder is ok ? If the Indexer did not output my guess is the motor would continually run until a mechanical crash would occur.
The RIM Tach M100 will check A,B,Z together in a neat little 9v supply package about the size of an old calculator, but the salesman would not change his wiring on his cable adapter to match 3,4 because the M100 cost is $1500.00, but he said he THOUGHT it would work.It uses a 68HC11 microcomputer and I think a TI-SN75175 could work with some display.Let me know what you think about EX:3,4

 

[Bucknut]

AeroPD850,
After prying the encoder cover plate of 1 Yaskawa Motor, I decovered the pcb that you described and it does contain all of the resistive circuits and LED.I was wrong when I thaught it was in the servo pack.

 

[AeroPD805]

Buck,
Yes, you're right channels S & Z are the index channels. I'm not sure about
you're application, but if it has a home position and has to travel a given
distance, then an index pulse is used. If this channel is out of adjustment
(high waterfall [distortion] or low amplitude), then you'll have the problems
you mentioned.
On the units with data+ and data-, most likely those are duplicate channels.
In other words the (+) is the positive TTL signal and the (-) is a inverse
duplicate. So while testing you don't really have the monitor the (-)
channel, it will mirror adjustment is made to the (+). If you want to
check the channels, hook-up to them seperately.
FWIW, depending on the chip used, some models require a (+) &( -)
input voltage, (although with your application +5 should do).
A word of caution - if you do R&R the encoder, be extremely careful
with the disc, it's either made of [thin] glass or steel! Very few if any
disc are interchangeable unless they're from the same p/n encdoer,
(to the letter!) and are hard to acquire.
Also, it's critical that the encoder must go back in the same location
relative the end cover it's mounted to. Mark around the encoder before
removal.
When testing check for the following [with a multi-channel scope]:
-Waterfall < 5% [signal distortion]
-Quadrature 90 degrees +/- 20 [A phase to B phase]
-Symmetry 50% +/- 10% [looking at 1 full cycle]
The A & B channels define direction and the S [or Z], the index define
distance and home positioning. More channels beyond A&B are added for
more precision.
re: RIM; wow for that much $ he should do the testing for you.
If you have a Servo Box, they typically have test points to pick-up
the encoder. In fact I'd be surprised if it didn't - encoder adjustment
is a maintenance function on robotic systems.
Aero

 

[AeroPD805]

p.s. Let's see my consulting fee is....... [just kidding]

 

[nbucska]

Can you give me more data about the motor and its application?
<nbucska@pcperipherals.com>

 

[precisionzone]

I was really surprised when you said that Yaskawa doesn't repair their motors here in US, or won't be able to help you. The problem is to find right contact person. Yaskwa has full servo motor repair shop in Illinois and they do a good job, expensive but relaible.

 

[Bucknut]

Thank you to all that have supported this thread! I have learned alot from this.I have a good basis for my starting point.
We have had these systems for a year now and they have ran almost perfect. We are just trying to be prepared for what the 2nd year brings us. Usually we start seeing a reprocussion from wear and tear then.
RE:Repair from Yaskawa, Illinois.
They do repair everything else for use, but we have sent them 5 Servo Packs that drive these motors, and they all came back N/G but all with different problems then when we sent them out !