I do NOT think 4.5 V is acceptable.

I am a long way away and I do not know, but I suspect you might have a bad or loose connection in the power supply chain. I have seen loose screws, broken wires, and even the terminal block screw bearing down on the insulation rather than the copper in one case.

In the power supply lead to a power tool I found a completely broken conductor - deep inside the cable. Gave me hell for a while.

Cheers
Roger

I do NOT think 4.5 V is acceptable.

I am a long way away and I do not know, but I suspect you might have a bad or loose connection in the power supply chain. I have seen loose screws, broken wires, and even the terminal block screw bearing down on the insulation rather than the copper in one case.

In the power supply lead to a power tool I found a completely broken conductor - deep inside the cable. Gave me hell for a while.

Cheers
Roger

Let's table that thought for a second. I only say this because

I FOUND A NEW SYMPTOM!!

Sorry, but I got a little excited.

When jogging with the ctrl key I thought I was doing something odd like not waiting enough time between steps so when it missed a step I didn't think much about it.

But there's a pattern there.

Using the X-axis alone, I discovered the pattern for the skips is as follows (1 being a step, 0 being a missed step)

011010110101101

That's a pattern of FIVE in a dual coil system.
My brain says it should only be 4 or 8 in a pattern, not FIVE.

So, once I figure out why it has a FIVE step pattern, I think I will be on to the solution.

I did test this over and over and came with the same pattern. This will explain the missing and jumping steps.

Any ideas?

Absolutely no idea at all. It makes no sense to me.
Personally, and this is just simple me you understand, I would focus on the power supply.

Cheers
Roger

I've ordered an oscilloscope but it won't be here for a week. I plan on using it on the power supply and the stepper signals.
I've checked power in just about every area I can with a multimeter and get no discrepancies along the way.

The 24v power supply has the same numbers across all the drivers whether under load or not. If there's a weird signal there, I won't pick it up til I get it on the scope.

The only weird thing I see at this point is the 4.5v signal when I am supplying 5.1v but it's less odd when I was getting 4.45 when only supplying 3.3v. I suspect it's a regulated signal to be @4.5v.

Where is the 4.5 V? If it is on a rail, very strange. If it is on an output, fine.
If it is on a rail, can you identify the convertor chip, and is there a small adjustable pot nearby?

Cheers
Roger

Where is the 4.5 V? If it is on a rail, very strange. If it is on an output, fine.
If it is on a rail, can you identify the convertor chip, and is there a small adjustable pot nearby?

Cheers
Roger

The 4.5 is on the output.
As near as I can remember there's no pot on the board. I'd look but ended up with a lap parasite (neighborhood cat) keeping me from playing with the board.

EDIT


Parasite detached. There are no pots on the board. A total of a dozen resistors only. No way to adjust the output voltage without knowing which resistor to remove and what value to install.

Progress but no progress.

I found the reason the pattern was what it was. It was simple math.

I had my tuning set at 600 steps per inch. That means that the program needed to drop two steps out of each five to maintain the .001" accuracy that's built in by default.
I set the tuning to 1000 steps per inch and it now jog-steps one step at a time with no missed steps.

Then I ran my g-code file and checked and it lost accuracy again. So there's still the original issue. It drops steps or, rather, reverses some. And it seems to do it on slow steps only.

Reversing some steps?
That sounds as though the driver is not sequencing properly. It should step through a phase cycle in a uniform manner, but if it misses a step or two it can appear to reverse briefly. This is especially possible near resonance.

It would be interesting to beg, borrow or even buy a Gecko 203V stepper motor driver, and to test the system with it. These drivers WILL work properly. The owner of the company has a famous published guarantee: if you can blow it up he will repair or replace for free. All he asks is that you keep the metal swarf out of it! Afaik, there have not been many claims on the company.

Cheers
Roger

Hi,
Roger is right, no matter what BoB or even external controller you use or even the steppers you use you'll need good drivers. Even if all the other components are prefect nothing
good will happen unless you have good drivers and Gecko are consistently rated the best. They are not cheap, howvever you only buy them once. Other experienced CNCers on the forum
recommend Leadshine AM882 and can be had for under $100.

Note that both the top line Geckos and the top rated Leadshine drivers are rated to 72V and 80V respectively, that's not a typo. To get the best from your steppers you should be planning to have a
power supply (evevntually) of 60V or more. You simply MUST have high voltages if you wish your steppers to run at moderate to high speeds...don't buy bull********* 50V drivers get the real deal from
the get go.

Craig

I'm not discounting what you guys are saying. I just don't have a way to test it yet. That's why I have the oscilloscope on the way. I can't keep throwing money out there looking for the problem. Once I have it checked with an oscilloscope I will know what's what.

Until then, I tinker. LOL

Also, when I say it appears to back up, it can back up 10 steps and then flip forward again.

I also need to check to make sure I have the coils in the proper sequence. You'd think that if I had one of them flipped then it wouldn't turn at all but who knows at this point.