Need help with manual pulse generator

[Jackal66]

I have a plug-in pendant:

CNC4PC

This is the board that connects it to the2nd parallel port:

CNC4PC

The problem is, it will only jog in one direction.

When it is turned clockwise, the machine will go in the -neg direction.

When turned CCW, nothing??

I have traced down the wire and they appear to be on the correct pins.

How do you test an MPG?

If it pulses clockwise, shouldn't CCW be ok also?

Does it require A & B both to pulse?


Thanks for the help,

JAckal

[keebler303]

A and B are required to determine direction of rotation. If A is logic high when B goes high, it is going one direction. When A is logic low when B goes high, it is going the other direction. You need both connected. To troubleshoot, you could swap the A and B wires where they connect to the board. If its still clockwise only, your software or board probably has an issue. If it will only go CCW, then the encoder probably has a problem.

Matt

[Al_The_Man]

The symptom is what you would expect with one channel missing, most MPG's use a 100p/rev quadrature encoder as a source, and has been said, you need both A & B.
to count up and down.
Al.

[Jackal66]

Thanks Keebler & Al.


I reversed the wires A&B on the back of the Hedds Encoder (handwheel).

It still only jogs in the -neg direction, but with a CCW rotation of handwheel.

JAckal

[Al_The_Man]

You can use a meter from common to each A & B output and if you turn VERY slowly while monitoring the A pulse, as soon as it goes high, check the B it should always be low in one direction and always high in the opposite dir, whenever the A transitions to +5v.
Easier with a 'scope though.
Al.

[Jackal66]

Al,

I tried checking itwith a meter. When the (A) side is at 4.87 volts+, the (B) is at 0.08 volts or 2.87v max.

It was real tricky to hold the (A) at the certain click where it would stay at high side.

Something that I did notice in the Mach diagnostics page:



The (2) LEDs that are circled, were flickering.

It was like a Morse-code or something. They would go back and forth making long or short flashes while the hand-wheel was being turned.

They did this in both dir. of the hand-wheel.


Thanks,

JAckal

[Al_The_Man]

They should both transition from 0v to +5v, or close, these are not open collector outputs by any chance?
Al.

[Jackal66]

Al,

I'm not sure. I don't really know enough to say either way.

Here is the technical data for the encoder in the pendant:




Hope that helps.

If not, maybe Arturo at CNC4PC can give me more info.

Thanks,

Jack

[keebler303]

This is what the encoder should output: http://upload.wikimedia.org/wikipedi...re_Diagram.svg The flickering LED's should also blink in this pattern. If they are not, the voltage level you are seeing may not be sufficient. It appears that your encoder has a line driver output rather than an open collector output. That is good for you. You could also try to disconnect the A and B wires and measure the voltages without anything connected to them. You could further emulate the encoder action by jumpering the B signal to either the common or the 5v encoder supply and pulsing the A input by rapidly connecting and disconnecting it from the 5v supply. It should make the axis move one direction when B is held low and the other when B is held high.

Matt

[Jackal66]

Keebler,

I am getting pretty much the same pattern in clockwise to CCW direction.

Cw is:

left
left
left then >right on same click
right
right
both
right right
none

Ccw is the same pattern, with a couple of differences.

I am going to reload my brains in Mach3, for about the 5th or 6th time.

All of the sites keep saying to do that as many times as it takes.

I'll keep everyone posted,

Thanks for everyone's help,

JAckal

[keebler303]

I realize now my link was not good. This is what I meant: File:Quadrature Diagram.svg - Wikipedia, the free encyclopedia

So your pattern should be:

right
both
left
none
right
both
left
none

and the other way:

left
both
right
none
left
both
right
none

[Jackal66]

After jumping the wires manually ( momentary short pulses) it still only works in one direction. (With "B" at both neg- and at 5+pos)

I'll check the LED pattern again, compared to the sample you gave me.

Does the 5 volts have to be exactly 5?

The voltage reading was 4.82 to 4.89.

Does the parallel port need to be EPP or SPP?

Mach was communicating perfectly through this parallel port.

Could it be the c-22 interface board is bad?

Next, I will connect the MPG directly to the parallel-port and get 5 volts to it.

Any other ideas are welcome.

Thanks a lot,

Jack