Thread: Lost steps on A/X axis

Originally Posted by rodm

So it was the BOB then

Please direct all further correspondence to: [email protected]

Oh damn, I noticed you again.

Bob

Try your steps at 136.3636
(200 motor steps X 10 microsteps) / (5mm pitch screw X 44 pulley / 15 pulley)
I am only guessing the calcs above but the only combination I can see you get around 138 steps /mm. I have done a spreadsheet template and you can't argue with Microsoft can you.
If you keep your current setting you have an error of 0.017 mm in travel.

yendor

So, you followed the link yendor.

Your calculations are better than my memory. That is what Mach3 set the steps to.

Since then, I have had further but possibly related, possibly different problems.

The touch pad that I made for auto Z keeps losing the circuit. Luckily, I had someone with me when it last happened and he confirmed that although the circuit would work flawlessly for 5 minutes, it was just as likely to stop working for the next half hour before mysteriously starting to work again. If I tap the connector plate rapidly against the framework, the light on the screen will often miss several blinks before starting to work again.

At the same time as this was happening, I was demonstrating to him how the machine works by cutting a simple name plaque, when machine suddenly lost steps on the X axis again. I stopped everything and then started it all again and no problems.

Check with a multimeter from earth to the touchplate and see if your voltage is steady.

That was one of the first things I did. Rock steady at 5v on the positive side and just as good on the earth side

Bob,
This will need the elctronics experts to chime in on this.
It does suggest to me a computer problem - Adobe or some other intrusive software affecting the Mach3 timing or even in the macro or the macropump (if you are running one) code a bit flakey. Dare I say that it could also be THE BOB as in the electroinc version this time.

Originally Posted by rodm

Bob,
This will need the electronics experts to chime in on this.
It does suggest to me a computer problem - Adobe or some other intrusive software affecting the Mach3 timing or even in the macro or the macropump (if you are running one) code a bit flakey. Dare I say that it could also be THE BOB as in the electronic version this time.

I have thought of many things Rod. It is possible that it is really the Break Out Board and that is high on my list of suspects, but I do not have a pump and the computer has been optimised for Mach.

It is DEFINITELY not Adobe as I do not allow that piece of bloated software on any of my computers. I use Foxit. Much smaller and faster and doesn't insist on compromising the integrity of my Windows computers.

Is there any software that will test the parallel port for speed and continuity and maybe even give a waveform reading like a scope can?

Bob Willson

Hi Bob
If you have optimised the computer it should be OK then.
Have you run the driver test in Mach3 directory? Any large spikes would indicate a problem. Like you have summised this is only going to read the internal timing and not what the parallel port is delivering.
I don't know of anything else that will test it for you - electrickery isn't my strong point..
Did the BOB or anything else giving trouble take a swim in the floods? I suppose you have popped the chips and re-seated them on the C10. Ausxmods have the C10 in stock by the way.
Sorry I'm not much help.

Hi Rod

I managed to save the electrics (apart from the motors) when the flood hit, so none of that was damaged.

Yes, I have run the driver test and it doesn't APPEAR to have any problems.

As to the chips on the C10, I have taken them out and moved them around at least 4 times.

Bob

I've got some broken bobs here if you want to give another bob a try, pin 7 doesnt work on them other than that they seem to work just fine, I use them for test setups they are the MB-03-V1 from homann.

Send me an email or pm with your address if you would like it, should atleast rule out the bob then, I have a feeling it may be the pc but I'm no expert on the electronics either.

Bob
I use cnc c10's and have a number of issues from time to time, but find them pretty reliable for most things, but they can have some idiosyncracies.

When I first fitted my Z-tool zero plate, it worked well for a while then started to give me the same sort of intermittent problems that you describe.

After a lot of head-banging I finally sorted it out.

You have to have the plate "pulled high" first - this means that when it touches it grounds - if you set it up the opposite way it really shouldn't work, but in fact will from time to time - go figure!!

Anyway - if you have it set to "high", you then need to connect a 0.10uf capacitor between the active terminal (in my case pin 17, but it can be any input pin) and ground. This should stop the intermittent fault.

Cheers
Noel

Hi Noel

Thanks for that. The pin is held high and has 5Volts between it and ground, steady as a rock.
I also have a small cap across the terminals from 5v to ground.
The plate circuit seems to be steady enough, it just doesn't seem to work all the time ... as it should. If it works once then it should continue to work.

I am also having some problems with the limit switches and it is possible the two things are connected. It may be a ground loop somewhere or something similar. My equipment is not good enough to measure momentary pulses. And on top of that, I am unaware of any way to actually test for an earth loop.

I am awaiting a second B.O.B. (Shurrup Rod) to see if that makes any great difference.

Bob Willson

Bob
Only other thing I can think of is that most of us run the circuit through the spindle, and I know this can give spurious faults from time to time as the earth/ground is often suspect.
You could try using an aligator clip/wire and earth the spindle to the frame and see if that helps.
Cheers
Noel