Machine moves by itself

we need pics, documents and probably some oscilloscope data to help you

Ingo wrote:

Exactly that Todd my card is configured with 5v and my computer is already sending 3v and the card doesn't have a signal filter that goes straight to the drive…. Sorry I didn't understand two things what would be a beakout board and another would be pull ups?

A breakout board is a small circuit board that plugs into your parallel port or parallel port cable, that gives terminals for connecting individual wires that you can then run to devices on your machine. They can be utterly basic and just have connector terminals or they can have features such as optical isolation, signal buffering or more.

Here is quality example of a breakout board with 5v buffering, a C10. They are not very expensive.
www.ebay.com/itm/193156666873

A pull up resistor is a resistor connected to a supply voltage on one side and connected to a logic circuit on the other. This causes the voltage in the logic circuit to "float" at the supply voltage level until the logic circuit is grounded. (A pull down resistor is similar except it connects to a ground.)

"Buffering" a signal, is a way of changing or boosting the voltage and or current level of a signal usually with some combination of isolation chips and pull up and pull down resistors.

Hi,

So I'm suffering with the machine….  With your tip I ended up buying this sign attached, as it was delivered faster because there was a close one and it was fast.  I'm using it on parallel port 1 (with pulse and direction outputs + spindle) and parallel port 2 I keep using my old one with inputs is working almost perfect… The moving part alone stopped is fine.  Then I did the precision calibration there in the “SCALE” which was also good.  Now I'm in a delicate situation where I make a generated program (making a test square) the measurements are being executed on the part is perfect according to the program, but if I release the program again with the same zero part (G54) the tool starts to take material always in the direction of X+ Y+ and Z+, the program ended, restart again even G54 it takes material, it always takes it in the positive direction without changing the zero point of the G54.  I thought the mechanical square part was off but it's not perfect either.  I have the impression that it works on tolerance, but is leaning towards the positive direction.  In fact, I checked the feeler clock and it was, so I checked the granite square and it wasn't…..

Ingo

From your description it sounds like you are either loosing some missed steps or are gaining a few extra steps from noise. This sort of thing can be difficult to diagnose remotely.

Noise related issues are usually because of wiring mistakes such as poor grounding practices (ground loops), unshielded or improperly shielded wires ran near noise sources.

Missed or lost steps can be because of improper step timing settings, too high acceleration, too high velocities, or bad latency on the host pc.

There is also the possibility of a mechanical issue resulting in lost motion. A common example would be a loose shaft coupling or pully slipping on the motor shaft.

Hi Todd,

I'm in a doubt….  Before, the machine moved by itself and I really changed cables position, I improved grounding, I changed the Bob and this problem solved it doesn't move anymore.  I'm struggling and I applied Renishaw's Ballbar I managed to leave the interpolation very zero, it was pretty cool, but Ballbar didn't catch this problem... Every time I release the same program it always adds (increments) for X+, Y+ and Z-.  Do you really believe it's due to grounding, noise, etc…?

It is possible, but more likely a step timing issue. Sometimes a simple thing such as inverting the step and or direction pin value can cure something like this. (May also have to change the scale sign when inverting the dir pin.) If the signal is inverted from what the drive is expecting, it can make the direction change setup timing wrong causing a missed step or two at the beginning or end of a move in one direction. Or too high accelerations can cause missed a few missed steps at the beginning or end of a move.

Hi Andy, so I still haven't managed to improve on this issue.  I didn't understand Todd's position on the question of reversing the step or direction pin?  Was it electrically or in Hal?  When I need to invert the direction, I do it on the scale (this didn't work).  The acceleration is at 300. I tested it with another computer, it didn't work, I bought another card, I'm going to test it, I'm out of options…

do you have your drive manual ???