EZ-TRAK - “Possible Pointer Error - General Protection Fault”

Esteemed members, I would greatly appreciate your help with a rather serious problem I am having with my Series I EZ-Trak. My nearest machine tech is a 6 hour drive away in Denver!

I did a search on the forum and found this thread, but I believe my circumstances are a bit different, possibly requiring a different approach.
http://www.cnczone.com/forums/showth...=pointer+error

In a nutshell:
1.) Machine generally boots up fine. BMDC countdown to 14, parameters load, XY&Z all home like they should, all seems good.

2.) After hitting some sort of keystroke (could be anything), an error message pops up saying “Possible Pointer Error”,
http://www.louderthanwords.us/smallf...5854686791.jpg
immediately followed by “General Protection Fault – Fatal Error Software will try and clear error”.
http://www.louderthanwords.us/smallf...5854686792.jpg
I hit enter, and get one of three scenarios:
a. Screen completely freezes and goes white, requiring reboot OR

b. One of the following screens appears, and everything freezes, requiring reboot
http://www.louderthanwords.us/smallf...5854686793.jpg
http://www.louderthanwords.us/smallf...5854686794.jpg

c. Screen briefly goes to a startup screen, then goes right back to what I was doing. This may happen 2 or 3 times before finally going to 2a or 2b.

3.) Problem has not occurred while actually running a program, or just spinning the dials. Sometimes the error occurs right away after homing, sometimes the machine runs for a while (10 minutes or so) before crashing.

Machine Particulars:

 3 Axis Series I EZ-Trak, c. 1998 (late DX early BPCM)
 8mb Flashdisk memory (i.e. no Hard Disk Drive)
 Version 6.0/5.78 software
 Z axis encoder
 Floppy in rear
 230 Volt
 Machine runs on generated 3 phase via Phaseamatic rotary phase converter. PC power supply is wired separately from motor/drives; runs directly from single phase 120v. I generally only turn on the phase converter and main switch at the back of the machine when I’m just about to cut a part
 Measured line voltage to PC power supply is 123v.
 Flashing green LED on BMDC card after startup. Green lights on all axis drive cards when powered up.
The following all seem to work fine: power supply fan, CPU fan, main fan

I have owned the machine for 5 years. Though it is used daily for prototype and gunsmithing work, use would be considered “light” compared to production mills. It has been relatively trouble free up to this point.

Stuff I’ve Tried/Other info:

1.) Went into CMOS to be sure “LBA” was not turned on. The setting here was “Normal”.
2.) Went into DOS via the “Utilities” screen, then “4 COMM” and tried to C:\defrag and C:\scandisk. The machine came back with “Bad command or file name”. I was later told this is N/A because I don’t have a Hard Disk Drive, but I don’t know much about this stuff.
3.) Reloaded software from my Bridgeport 6.00/5.78 floppy. This seemed to help for a while, the machine ran for 4 hours before the crashing started again. Oddly, I began to intermittently lose my Z axis DRO reading (i.e. reading did not change) when I switched from 3 axis to full manual.
4.) Tried booting direct from floppy. Machine doesn’t seem to want to do that, it only wants to load the software onto “C:” I checked in the CMOS/BIOS settings to be sure the machine would load from A first. I even removed the Flashdisk card from the motherboard to get it to boot from A. It did not want to boot at all. I put the Flashdisk back in, and I’m right back where I started.

Other:
This MAY be totally coincidental, but if memory serves, the problems seemed to begin around the time I started loading G-code as .TXT files (via floppy) that was written on my laptop PC. These files did 3 axis contouring and ran just fine. Up to that point, I had just been doing PGM files right on the EZTrak.

The same problem DID occur once or twice last summer, but just went away on its own.

CMOS battery was changed a few months ago, as I was getting “CMOS checksum errors” at startup and finally got tired of plugging in a keyboard to bypass. Didn’t really know what I was doing, just pulled the battery out and stuck a new one in of the same type.

Quick shot of the main guts, we are looking straight on at the 8mb flash card:
http://www.louderthanwords.us/smallf...4636328543.jpg

Well enough rambling, again I would appreciate any help, or pointers in the right direction! Many thanks for enduring to the end of this post!

Stan

The other post said:
>I have also seen GENERAL PROTECTION faults with flaky power suppies, >overheating processor, or memory getting flaky.

Check and/or clean your power supply (blow out with some air)
be sure fan is turning.

From you pictures the machine looks very clean inside.

Reseating the memory chips can't hurt.

When you changed the battery, did you reset the CMOS and re-enter the correct data? If not, there may be some funky code in there.

floppy drive error won't give this type of info.

If you have a speaker on the MB, does it beep on startup?

enat

Start looking for a motherboard, especially if yours is the OEM original.

We had similar problems - first we lost the HDD/FDD controller, then we had boot problems. After replacing HDD/FDD card (ours had one), things were OK for a while but then started to get flakey. If you're not reading the flash drive, CMOS could be corrupted or M/B could be going flakey. Afterall, the M/B's are only desktop units that might not have been the 'best' for use in a hard core commercial CNC machine, especially one with and unsealed cabinet in a shop environment.

IF you have a HDD/FDD card or M/B going bad, you often won't boot and then you might lock up during operation.

Once we replaced the M/B, the thing booted and has run perfectly since with no other changes - period.

Finally, the fact that the thing works OK and then craps out after time does suggest electronic gremlins are creeping in once the M/B and/or other support boards are getting warm.

Keep in mind that a flakey power supply might also cause some issues. However, our experience is that power drop outs from a P/S glitch result in reboots.

Also, check the capacitors on the MB.
The Abit PX5 that's in there was of the year where many computers
had the problem with bad capacitors. The row of capacitors to the right of the CPU are for the voltage regulator. If they are swelled or loose they are bad. And the tops should be flat not bubbled, if not flat where the 'K' is they are going bad. I've only had one MB with this problem, a SOYO, but I've used many Abit MBs for tears and never had a bad one. But in the Abit newsgroup you'll find many did have problems with capacitors.

enat

I have a Bmdc in an easy path lathe and get the same GP fault at random intervals. I found that by making any keyboard entry every few minutes will keep the faults ( gremlins ) away. I suspect the screen saver may be timing out and messing with dos. Any suggestions here?

My first answer was deleted by another moderator.
This particular EZTRAK was one of a very narrow batch of production. The top of the flash card has jumpers that are set for a particular mother board. There were only 2 that Bridgeport used with this flash card. The flash card itself requiresd the CMOS to be set to NONE for the hard drives. A special piece of software made the flash card bootable and a very bare (limited) DOS was installed thus the reason you do not have SCANDISK or DEFRAG. The flash card was NOT a large storage device. Only 8 MB. Thus if you are loading 3D files on it, it may be full.
GP faults (general protection) are usually generated by the mother board. It could be the logic power supply, heat, software, memory, processor or the mother board. Reseat the processor, and memory. Check the logic power supply. I have even had a bad floppy drive create faults. It was the last item on my list of possibilities.
Loading the Bridgeport software does not replace the boot software nor the DOS files. another factor is that these old flash memory did not have the read write cycles of the newer DOM (disk on a module) Even the disk on a chip used by the next generation of mother boards used in these machines were known to have a short life span.

George

Screen savers are notorious for creeating havoc with programs. It has been my experience with the BMDC code that it wants to run, by itself, on the system.

We tried to get it to run in/with DOSSHELL and some other stuff and, apparently due to the memory management scheme of the BMDC, it would not cooperate.

If you have a screen saver loading in the AUTOEXEC.BAT or CONFIG.SYS programs, deactivate them as part of your trouble shooting effort.

If you only have the 8MB flash, perhaps it may be possible to upgrade to a newer, larger DOM. You may have "used up" the read/write cycles of yours and it is now starting to go south....

Wow, many thanks for all the responses, suggestions, and help! There is a bunch of good information in this thread.

In short, I tried all the "easy" troubleshooting things, i.e. power supply, reseat cards, etc. Nothing worked, and every minute the machine was down, I was not making money.

As suggested here, I went and got a new MB and 64M DOM. Got them from EMI. No, it was not cheap. BUT Mark at EMI was extremely helpful, patient and knowledgeable. I had the stuff the next day. For a chip sweeper like myself, it was a bit intimidating pulling out all these cards and wires and such, but I went real slow and kept track of where everything went. At the end of the day, I figured almost 10 years on a MB, not a bad idea to replace anyways...I mean, I put new timing belts in my truck before the break, right? I am now budgeting for a new MB and memory every few years, as long as I can get them.

So far it is running great after a few days and I'm breathing a huge sigh of relief for now. As you know it is very different owning a machine to make a living vs. one for fun. It was a very stressful time, esp. due to my remote location, and I am truly grateful for all the help posted here. Machinetek, NC Cams and the rest, you are great resources to the community!

Couple other things I learned, that may be of help to others.

When I bought the machine, it had a Fram air filter on it, with the model # written on the cabinet. So that's what I bought and used on the machine. Turns out this is completely the wrong type filter, even though it fits. The real Bridgeport part has a lot less air resistance than the auto filters, allowing greater airflow to the cabinet. I switched filters and the difference was dramatic. Pains me to think for several years, the electronics weren't getting the cooling they needed. Thankfully I work in a very cool environment.

Write down all those settings/parameters! I hadn't bothered, so I had no idea what my CMOS etc. settings were to begin with. I went and took pictures of all those screens for the future. From the main menu, *11235 gets you to the parameters.

Re: the Fram air filter.

I seriously doubt that, if you have the RIGHT automotive air filter, that you'll be able or need to suck in as much air to cool the BMDC mother board and internals as you need to feed a decent sized V-8.

We did some searching and found some automotive filters that, when stacked 2 or 3 high, EASILY fed plenty of air (checked with water monometer for pressure drop in cabinet).

These were readily serviceable via local K-mart or other automotive repair parts sources and a fraction of the cost of the "Bridgeport" filters, even when you buy several. BUT YOU HAVE TO GET/USE/SELECT THE RIGHT ONES!!!!

The advice about CMOS settings and problem symtoms that crop up when motherboards and/or other hardware start going flakey are good to document. Maybe somebody will research them the next time they have problems instead of re-instituting a thread to reinvent the wheel.

After all, there are only so many ways to boil a 3 minute egg....

I am getting the same "Possible Pointer Error" on my 1995 EZ Path CNC Lathe. Did you ever solve the problem? And if so, how? I would appreciate any help you could provide.
Thank you.