Hello Guest it is April 16, 2024, 11:07:16 PM

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - BluePinnacle

Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 »
Another possibility... our power is not the best. I did have a sag a week ago which forced a reboot; luckily the chargepump board dropped out and I heard the spindle slam to a stop before I'd spotted the lights go down. Everything stopped dead, just like I'd designed it. :D I'll see if the IT lads have a spare UPS anywhere.

Another thought - USB memory sticks make mine go mad, the steppers bang and groan and won't work worth a sht. It's impossible to do anything. Funny business  ???

Edit again! i checked - the stalling at the end of the sub is down to a missing carriage return. Very peculiar, I'd never have thought to check for that. I'd just gone through and deleted any and all unnecessary spaces. I'll make a point of putting in

(leave this line in, mach3 needs a carriage return above here) at the end of my subroutine file. Cheers for the tip :)

Hi Jim

Offset tables - Going by what you've said i think we can rule these out. I'd done several parts that day, the usual routine being followed thus:

start computer
reference all axes
jog to approximate start point
fit blanks
load G-code
manually G0 to x100 y130 z10 (nominated start/return point)
run cycle
repeat until hometime

When I started the computer today and did a spindle-off thin-air run, it went to the right places on the fixtures with no surprises, so i can only imagine that my offset tables are robust. I've never zeroed DROs in-flight. Wierdly, it now stops at the end of a subroutine a few steps from the end of the program! arrrgh.

The other possibility you raise is that of reading ahead into a subroutine, I wrote all mine into the same text file. This may be giving it headaches. I'll re-write using seperate text files for subroutines and see if this helps.

There were no orphan spaces in the code btw.

thanks again :)

Hi, thanks for the tips so far, I'll check on these.

I didn't check the offset table, but I'll do that today. After the first crash I re-homed the machine and carried on - somebody had stood a part on the keyboard ( !!! ) so i put the spectacular crash down to bad luck and dumbness. The second time I was stood next to it when it halted halfway through a safe move, fed Z down onto the part and stubbed out a brand new 8mm cutter like a freshly finished smoke. I punched the Estop button and halted everything, and the G-code readout on the program run screen was displaying a Z move which is part of a subroutine for boring a 10.2mm hole (feed down, circle, feed down, circle, pullback, rinse-and-repeat). It seems to have jumped there from halfway through a move. Why? This is getting expensive! And it's making me look stupid! I'm not happy! More to the point I've done 40 or so of these parts already with only minor edits for gradually raising feed rates and reducing cycle times. I haven't altered any of the wiring or changed PC settings at all.

I'll investigate and get back on this, thanks again for the suggestions thus far. I'll let you know what I find. The code is pretty beaten-around so there may be blank or orphaned lines in there.

jim - What am I looking for in the offset table? Would its values change for any reason? ta.

more later ... BP

I've got a program spread across two fixtures for making a large waveguide flange. The start point is at G54 X100 Y130 (left hand fixture), and the whole cycle finishes there, returning from a similar position on the right hand (G55) fixture. This normally executes well, using a 3-axis movement to clear the fixtures and a fast return to the start point. I've done about 40 flanges so far, and they're all good, but today Mach3 has run away twice. The first time it drove the cutter up (Y+) without any provocation, shredding the jig and snapping the cutter. Secondly, without any warning, it was returning to the start point when it seemed to hop to a subroutine I'd written for boring a 10mm clearance hole. Again, the cutter broke as Z fed down and stubbed it into the part.

This isn't funny, I've trashed over £60 of carbide cutters and aluminium blanks today and we're going to be in trouble if I can't sort this out. Any clues?

General Mach Discussion / Re: random e stop error
« on: November 20, 2008, 12:02:30 PM »
could be earth leakage, that held me up for a while. It takes very little current to ground a pin. Use a seperate supply like a 'phone charger, feeding your estop and limits, and use it to hold a relay closed. Wire the relay contacts to your logic pins.

General Mach Discussion / Re: Deleting spammers
« on: October 09, 2008, 04:48:22 PM »
it's a possibility. This isn't usually an inflammatory board and there are lots of responsible adults.

General Mach Discussion / Re: Home and/or limit switch
« on: September 12, 2008, 07:42:12 AM »
Jim - good call, mine uses DIP16 packaged darlington phototransistors (ISP845X). Four isolation channels per chip and 5KV isolation. nice sharp switch-on with very little forward current required, ideal for a first stage from one's LPT. And like you say, very cheap :)

General Mach Discussion / Re: Home and/or limit switch
« on: September 11, 2008, 11:31:09 AM »
Spindle control is pretty easy and by implication so is coolant control. Of course the cheaty way is to have the coolant pump in series with your spindle motor (if you don't use reverse) and switch it out if you want to cut dry. I didn't think of this, put in software control and the computer runs the whole thing.

A quick opto-isolator circuit will drive a small 5v coil relay, and the relay, if selected for its contact properties and isolation values, will drive a hefty 3-phase contactor with no gripes. While we're here, mobile phone chargers power relays and opto circuits very nicely :)

General Mach Discussion / Re: Fried 2nd Computer (Newbie)
« on: June 06, 2008, 07:15:05 AM »
Words to the wise, if you want a quick 5v supply, an old Nokia phone charger does very well. Almost everybody i know has old Nokia chargers kicking about somewhere.

General Mach Discussion / Re: PCI port card addressing woes
« on: June 06, 2008, 07:11:49 AM »
Who doesn't need an emergency stop? I don't like the sound of that one bit.

This has been solved anyway, thanks for the help, i think we've established that MosChip boards are the way to go and not the cheapy triangular ones. :)

Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 »