Encoder losing position

jonrya91 · May 23, 2017

Here is my setup:

AB compactlogix PLC
1769HSC card
Automation direct 1024 incremental encoder
AB panelview 1500
104:1 gearbox (I've verified this is correct)

This is a simple mechanical arrangement of a motor and gearbox turning a caster wheel.
106,496 pulses per rev. represents one turn of the caster wheel. (104*1024)

Here is my problem:
If I mark the caster wheel with a reference to locate one turn, my display shows a complete turn that is roughly 10-15 degrees behind where it should be. Each turn gets more and more out of time.

I've ruled out the encoder being bad (replaced with new), the encoder slipping and the caster slipping.

I've had to increase my full rotation encoder count to roughly 118,600 in order for the machine to limp by.

Does anyone have any ideas?

PLC Pie Guy · May 23, 2017

There must be an offset somewhere that is adding to each turn's position. Or is it possible that the coupling to the encoder is slipping a bit, not likely but.....

Is the error exactly the same amount of pulses each turn???

bernie_carlton · May 23, 2017

Are you using quadrature counting? Possible noise?

PLC Pie Guy · May 23, 2017

Bernie,,, I was thinking noise as well, that's why I asked if the error was always the same amount of pulses. I had the same set up mounted near a motor before and the noise drove it crazy... Relocated away from the motor and all was well.

I am assuming you used shielded cable for your encoder wiring right??

jonrya91 · May 23, 2017

There's no offset in the program that I can see.

The encoder isn't slipping - I've checked the set screws several times.

I'm not familiar with quadrature counting - however; I will check the noise issue shortly.

Thanks for the help so far!!!

PLC Pie Guy · May 23, 2017

Perhaps some screen shots of the HSC card configuration for that input could be helpful.

ALSO,,, Is the error exactly the same with every turn? or does it count 100 extra pulses on one rev and 1000 on the next?

Is it negative on the actual position or positive on the actual position?

jonrya91 · May 23, 2017

Here's some screen shot attachments of the configurations:

I don't think this is noise related - the encoder cables are shielded and grounded. They are near motors running on AC drives so I still might isolate the cable.

It's not counting extra pulses - there's a reset to show % rotation. (see attachment) The caster wheel is going slower than the PLC thinks. I can't put a number on the counts, but it's roughly 10-15 degrees lost per turn on the caster wheel.

jonrya91 · May 23, 2017

Here's the encoder count reset.

PLC Pie Guy · May 23, 2017

Is it at all possible that your not running EXACLTY 104 to 1? when I am setting up encoders on machines where there were no encoders before,, I like to try to shaft couple it to the machine or motor. (flexible coupling of course) Making the ratio null and programming simple. I'm NOT super experienced in the field of encoders but 104 just seems like a lot. especially if its not perfectly 104.

Also, if your running pulleys or sprockets or any type of chain or belting you could have a mechanical stretch issue you are dealing with.

I have encoders on chain drive production lines give different readings at two ends of the line simply due to mechanical wear. I have to compensate for the error caused by the chain wear to keep accuracy.

Possible??? Is this a new install or you are fixing?

jonrya91 · May 23, 2017

Old install I am fixing. As far as I know, it's done this forever and we've just dealt with it.

The gearbox ratio is definitely 104:1. I ran the machine very slow and actually counted the encoder to one rotation of the caster. It's close enough that I shouldn't see 10-15 degrees error (This is a large 48" wheel too so it's easy to see)

The motor is direct into the gear box no pulleys etc.

Also turning the encoder by hand I see roughly 1000 counts - it's supposed to be 1024/rev.

PLC Pie Guy · May 23, 2017

jonrya91 said:
Here is my setup:

This is a simple mechanical arrangement of a motor and gearbox turning a caster wheel.
106,496 pulses per rev. represents one turn of the caster wheel. (104*1024)

?

Perhaps Im missing something.... really possible... but wouldn't you use the value of 104*1024 being 106496 as the SCP instruction input maximum and the HSC reset???? you are using a value of 118550.... Where did you get that?

jonrya91 · May 23, 2017

PLC Pie Guy said:
Perhaps Im missing something.... really possible... but wouldn't you use the value of 104*1024 being 106496 as the SCP instruction input maximum and the HSC reset???? you are using a value of 118550.... Where did you get that?

You are correct. I increased the number in order to get the machine to run "properly".

If I run the machine at 106496, it's waaaay off and can't be used.

keithkyll · May 23, 2017

On image 5274, I see a problem with the pullup resistors. It's drawn with both resistors on the A channel, nothing on the B channel. Should be one on each channel.

mendonsy · May 23, 2017

You said that you counted 1000 counts/rev when turning it by hand, but the results you are getting say it is about 1140 counts/rev??? Something isn't adding up here.

jonrya91 · May 23, 2017

keithkyll said:
On image 5274, I see a problem with the pullup resistors. It's drawn with both resistors on the A channel, nothing on the B channel. Should be one on each channel.

I don't think we have them installed at all - or they were left out. I see none in our panel or junction box.

Encoder losing position

jonrya91

Member

PLC Pie Guy

Member

bernie_carlton

Lifetime Supporting Member + Moderator

PLC Pie Guy

Member

jonrya91

Member

PLC Pie Guy

Member

jonrya91

Member

jonrya91

Member

PLC Pie Guy

Member

jonrya91

Member

PLC Pie Guy

Member

jonrya91

Member

keithkyll

Member

mendonsy

Member

jonrya91

Member

Similar Topics