You could be very correct. The wiring to each switch was replaced when this problem began with shielded wire and they tie to the main chassis. I have noticed a small change in the frequency of the limit resets by different wire routing even then. I'm seriously thinking about ripping it all off the machine once more and rewire the switches parallel and active high as Tweakie.CNC suggested. It makes sense to the debouncing feature of Mach3 but the only hesitation is that a broken connection on a specific switch or wire is not known until possibly too late with NO active high convention. I have some four conductor twin shielded pair wire that would do well.

Thanks,
WAC

I think you need to check your machine for bad grounds, bad ground strategy, no sheilding,etc.

SOme spindle drive put out a ton of noise that if not accounted for create havok for the electronics side. Especially the 0-5v logic side of PC setup.

SOME motor drives are very noisey as well and IF you stack the noisey drive up next to an open 0-5v logic board (;-)

As to the solder joints SOme manf have the basic board built then ADD in there own local devices of choice at a later date.

Just a thought, (;-)TP

ACTIVE LOW is what I have suggested.

Tweakie.

If you are having any trouble with the second Parallel port, that is a big red flag and you may have a conflict in the computer. Check to see if the two parallel ports are sharing and interrupt or have an address conflict. IBM defined these many moons ago, but add-in cards often come pre-configured as LPT3 (and LPT4 plus COM3 etc, if it is a multi card) or otherwise break the rules. It may be that you are chasing a problem in wiring that the PC itself is generating internally.

I can tell you from experience that bad, cheap or failing PC power supplied can make a horrific amount of noise. So much so in some cases that you can actually hear it in the motherboard speaker!

Use the task manager to check the CPU utilization. If it maxes out, it probably would not be a good thing with such a timing sensitve application as Mach. If another process is runing with a higher priority, it can suspend the lower programs. Also check to see if you have antivirus or a similar 'watcher' program running which could be intercepting the LPT1. Parrallel ports were often used for file transfer and networking at one time and it is possible that security software is monitoring it.

The only other though I had was based on my recent discovery that a wireless network adapter was making the whole CNC go nuts. False estops, false limits, and the steppers would suddenly stall for no reason. Unplugging the USB wireless adapter made ALL of the problems go away instantly.

Something else . . and this may be a stretch . . is if you are in an industrial area where some company has radio dispatched trucks. If you are getting slammed by a powerful transmitter, it would probably behave exactly like you are experiencing because it would all look fine until someone keyed the mike and it would just be an amazing coincidence if you happened to be looking at your scope at that moment.

I am still a newbee at this stuff, but still, I would put a spotlight on that PC as a possible suspect.

This machine has not had the spindle turned on yet! I have a VFD and a three phase motor for it but haven't made the exchange from the single phase motor. I'll check the machine grounds and see that the computer case and the machine axis drive cases are grounded as well - as - all one. This parallel port problem may be that it was named wrong  -  I'll have to check on that. Seems to me that I did rename it LPT2 for some reason but it was installed as LPT3 originally and in PCI slot 3. This book PC is brand new, bought specifically for this machine. But, there may be some issues. I installed the additional LPT port a year or so ago and then stored the PC until a few weeks ago.

I get the active high/low convention mixed. I now have active high and you (Tweakie) suggested active low... The actual process is there and correct, I just keep calling it backwards. Someone said a few posts ago that his only problem is his own. Well, I must agree and I'm betting the same here. I have done something stupid that is causing these problems and it's also probably right in my face and I don't see it. 

Thanks for the help!

The nice thing about this forum is that although there must be a million possible causes for any particular problem, everybody is eager to help and sooner or later the answer to your problem will be found.

Tweakie.

Just for your information, it does not matter what the port is called in device manager as far as Mach is concerned, you just need the address. You can put the address of what windows is calling LPT1 in port two in Mach and it will work as long as you set your port number in all of your setup as port 2.

Hood

Thanks Hood, that question was just going around 'in there' as I was changing some names and numbers.

THe active high senario is also a safety feature. IF you had a wire break in you r limit scheme the machine would trigger an error BEFORE you found out by slamming into a hard stop and the limit did not work.

(;-) TP

As a test what about just connecting  a wire direct from the pin to ground on the BOB instead of the switch and see if your problems still exist.
Hood