Messages

121

General Mach Discussion / Re: Unprogrammed Z axis down feed

« on: July 20, 2011, 11:44:06 PM »

Thanks ERP.

I'm thinking it's a noise problem on the step input to the driver.  Probably related to power line conditions as it's very intermittent.  Since the last direction for the Z axis when cutting is down, any noise on the step input big enough to trigger movement is going to result in down motion.  I'll search the forum for advice on power line noise suppression.

The lack of change in the DRO value does seem to vindicate MACH3.

I was kinda hoping someone else out there had a similar problem and a cure.

Peter

122

General Mach Discussion / Unprogrammed Z axis down feed

« on: July 20, 2011, 09:43:13 PM »

I've searched the forums for my problem but can't find any similar ones so here goes.  Every once in a while (usually less than once every several days), my mill does a z axis ramp down while cutting a part.  This has happened with G code that has been used successfully several times before and not altered since.  The same code will work fine again later.  The result is a ruined piece and often a broken cutter.  The error never occurs upwards, always down into the piece.  The DRO value does not change.  The ramp rate appears to be at the specified plunge rate for the cutter.  The ramp will stop on it's own, then may start again further on.

Any help would be appreciated.

123

VB and the development of wizards / Re: How to do protected moves for homemade probe?

« on: October 24, 2010, 11:46:47 PM »

I have read (and I apologize, I cannot remember the source) that a G31 move that goes to the end of the programmed distance without a trip is considered a fault.  However, I haven't found any error flag/register in Mach3 so I guess I'll have to craft my own as I described.

I also read that a G31 usually results in some overrun but I haven't seen that.  It could just be that my feed rate is very low.

At least software changes are easy to do.  The only costs are time and any collateral damages from mistakes.   

Either way, the probe is working great and I'm having a ball.  Repeatability to .0001" and I've got it centered to about .0003".  I suspect my spindle bearing isn't much better if at all.  The screen set has everything I want in it and it's a snap to add/modify.

Regards

124

VB and the development of wizards / Re: How to do protected moves for homemade probe?

« on: October 24, 2010, 10:08:17 PM »

Okay, I anticipated the need to move off the trip point but why does Mach 3 go into 'hiccup mode' if the probe is already activated when the G31 is executed?  It knows there is a fault because it tells me in the message window.  I would have anticipated a simple exit from the script.  Instead it starts moving all three axis.

The whole reason I'm using G31 is for the protected move.  I don't want to have to rebuild my probe because of a setup error.  I know I can't prevent a crash into the side of the probe but at least I can avert disaster from ramming the tip into something.

I don't know what you mean by a protected trip move.  Some protected moves are just to reposition the probe and no trip is expected.  I can detect an error if the actual travel distance does no equal the expected.  For moves that expect a trip, one just measures the actual distance and makes sure it is less than the programmed distance.

Thank for the fast response.

 

125

VB and the development of wizards / How to do protected moves for homemade probe?

« on: October 24, 2010, 09:14:22 PM »

My apologies if this question has been addressed elsewhere. I have attempted due diligence both here and on CNC Zone but have not found an answer. I might
have missed it because of the sheer volume of information and posts. The closest data I found related to commercial probes with vendor supplied software.

I have just completed a homemade contact probe. In an effort to avoid crashing it through dumb setup errors, I wish to incorporate protected moves in my
routines. To do this, I envision using G31 for all moves, then testing the results to see if either I got to where I was supposed to and didn't hit something when I wasn't supposed to or I got to the end of a probe move and didn't hit something when I was supposed to.  I have empirically determined that a G31 move with no probe contact acts like a G31 with a probe contact at the end of the programmed move. In others words, there is no error detected and, at least for slow feed rates, the probe is sitting at the end of the move.

However, this creates a problem. If the probe does make contact, the next move will start with the probe activated, causing Mach3 to object saying the probe is
already activated. Mach3 then proceeds to start moving all three axis in small steps, about 1 full step, 5 times a second. This won't stop until E stop is
activated.

What's going on here? I suppose I could do a small G0 to get the probe off the piece before doing the next G31 but I think something is wrong here.

Thank you for your patience and help.