Encoder Hookup

[murphy625]

Ok..
Here is a cut-out from my servo manufacturers page on hooking up the "Encoder Simulation" port to the cnc controller.

Keep in mind that the motor has a "Resolver" that hooks up to the servo drive. The servo drive has an "Encoder Simulation" port that gets wired to my ICM2900 from Galil.
But the manufacturer does not specifiy what each pin is. They only reference them as shown in the drawing.
The ICM2900 has +max, -max, +mbx, -mbx, +inx, -inx, +5v, gnd.

I would assume that max is A chanel, mbx is B channel. but what about the 3rd set of twisted pairs that are marked as C and C- on the servo drive module? Do those go the [inx] ?? The servo resolver has an Excite channel also.. Is that the C and C- ??
The system seems to work with just the A an B channels hooked up but I wonder if I did the right thing or not by leaving the C channel unhooked.

Any suggestions?

[HillBilly]

C+ and C- should be the index pulse from the drive to the Galil. You will not notice it is disconnected until you issue a FINDINDEX command. Like Al stated this is a precision marker for homing. It occures either once per encoder cycle or once per revolution.

The mechanical home switch will vary but going to the next encoder marker pulse after the switch is triggered is very repeatable.

Darek

[Al_The_Man]

Murphy, I previously gave you a bum steer, I thought that the C wires were the power supply, I am used to seeing Z designated as the marker pulse, Having said that your diagram does not show encoder signal common? This is what I meant when I mentioned that some simulator ports have a common and a +5v, in this application the +5v would not be connected.
Your resolver is completely separate from the encoder signals, as the resolver only has a sine/cosine signal and is very low resolution, so apart from being used for motor commutation, it is electronically multiplied and conditioned for Differential line driver output for the encoder simulation plus marker.
Al

[murphy625]

Is Hillbilly right?
Should I connect the c and c- from the encoder simulation to the index + and index - terminals on the ICM2900??
Please be patient with me. I will learn encoders but right now I have my hands full working out wiring mistakes and other bugs.
Also, what I found very interesting is that when I look at the block processing diagram in the servo manufactures drawing, it looks like it shows all the negative (Low) terminals on the resolver are hooked to gnd anyhow. They are all marked "0 Volts" on the drawing. SinHigh, SinLow, CosHigh, CosLow, ExciteHigh, ExciteLow. <<<as marked on the resolver terminals...
I guess that would make sense if the resolver is really nothing more than a spinning transformer. (but what do I know).

From reading both of your replies, am I to understand that I don't even need to hook up the index wires? (remember my tolerance is + or - 1/4 inch or more!)

As I understand Hillbilly's post, the marker pulse or index pulse is only there because a mechanical limitswith will trip at different points varying a few thousands here and there? (I understand those fine accuracies are required for most cnc applications)
So a normal homing routine would direct the machine to the home switch. When the machine hit the home switch it would slow down greatly and then look for the next marker or index within the following motor shaft rotation?
So in most applications, the home switch is nothing more than an approximate location and the shaft orientation will determine the actual physical position of home?
I must have this correct because it sounds logical. Yes?

Thanks for the help..
Murph

[Al_The_Man]

Murph, Yes Darek was right about the C channel signals, and if you have them why not hook them in, Galil has instructions for finding a switch trip as well as finding the marker, the normal way is to use both as you point out. Did you check the simulation port to see if it mentions a signal common?
Al

[murphy625]

The simulation port does not mention any signal common.
The manufacturer included very little information about it in the manual. (and I have both versions of the manual) LOL..

They only list the pins as described.
In the DigitAx brochure (and I just found a single reference to it ) they list the encoder simulation as dual quadrature channel with marker pulse.
(Yes, after a 4months of referencing this manual, I just found this) Its listed under the "features" section..

So now, I look like a moron, and I would assume the C and C- are Index and Index -.

It would have took me much longer had someone here not refereed to that index as a "marker".

Now I am learning.

Also, if I was to check out the encoder pulses with a scope, would I be correct in assuming the timing I needed would be RPM X Encoder Counts per Revolution?

Thanks for your help,

[Al_The_Man]

Originally Posted by murphy625 Also, if I was to check out the encoder pulses with a scope, would I be correct in assuming the timing I needed would be RPM X Encoder Counts per Revolution?

That should display one encoder pulse, also remember that the motion card can increase the resolution by a factor of 2 or 4, it does this by counting both the two leading edges of the A & B pulse, or the leading/Trailing edges to effectively give for a 2000 count encoder, either a 4000 or 8000 count/rev resolution.
Al

[HillBilly]

Common should only matter when using single ended encoder signals. When using the differential inputs opposite signals are detected and common signals (noise) are rejected.

From what you describe about your resolvers they are being used single ended.

1/4" Tolerance that is loose. The exact home is a nice feature to resume a cut after a mishap, say a power failure ect...

Darek

[murphy625]

Thank you AL,

Thank you Hillbilly,

Murphy