Retrofiting a Hurco BMC 30

Hi

I could buy a malfunction Hurco BMC 30 with manufacturing year around 1990.
The controller wasn’t starting up correctly (it was not possible anymore to home the axis and some other wired issues).
The previous owner blamed the holding battery, so he changed it but he wasn’t able to bring it back to live even with a Hurco Technician.
But it was possible to move all the axis and the encoder were all working to.
I wasn’t able to start the Spindle (due to the homing issue) but according to the previous owner it should work to.

So I was thinking remove the controller and put Mesa cards in instead.
The Servo Controller is a Funac “A06B-6050-H401” (3 axis AC servo Controller) with analogue velocity control and encoder out.
I bought a 7i97 and a 7i64 to replace the old controller.
So far so good.
Now I started with the replacement. I let the Relays control board in place (PCB which connects everything together). And just removed the old controller. So I had a cable for each axis witch was before connected to a “double axis Control board”.
I connected the necessary stuff to the mesa Card (Analog out, Encoder, limit switch, overload digital in).
I now can move the axis but I don’t get the position back (it just stays at 0.000).
I tested my LinuxCNC configuration with an encoder I had laying around and it worked.
Next thing I checked was the signal with an Oscilloscope (see screenshot).
I don’t know what’s wrong but the signal looks ugly -.-

Does anybody have an Idea what’s wrong / has anybody worked with a similar Funac Servocontroller?

Looks like regular quadrature encoders, right?
Where did you probe that encoder signal? Before or after the relay board?
Have you probed the other encoder signals?

[edit]
I presume that's the A and the B channel showing on your scope?

[edit2]
Do you know what the signals on the encoder pins G,D and E are? Maybe those are inverted encoder signals you could use?

@cnc4ever

I am assuming you mean Fanuc for your control.

If they are the red cap servo's, then they have a special signal that is not like a standard type encoder with nice square wave. I am positive Jon Elson (pico-systems) built boards for these, but not 100% sure... all memory. I just remember trying to stay away from red cap because of this problem.

EDIT: I went and checked Jon's website... here is the link to the board. Hopefully something to help. I don't know if MESA makes one but you can check with Pete (PCW).
pico-systems.com/osc2.5/catalog/product_info.php?products_id=33

Robert

It looks like standard quadrature in the drawing.
As Aciera said I would think you would want to use differential mode
with all 6 encoder signals (especially since all the scope shows is noise...)

I am glad Pete checks behind me, definitely don't want to misinform someone....

Oh I see, never looked at this pins G, D and E.
I thaught that the "A ref" signals where the inverted one's.
I will try to get these signals and use "truly" differential mode.
But why did they had all these "ref" Signals?
It looks like the controll board (removed one) was differential but the connected it like "TTL" right? Because all the "ref" signals go to "zero Volt reference".

What I don't realy get is why the oscilloscope signals are looking so bad.
I put the oscilloscope like if the signals were differential ch1: "A" ch2: "A ref" and the ground to a my ground.
But it looks like this "zero Volt reference" has nothing to do with my ground.
I also have a small Battery powerd oscilloscope, witch I connected ch: "A" ground: "A ref".
And the signal looks good (see attachment).

I today checked the servo's they have red cups Type is "A06B-0512-B002" it is written "Pulse Coder" on the type Plate (see attachment). As far as I know the red has more to do with the type (yellow is DC motor).
But for the encoder I don't know if they are normal one's or not.
But I think anyway as long as I use the original servo Controller this shouldn't matter right?
Because the servo Controller will "translate this Signal".

And yes it should be "Fanuc" not "Funac". Sorry

I rewired the Encoder to the "real" differential signals and now it works perfectly.
I guess the signal was just to weak.

All the axis are working flewless
I connected the Spindle now to and I can make it turn.
I am realy impressed how easy it was with LinuxCNC and the Mesa Card.
Realy great open source work was done here.

Next work is to fix a problem with the spindle and get the toolchanger to work.
Maybe somebody here can help me.

The problem with the spindle is. It always turns the same direction.
I have a relay for CW or CCW on the original controll board.
I tried to swap them but then it yust turns always the other direction.
I am not entirly shure what the problem is. But I would like to test what happen if I would yust output positiv signals.
So 0-10v when CW is active and 0-10v when CCW is active.
How can I do this on my 7i97? (config in the appendix, mostly from tommy's Huroc Project thanks a lot for sharing).

To I would like to get the Loadmeater (analog Signal) from the spindle on my linuxCNC dash.
Is there a Mesa option to read Analog witch is isolated from the controll ground?
My 7i64 would have a Analog In but then I would need to connect the ground from the spindle driver together with the ground from the Mesa cards and I think this is not a good idea. Or do I get something wrong?

So my last question is about the spindle Orientation.
What I found out so fare is there is a Magnetic sensor witch is responsible for this.
It is connected to the Spindle drive and from there I get a Orientation complete.
To I have a realy where I can enable this.
But so fare it does about nothing. How is it supouse to work? Do I have to enable the realy while the axis is enable and spinning?

Sorry for long post.

You can get just 0 --> +10V by using feeding the spindle speed through the
abs hal component

One possibility for reading the spindle load would be a 7I87 its overkill
(8 analog input channels) but it is isolated