[chozo]

I have uploaded a high-res pic of the board I took with my camera.

It's around 220kb, so here's a link:
PICTURE

[sbrpollock]

Sorry.....my first suggestion is almost useless.....

I see that this board has a Db 25 connector to connect your parallel port cable directly to the board. This won't lend itself to swapping the step & direction signals as I suggested earlier. I thought that the board had a terminal strip where you hooked up the lines from the parallel port cable.

You could try using the method you mentioned earlier using some Db 25 connectors in between your parallel port cable and your driver board to be able to manipulate the signals comming from the parallel port; but this may be difficult to do. This is what a "Break Out Board" enables you to do.

Somebody here has to have a better Idea than mine, Anybody?

[chozo]

I just went to Radio Shack to buy another gender changer. I'll just run wires in between the two female connetors with the exposed ends folded over so they stay snug, and switch the wires around(2,3 for x with 4,5 for y) to go along with your idea.

I'll report back with the results.

[pminmo]

Patrick,
Both are good suggestions. Just from the picture on Hobbycnc's site it looks like there is an electrolitic associated with each module, but nothing else. It would have bee good to have a .1uf for decopling in addition to the electrolitic. I don't think those modules were meant to be driven directly from a parallel port.

Phil

[chozo]

Patrick,
Both are good suggestions. Just from the picture on Hobbycnc's site it looks like there is an electrolitic associated with each module, but nothing else. It would have bee good to have a .1uf for decopling in addition to the electrolitic. I don't think those modules were meant to be driven directly from a parallel port.

Phil

There is a .1uf for each axis. I didn't even notice it 'til now, but in my picture I uploaded, the chips are blocking the view.
You can see them here:
http://www.rpgdesign.com/hcnc.jpg

[pminmo]

Chozo,
You may windup putting in a buffer/signal conditioner like this as a fix. Where signal 4 and 5 is from the computer, and the corresponding outputs go to 4 and 5 of the driver board.

Phil

[chozo]

Chozo,
You may windup putting in a buffer/signal conditioner like this as a fix. Where signal 4 and 5 is from the computer, and the corresponding outputs go to 4 and 5 of the driver board.

Phil

Thanks. Sadly, I'm an electronics noob. I'll try to gain some knowledge in this area.

[sbrpollock]

I want to make sure that I'm not stepping on anyones toes at HobbyCNC, I don't think this is a design problem as this is a finished product they are selling and they have surely dealt with this issue. If anyone from HobbyCNC reads this please don't feel I'm being critical.

Though, I think that I would have done a little more to decouple and filter, I really believe this is a tested design, and that decoupling is probably not the problem unless one of these caps is bad (It wouldn't be the first time).

Allegro recomends that in addition to the decoupling, the inputs should be driven with low impedance drivers to minimize noise vulnerability. Since in situations like this I never have any control over what the parallel port is using as drivers, I always include Opto-Isolators (Preferably with Schmidt Trigger type inputs). This way I'm confident I have good signals driving the Allegro device. This also serves to protect the PC parallel port from those occasions when I lay a screwdrived down in the wrong place while the circuit is in operation

All of this might be a moot point if chozo swaps the step & direction signals and the problem moves to a diferent axis.

Now that I think about it, Do you have more than one parallel port cable? Has this been eliminated as a posible cause?

Also, in the picture, from left to right which of the drivers is your Y axis driver? If it is the second one from the left, take a look at jumper J2, it looks a little out of line with the rest. Is it securely connected to both pins?

A buffer driver is not hard to make and may make up for a parallel port that is not getting the job done, but it sure doesnt seem like it should have to come to that.

[sbrpollock]

Chozo,
You may windup putting in a buffer/signal conditioner like this as a fix. Where signal 4 and 5 is from the computer, and the corresponding outputs go to 4 and 5 of the driver board.

Phil

Phil:

That's what I was thinking in the last line of my last post, either you guys posted while I was typing or I'm just going blind

[chozo]

I want to make sure that I'm not stepping on anyones toes at HobbyCNC, I don't think this is a design problem as this is a finished product they are selling and they have surely dealt with this issue. If anyone from HobbyCNC reads this please don't feel I'm being critical.

Though, I think that I would have done a little more to decouple and filter, I really believe this is a tested design, and that decoupling is probably not the problem unless one of these caps is bad (It wouldn't be the first time).

Allegro recomends that in addition to the decoupling, the inputs should be driven with low impedance drivers to minimize noise vulnerability. Since in situations like this I never have any control over what the parallel port is using as drivers, I always include Opto-Isolators (Preferably with Schmidt Trigger type inputs). This way I'm confident I have good signals driving the Allegro device. This also serves to protect the PC parallel port from those occasions when I lay a screwdrived down in the wrong place while the circuit is in operation

All of this might be a moot point if chozo swaps the step & direction signals and the problem moves to a diferent axis.

Now that I think about it, Do you have more than one parallel port cable? Has this been eliminated as a posible cause?

Also, in the picture, from left to right which of the drivers is your Y axis driver? If it is the second one from the left, take a look at jumper J2, it looks a little out of line with the rest. Is it securely connected to both pins?

A buffer driver is not hard to make and may make up for a parallel port that is not getting the job done, but it sure doesnt seem like it should have to come to that.

The J2 jumper isn't a problem. I messed up it's looks(I have a jumper puller, but I was lazy and used pliers), but it's fine. I took it out, and put it back in, but that was after the problem was known.

I have tried two different cables, and it's the same.

I still haven't swapped the signals, so I'm curious as to how that turns out.

THANKS EVERYONE! You guys are great.
Hopefully this will get worked out.

[balsaman]

Lots of people are using this board sucessfully. My suggestion is to contact HobbyCNC before you go any further. he may have an answer and it will save you a lot of headache. Plus, he is excellent to deal with.

Eric

[badbubba]

Chozo,

I had a similar problem. Some how I had missed soldering one of the pins on the controller chips.

Bubba

[chozo]

Here's an update:

I was getting tired of cutting a million wires for the parallel cable to try to switch the step/dir signals for X and Y.

So, I remembered that I had the A-axis available, and also that I had been wanting to try it before.

I hooked up the motor that was at the Y-axis to the A-axis connection.

I changed the setting for the Y-axis in Mach2 from pins 5,4 to pins 9,8 which are normally for the A-axis.

Works fine.

Does this tell anyone anything, or no?

I don't really want to settle for this "solution"(although it does the job) because the Y-axis still has a problem somewhere along the line.

badbubba: I did examine the Y-axis chip when I decided solder in the A-axis chip. I didn't see any broken connections or anything out of the ordinary.

Thanks, everyone.
I'm still interested in solving this, of course.

[sbrpollock]

As I said earlier, I believe this is a well tested product and that HobbyCNC has probably already dealt with this. BalsaMan has also said that a lot of people are using this board successfully and that HobbyCNC is great to deal with.

Although I enjoy troubleshooting type work like this, it is not a good justification for reinventing the wheel! Try contacting HobbyCNC like BalsaMan suggested.

But, Please do let us know If you find out what the problem is!

I'm sorry I didn't suggest this myself sooner as sometimes I get caught up in the idea trying to find a problem myself instead of using the most obvious help available. This is O.K. when I'm wasting my time on my equipement, but is kind of rude when I'm wasting someone elses time when an answer might be available quicker somewhere else.

Sorry about that.

BalsaMan.....Thanks for the catch before I made more of a jack*** of myself!

[anoel]

Yeah, Dave is excellent to deal with...

But I'll chime in... I built the same board and I had a problem that sounds just like what you describe... I had the orientation of the resistor network for my axis installed backward. Check the orientation on that resistor network on the axis that is not working. And make sure that the dot on the RN is oriented the same as the other axis that work correctly.

"I CANNOT BELIEVE THAT THIS THREAD GOT TO BE 11 PAGES LONG WITHOUT CONTACTING HOBBYCNC AT SOME POINT. WOW!!!"

[pminmo]

Good put on the resistor network, that could definitely affect how one axis works.
I too believe Hobbycnc should be contacted. In fact I thought I had suggested that, but looking back I don't see that I did. Tell us what you figure out.

Phil

[chozo]

Yeah, Dave is excellent to deal with...

But I'll chime in... I built the same board and I had a problem that sounds just like what you describe... I had the orientation of the resistor network for my axis installed backward. Check the orientation on that resistor network on the axis that is not working. And make sure that the dot on the RN is oriented the same as the other axis that work correctly.

"I CANNOT BELIEVE THAT THIS THREAD GOT TO BE 11 PAGES LONG WITHOUT CONTACTING HOBBYCNC AT SOME POINT. WOW!!!"

I'm contacting HobbyCNC now. LOL Yeah, I suppose that could've been done earlier.

The RN for the Y-axis is orientated correctly, sadly.

[chozo]

Okay, I've been in contact with Dave from HobbyCNC. He's been great.

To sum things up, here's what I had to do:

I basically looked over every single solder joint, and retouched almost all of them, making sure they were as smooth and full as possible. When that was all done, I thoroughly cleaned the board with alcohol and toothbrush.

Hooked it up. Tested each axis individually, one at a time.

X: PERFECT
Y: PERFECT
Z: PERFECT

Then I hooked them all up.
ALL: PERFECT

I thought I did a good soldering job when I first did the job, but now I know exactly how precise and clean you have to be. I'm better now because of it.

THANK YOU TO EVERYONE!

I'm sorry I took up your time, when the problem was completely my fault, and wasn't anything you guys could've guessed.

I'll try to make sure I eliminate any possibility of error on my own part if I ever come to you guys with another question.

Thanks again.