[chozo]

Which hobby cncboard? The new chopper or the old one with power resistors? Did you measure the 5v with the driver board cabled to the computer? If its the chopped hobbycnc board, I suspect it's using a Sanyo driver. It needs 4V for a logic high. The only way you can verify that your card is outputting 4V into the driver board is to measure it with the board hooked up. It really sounds like the motor on bad channel is seeing step/clk signals that are too fast. An incompatable logic level or noise could be the culpret. Which axis?

Phil

It's the new 4-axis chopper board.

The Y-axis is having the problem, regardless of which motor is hooked up.

The only way you can verify that your card is outputting 4V into the driver board is to measure it with the board hooked up.

Is there any easy way to do this? I guess I could unscrew the board and lift it up, then check the connection from underneath.

I'm assuming you measured this at the PC/PCI card parallel port with nothing connected to it? If so what pins exactly are reading 5vdc? Also where did you buy the card? Brand?

I measured at the end of the cable. Pins 3-7, 10-13, 15, and 17 read 5vdc.

I bought the card from TigerDirect. It's an I/O Flex.

[CNCPlastic]

That Tiger card isn't going to work, all of your input pins and a bunch of the data pins are lit up with 5vdc. You need a dirt cheap do nothing special parallel card, as I mentioned before I use the cheap parallel cards from CompUSA, here's a link.

Aaron at www.industrialhobbies.com uses another brand on all of their cnc mills, maybe shoot him an email.

If you get a CompUSA card send me an email when you go to configure the Mach2 port, you have to dig through the Windows system info to find the port range. [email protected]

http://www.compusa.com/products/prod...726&pfp=BROWSE

[CNCPlastic]

PS: Hang in there, assuring that your parallel port is not doing anything whacky is a significant first step.

This whole mess of cnc electronics will kind of sort of work with a whacked out parallel port but no amount of tuning will cure it. When we were trying to get my mill tuned with the first twilight zone built in parallel port it sounded like the mill was going to blow parts across the shop. The servo's would start, stall, fault, reset and repeat, resulting in a herky jerky table movement you would swear was a mechanical bind. I thought the whole balls crew setup was bound up or something was way too tight. The big stepper couldn't lift the head without stalling and sounded like crap. But all that seemingly mechanical noise was just the servo's and stepper reacting to the whacked out parallel port. I got the new card configured and she sounded so sweet, and smooth, just how you would expect a cnc machine to sound.

[pminmo]

I like cncplastic suspect an incompatability with your parallel port. Although I suspect it's more an issue with the Hobbycnc board than the parallel port. Are there mfg or pn's on the heat sinked modules? Pin 5 of the db25 on the hobbycnc board, can you physically see if it routs directly to a pin on the module? If so which one?

Phil

[chozo]

That Tiger card isn't going to work, all of your input pins and a bunch of the data pins are lit up with 5vdc. You need a dirt cheap do nothing special parallel card, as I mentioned before I use the cheap parallel cards from CompUSA, here's a link.

Aaron at www.industrialhobbies.com uses another brand on all of their cnc mills, maybe shoot him an email.

If you get a CompUSA card send me an email when you go to configure the Mach2 port, you have to dig through the Windows system info to find the port range. [email protected]

http://www.compusa.com/products/prod...726&pfp=BROWSE

I'll check that card out. What is different about that one? I thought the one I bought was pretty "bare minimum" as it is.

As far as configuring it, I know how to do it(but thanks for the offer). I got this one working. Just has the same problem as when I was using the integrated port. X and Z axis spin perfect, but Y locks up half the time, and actually spins the wrong direction sometimes.

I saw that webpage of Aaron's when looking up parallel cards. hehe
That card he uses is for sale at CompUSA as well. I don't see any difference from mine, unless it's just a quality issue.
http://www.compusa.com/products/prod...03893&pfp=cat3

[chozo]

I like cncplastic suspect an incompatability with your parallel port. Although I suspect it's more an issue with the Hobbycnc board than the parallel port. Are there mfg or pn's on the heat sinked modules? Pin 5 of the db25 on the hobbycnc board, can you physically see if it routs directly to a pin on the module? If so which one?

Phil

I'll check it, and get back with the results later.

Thanks.

[CNCPlastic]

I would expect that addon PCI parallel port cards probably vary as much as the built in ports, you just happen to pick one that isn't going to work is all. Who knows probably the next batch of CompUSA cards will be different and then we'll have to hunt down another source.

I do agree with the other guy, it sounds like there is a reasonable chance that there is a problem with your breakout board and/or configuration, but I think you really have to eliminate the parallel port issues as step one.

[chozo]

The driver chips look like this:

O------------O
|_S__K JAPAN_|
|__SLA7062M__|
|____4507____|
--------------
||||||||||||||

(nice ascii drawing? lol)

Pin 5 does not route directly to any chip that I can see.

One thing to note is that I do not have the driver chip for the A axis soldered in. I didn't think that would matter, since each axis seems independent on the board. Seems like something that "couldn't be the problem" like this is the no-brainer solution sometimes.

I just soldered it in, not that it will change anything at all.

However, even so, I could try using the A axis instead of the Y, and see if the motor spins fine.

BBL

Thanks to everyone again.

[pminmo]

Allegro part, you probably need 3.75v for a logic hi from your parallel port. looking from the parrallel port there are 2 three pin devices, looks like U5 and U6. P/N's? LM7805 or LM317, or???

Phil

[chozo]

U5 is LM317
U6 is LM317HV

[pminmo]

I would suggest you contact hobbycnc. I suspect the LM317 that powers the axis your having trouble with maybe running a little high. If there is no other logic on the board, your allegro part can run it's logic down to 3.0V. For more compatable input thresholds, the logic power should be closer to 4V, than 5V. If you measure from pin 6 to pin 11 (gnd) on the 7062 that your having trouble with is it 5V? higher or lower?
There is two resistors associated with the 317 to set the voltage, one of them could be changed slightly to be more compatable.

Phil

[chozo]

I would suggest you contact hobbycnc. I suspect the LM317 that powers the axis your having trouble with maybe running a little high. If there is no other logic on the board, your allegro part can run it's logic down to 3.0V. For more compatable input thresholds, the logic power should be closer to 4V, than 5V. If you measure from pin 6 to pin 11 (gnd) on the 7062 that your having trouble with is it 5V? higher or lower?
There is two resistors associated with the 317 to set the voltage, one of them could be changed slightly to be more compatable.

Phil

There is only one LM317 on the board, so what do you mean by "the LM317 that powers the axis your having trouble with" ?

Pin 6 reads 5vdc on all chips.

Thanks a lot for your time and help.

[pminmo]

The LM317 and LM317HV function the same, the LM317HV will accept up to 60V input though, where the plain LM317 I believe is 35V. Without a schematic I'm shooting in the dark. http://cache.national.com/ds/LM/LM117.pdf is the datasheet. The center pin of the 3 pins is the output of the lm317(or lm317hv) regulator. Without power if you have a method to measure resistance you could see which one connects to you noisey channel. Them there are 2 resistors colse to it, does you kit have a datasheet on what the values of the resistor are?

Phil

[chozo]

The LM317 and LM317HV function the same, the LM317HV will accept up to 60V input though, where the plain LM317 I believe is 35V. Without a schematic I'm shooting in the dark. http://cache.national.com/ds/LM/LM117.pdf is the datasheet. The center pin of the 3 pins is the output of the lm317(or lm317hv) regulator. Without power if you have a method to measure resistance you could see which one connects to you noisey channel. Them there are 2 resistors colse to it, does you kit have a datasheet on what the values of the resistor are?

Phil

There are two resistors at the front and back of each regulator.

LM317 has
249 1/4 W 1% opposite the tab side of regulator
750 1/4 W 1%

LM317HV has
249 1/4 W 1% opposite the tab side of regulator
and depending on what fan voltage:
12v: 2.15K 1/4 W 1%
24v 4.53K 1/4 W 1%

As far as measuring resistance, I'll have to get back with you. I'm leaving.

Thanks again..alot.

I have a basic schematic of the board. I'll try to get it uploaded.

[pminmo]

It would be the LM317 then, not the HV. Reducing the value of the 750 ohm resistor will reduce the 5V output of the LM317. The minumum threshold voltage for the SLA7062 is .75 * Vdd (5V) in your case. That means the parallel port has to output a logic high of at least 3.75V when connected. The original PC parallel port would typically put out 3.1 or so volts in a high state. My pc puts out around 3.4V into a 10K load. This could possibly be the crux of the problem. Because as you near thresholds, it gets intermittent, works some, doesn't work all the time. But here is the other kicker, you can't just put it down to say 4volts because VDD of the 7062 should be no lower than .3V higher than the parallel port output high voltage. i.e. if you were truely getting 5V out, then 4.7V for Vdd would be the minimum.

Phil

[chozo]

It would be the LM317 then, not the HV. Reducing the value of the 750 ohm resistor will reduce the 5V output of the LM317. The minumum threshold voltage for the SLA7062 is .75 * Vdd (5V) in your case. That means the parallel port has to output a logic high of at least 3.75V when connected. The original PC parallel port would typically put out 3.1 or so volts in a high state. My pc puts out around 3.4V into a 10K load. This could possibly be the crux of the problem. Because as you near thresholds, it gets intermittent, works some, doesn't work all the time. But here is the other kicker, you can't just put it down to say 4volts because VDD of the 7062 should be no lower than .3V higher than the parallel port output high voltage. i.e. if you were truely getting 5V out, then 4.7V for Vdd would be the minimum.

Phil

I was understanding everything you said, but I don't quite get that last part. Sorry.
The voltage at the driver chip has to be at least .3V higher than the parallel port? 4.7V is the minimum for what?

[sbrpollock]

Before going too much further; I have a troublesooting suggestion:

If you have a break out board or some other means of hooking the signal wires from your parellel port to your driver board, try the following:

Assuming the Y axis has the problem and the X & Z axes are both O.K.

Swap the Step & Direction signal wires that are comming from your parallel port to your Y axis with one of the other axes.

In other words, the parallel port signals that are presently driving the X & Z axes are working fine, so use one of these to temporarily drive the Y axis.

If the Y axis starts working fine, and the problem moves to the axis you swapped with, then I would be suspicious of the parallel port.

If however the problem continues to be on the Y axis, this procedure pretty much rules out the parallel port as the problem.

Can you do this?

[chozo]

Before going too much further; I have a troublesooting suggestion:

If you have a break out board or some other means of hooking the signal wires from your parellel port to your driver board, try the following:

Assuming the Y axis has the problem and the X & Z axes are both O.K.

Swap the Step & Direction signal wires that are comming from your parallel port to your Y axis with one of the other axes.

In other words, the parallel port signals that are presently driving the X & Z axes are working fine, so use one of these to temporarily drive the Y axis.

If the Y axis starts working fine, and the problem moves to the axis you swapped with, then I would be suspicious of the parallel port.

If however the problem continues to be on the Y axis, this procedure pretty much rules out the parallel port as the problem.

Can you do this?

I don't have any means of doing it at the moment.

Maybe I could buy another gender changer so I have female connectors at the end of the cable, and at the controller board.
Then run wires in between (by pushing them in the pin holes), but switch the step/direction wires as you mentioned.

Wouldn't this accomplish what you are suggesting?

Thanks.

[sbrpollock]

I am not familliar with the HobbyCNC board or the schematic for that board, but I did have a devil of a time with this driver from Allegro the first time I used it. I was using it in a circuit that was not being driven by a parallel port but by a function generator. I experienced the type of intermittant problem you describe in your first post. The circuit was very tempermental. Here is what I found was my problem:

Both the driver supply "Vbb" (pin16) and the logic supply "Vdd" (pin 6) need to have decoupling capacitors running from these pins to ground. In my circuit these capcitors were extremly important and they had to be as close to the pins as possible in order for the circuit to operate correctly. In my case, my logic supply, step, and direction signals were all picking up a high frequency component from the PWM in the output. In the end, I ended up both rearanging the circuit and filtering EVERYTHING and all is well.

I'm not suggesting this is what is happening here since this is a design that these types of problems have already been wrung out of. However, I know the Allegro SLAxxxx devices are sensitive to this so I thought I would share it in case it might help. (Maybe the problem here is a bad capacitor?)

Since I don't have any access to the schematic for the HobbyCNC board, maybe someone here who does have access can tell if any of this is relevent.

[sbrpollock]

You posted while I was typing the above.

Yes that would accomplish what I was talking about.

Any means you can find to swap the step & direction signals coming from the parallel port to the drivers would do it.