[DickinsonM]

Hi there,

Im brand new to the community, and am in the process of hooking up my old Strippit FC-1000 to a new controller from CamSoft.

Im having some positioning problems that they cant seem to help me with, and Im not sure whether its software related, noise related, or servo/encoder related.

Basically, this is my problem....

When I move the X axis in a positive direction.
Every 1" of travel equals 2000 encoder counts.
And 1" does infact equal 1".

Yet when the same axis moves back on itself, in a negative direction.
1" then equals 1.040"
So it adds about 40 thou per inch, everytime it moves negatively.

I cannot for the life of me figure out why this would be.
Since the reading on my screen and encoder, is reading exactly what it should.



So for instance....an example move would be.....

Positive move to 10" (Reading 10.000 - True position 10.000)
Negative move back to 5" (Reading 5.000 - True position 4.800)

Anybody have ANY idea as to what could be causing this.

Camsoft seems to think it is noise in the amps/encoders.
But I think its far too consistent to be noise.

Any help wouldbe greatly appreciated.

Thanks

[mikpat]

hi there been doing cnc maching for 22yrs maybe I can help
This sounds like parameter problem ...1st what type of control fanuc
,mazak

[DickinsonM]

Controller is Camsoft.
Fanuc I think.
Galil motion card.

Machine is Strippit FC-1000

thanks

[mikpat]

check this web site it may help..strippittech.com
It say using a F2 limits axis speed to 2000 IPM (this is your speed control for axis movement.., to slow it down some

[gridley51]

There are a whole load of questions thrown up by this,so I`ll start at the beginning.
I assume that you have checked the machine mechanically and it`s backlash free?
You say servo/encoder fault so assume it`s servos.
Have you measured the actual travel with a length standard and dial gauge?
I`ve got more but prefer to do a bit at a time.

Mark.

[DickinsonM]

Yes, the machine is backlash free.
There might be like .005" backlash. But nothing like the 0.040 Im experiencing for every inch.

Yes Ive measured the travel.
And when it moves away from me, it moves perfectly 1" every time.
Then when it comes back to me. It moves 1.040".

The funny thing is.
Today, I tried another encoder, just to see.
And the problem actually reversed itself.

Now when it went away from me, it moves about 1.015"
and toward me was 1" perfect.

So the problem reversed itself, and got a tiny bit better.

I am completely stumped on this.
Why would the encoder reverse this problem, when the two encoders are IDENTICAL!!!

MAKES NO SENSE!

[Dorsal]

It sounds like the glass scale is loose, or the encoder is loose and following the scale for the first 0.015"
Tell us, does the error remain in the same direction, no matter which direction you traveled first. In other words, move 1.000 in +X, then move back in -X. Then move another 1.000 in -X, then move back 1.000 in +X. Same result, same direction?

[DickinsonM]

The encoder is a rotary encoder though.
So no glass scale.

Could it be how it is mounted??
Is mounting direction extremely critical?
Because right now its slightly angled so the wire can connect to it.
But perhaps it has to sit exactly straight?

[jdkuebler75]

it sounds to me like your encoder marker pulse "long pulse" is either not being seen or there is noise on that line do you have an ocillisscope? i know this is an old thread but if you are having problems still [email protected]

[TechGuy5002]

The encoder is a rotary encoder though.
So no glass scale.

Could it be how it is mounted??
Is mounting direction extremely critical?
Because right now its slightly angled so the wire can connect to it.
But perhaps it has to sit exactly straight?

Its possible the angle could cause a problem. Can the encoder shift? ie does it wiggle if you try to turn it with your fingers? It could be that its shifting when you can directions causing the error. It wouldn't take much movement to make that error.

Another thought is that its mis-wired or miss-labeled, or you have a faulty encoder. Can you swap an encoder from another axis. If the error goes away, then you likely have a bad encoder.