EZ-Trak Y DAC Overflow

[TRAK-Man]

I have an EZ-Trak DX that started giving me a Y DAC overflow while doing the initial homing on the Y axis. This just started after sitting for a month without power up.

Looking around on the forum, it appears that this could be an encoder problem. I've checked the Y-home switch, and it seems to works fine based on resistance measurement. Motor runs Y axis in both directions OK. DRO readout looks fine when I manually move the axis, which I guess means the encoder is OK.

I've read here that you should ohm out the cables to the encoder. There apprears to be two pairs of four wires. Should there be continuity between each wire of the four, and to any of the remaining three wires? I'm not sure how encoders work.

Also, apparently a MARKER signal comes from the encoder. Is there any way I can check for this? I have a scope, but don't know what to look for.

Thanks for any help you can provide.

[LeftCoastCNC]

Our EZ-Trak at work has been doing that about every 5th time we start it up. I am pretty sure it is the limit switch on ours.

My understanding of how it homes is; the start up routine tells the machine to go -Y until it hits the limit switch (not even looking at the encoder). Once the limit switch is tripped it reverses direction and the machine is looking for an index on the encoder to call home.

When trying to home, does your machine go past the switch and make a clunk or does it hit the switch and reverse?

When you were testing the switch, were you pressing it by hand or with the saddle? It might not get depressed as hard with the saddle...?

[machintek]

A DAC overflow means the control is expecting encoder pulses but does not see them.
If the motor actually moves then in theory the motor and drive are good.
The way the machine homes is that it moves toward the switch, makes the switch and now looks for the marker pulse. Sees this and now knows when to start counting.
Questions, when the drives are ON, is the Y axis drifting? (moving slowly).
When it gives the DAC overflow, has it hit the end of Y axis travel? If the axis is moving and giving a DAC overflow, it may not be counting.
Can you attach a full size keyboard? Go to DOS, root directory, PFM.exe ENTER, cursor down to tuner, ENTER. Now you can see if the control sees the home switch and if the axis are counting. Compare the X to the Y.

Depending on the encoder you have, it can be removed(the can type is easy, the built into the motor is not suitable), 5VDC power applied and run in a drill. There are 3 channels putting out a 5 VDC wave. This can be seen on a scope. The marker pulse typically occurs once per revolution. If the marker pulse fails to be seen, the control moves the axis until it hits the end of travel and you get a DAC overflow.

George

[JoBwan]

Great info. Mr. Moderator

[TRAK-Man]

Thank you for the great information.

I was able to fire up PFM and I can see the counts changing on the X and Y axis, and the home signal change on each. Neither axis drifts while under power.

Is there any way to see if a marker pulse has occurred?

So, if I understand properly, for each axis it travels until it hits the home signal, then continues the same direction until it gets a marker pulse, then stops and considers itself homed. Is this correct?

On mine, it moves out on the y, runs out of travel, then seems to jump back about an eighth of an inch, then stops with the error message. Should it never move back as part of a normal zeroing? I can't remember how it originally worked. If it is supposed to move back after the home, maybe the problem is that my motor isn't driving smoothly, and that's the reason for the jump, and DAC overflow error.

To check the drive, I was thinking it would be nice to jog the y-axis in both directions to see if it is driving smoothly. Is there a way to do that with PFM? Since it won't zero, the normal EZ-TRAK software won't let me get into jog mode, or enter a move command.

Any idea what wire on the encoder (YA, YB, YI, + or -) that the marker pulse shows up on? Is each a pair, for example, so you should measure between YA+ and YA- to see one of the signals? Are these opto-isolator outputs?

Your help and knowledge is appreciated.

[machintek]

It may not be seeing the marker pulse.
It is hitting the end of travel torquing up, faults out because control stops seeing encoder pulses (here is the DAC overflow), drives shuts down and jumps back as motor torque is relieved.
here is the best way to see if the home is not adjusted correctly or if the marker pulse is missing.
Move the axis by hand to the center of travel.
start the homing process and carefully press and hold the home switch.
If it keeps on coming for your finger by an inch or more, you do not have a marker pulse (you said the home switch was working correctly).
To make sure it is the encoder and not the cable, etc, swap X and Y motors. If it follows the motor, you have your answer.

George

[TRAK-Man]

Thanks for the good ideas George.

I tried your idea of closing the home switch as it travels, and as you suspected, it kept right on going. This makes me think that it must not be getting the marker pulse. I presume that this pulse is the one marked I+ and I- for index. I tried ohming that out, but the signal is very high impedance, so I didn't get anything useful from doing so. I ordered a cable extender which will come in a few days that will allow me to watch the signals with a scope to see what's there. It would be interesting to know what the circuit is in the encoder.

Your idea of switching the motors is interesting. Are they supposed to be the same and interchangable on X and Y? My X motor is a CMC MTE4225157G2F, and looks different than the Y motor, which doesn't have a tag or label on it. I have attached a picture of the Y motor. I popped the cap off the end, is that the encoder that I am seeing?

To change them do you just remove the cables and four mounting bolts and they slide out on from splines, or is more disassembly necessary?

Thanks for your fine thoughts.

[machintek]

There is a difference between a encoder and a switch. The ohm meter will not do you any good. This is a device that is triggered optically to create a wave form.
In theory, the motors should interchange. I would return them to the original position in case the drive board tuning is more specific than usual.
Yes, that little can is the encoder. If you carefully pop off the can, you will see the glass disk with the lines etched on it. It has a set screw holding the disk onto the motor shaft. This is a very fragile device.
The motors are held by those 4 screws. The end of the motor has a pulley that drives a belt. No splines.

George

[TRAK-Man]

Thank you again George.

I ohmed out the encoder wires from the motor to the PCB in back, and all was fine. So I pulled out the Y-motor, and hooked it up on my workbench so that I put 5V and ground into the encoder, then looked at each +output with the scope as I turned it with my fingers. The "A" and "B" gave nice 5V squarewaves as it spun.

The "I" output, which I presume is Index (or marker) output, was high all the time, except if I spun it, it would output a train of about 15 pulses on each revolution. I would have expected it to be low, and then give one positive pulse each revolution - but that's just a guess. I suspect a cheapo Radio Shack logic probe would also help provide diagnostic help if one doesn't have a scope.

Anyway, in the end I put the X motor in the Y position, and it worked fine, confirming that the EZ-TRAK isn't getting the marker properly from the Y motor.

So, I guess the motor/encoder is off for repair, and hopefully chips will be flying again soon. I thank you very much for your help. It's been very valuable to me, and I've learned a lot.

Brian

[BioMicro]

Not sure if you want to try this, but you can replace the encoder. I did one, and it isn't too bad - you have to be careful as it's very fragile like machintek says. But if you get the set screws loosened, it comes off easily. I replaced mine with and Encoder Products part, about $240 IIRC and this is what they use at EMI, since the Heidenhain ones are more expensive. But maybe <$400 to replace the encoder at EMI?

http://www.encoderproducts.com/ is the site, they have a dealer network. I believe I got ERO 1420-250 (old number was 1421, changed to 1420). 250 is the line count per revolution. I noted that "You must align “.275 inches past where home switch makes.” (not sure what that means, I don't recall having to align it.)

I was advised that sometimes the motors need internally cleaned, which (cleaning) can open or clear a short which causes missed counts. (possibly the tachometer? - both the motor and the tach have 4 brushes each.) They charge $460 for a motor refurb. at EMI. They don’t replace the brushes, said they usually don’t need it.

[TRAK-Man]

Thanks for the tips about how to do this, but I think I will let the pro's handle it this time, I'm a little nervous about screwing things up, then I'd be out the time and money of the encoder. Very tempting though!