[DerHammer]

I own an older cnc milling machine with a Fanuc 0ma controller on it and earlier this morning, the controller appears to have partially died. When I say partially, I mean that I can hear the machine make a power on and power off sound with the on and off buttons but that's about it. Both the monitor and indicator LED's on the controller buttons remain off.

I have tried to reset the breaker inside the power cabinet but have had no success. My next thoughts were to start pulling fuses and checking them individually but I'm hoping some of you experienced fanuc experts out there may have had similar experiences and could shed some light on the subject. Any suggestions are welcomed and appreciated. Thx

[WGoyer]

On the far right of the mother board is the black power supply. If you stoop down low and look up into the power supply, you will see a small bunch of test pins, each having a number indicating the voltage. Start with the one that has a "5" and see if he 5 Volts is up. Next, check the 24 Volts. The 24 volts can go all over the machine on some integrations, but mostly are used to power the pushbuttons and limit switches. If the 24 gets shorted due to wet limit switches etc, the control won't start. Also, the power supply can just go bad. Be EXTREMELY careful not to short the pins together. This could let some smoke out. Getting the smoke back in takes lots of money.

Warren

www.uptimecorp.com

[GPena]

If all voltages are ok, you can try to reboot the system pressing [reset] and [delete] buttons in the keyboard at the same time, then turn on the control and do not release those 2 buttons untill something appears on the screen, usually it won't take more than 10 seconds.

Please be carefull since this procedure will erase all memory contents like parameters, diagnostics, programs, offsets. You must have a backup before you do this.

Good luck.

Greg.

[DerHammer]

Ok, So I got a hold of a local service tech in my area and he gave me a few short walk through procedures to try over the phone, so hopefully I can update those watching this thread and further narrow the problem down.

I started off by turning on the machine as far as it would go with the cabinet door open to inspect what LED's illuminated inside. It appears that I do have a red "WDA" LED illuminated on the controller main board. I believe this stands for "watch dog alarm" and I am currently under the impression this indicates anything from a main board failure to a power supply issue.

Is there any other safe preliminary tests short of testing the power supply voltage (and potentially letting out the magic smoke), which I could perform at this point before bringing a tech in to figure her out? Thanks again to everyone for all the help.

[GPena]

It is maybe the master board, but first intialize the control with the procedure that I gave, this could solve the problem without any PCB replacement.

Remember the backup!!!

Greg.

[DerHammer]

Sounds like its worth a try, the only problem I see there is that I don't exactly know how to back up the parameters. Especially considering that it would require booting the machine up to a point where I could control it enough to initiate the output file. I have no display or control whatsoever the way the machine is sitting currently, so I'm not exactly sure how this would work.

Hopefully, this I'm not missing something here which is making me seem even more newbish than I am

[DerHammer]

A quick update for those watching this thread. A tech came out today to diagnose the mill and it would appear that there is a hardware failure in either the GPU card or the main board. As many of you know, pretty much anything fanuc is extremly expensive to fix, so if anyone has any good leads on where to get ganuc OMA controller parts, it would be appreciated.

Also, Many thanks goes out to everyone who has helped on this too!

[f-bu]

If you have some additional boards like PMC-M or additional I/O, I/O+4th axis etc. try starting the CNC without these boards. If it starts, there will be many alarms of course, but you can decide if the problem comes from the mother board or not. If it is a PMC-L machine, i.e. no additional board, you can also try removing PMC Eproms - marked 011, 012 on model A, 0E1, 0E2 on model B. If CNC starts with "no PMC program" alarm, there is the problem. I had such a case on 0MA with defective PMC eprom and WD alarm.
There is a check pin named "PARITY" or something similar. Shorting it to 0V check pin starts the control without parity checking of main EPROM memory. This could be your next test.

[DerHammer]

Thanks for the Info F-bu. You bring up a good point about having faulty E-proms. We checked the voltages off the main board test posts as well as the power input unit I believe its called. We also checked the battery voltages, which were also all good. We then removed the E-proms, cleaned the terminals and inspected the chips to make sure all the pins were straight and reseated them but the problem still existed.

However, We did not try to start the machine without the E-prom all together. As long as this doesn't run the risk of burning it up further, I don't see what it could hurt at this point. The fanuc Tech we called over the phone told us that a WDA alarm indicates a problem in a main board or GPU. As far as additional I0 boards go, its a fairly bare bones system. I dont have anything additional other than the GPU card and some expanded memory. I will try powering on the machine without the E-proms and let everyone know. Thx again

[dpuch]

There are several 3rd parties that repair/swap fanuc parts for cheaper than fanuc prices. Just google the part number, or perhaps fanuc repair. I think usually around %70 of fanuc prices. A calculated risk not dealing with Fanuc repairs though. The few times we have done this have worked out well enough though. You need to know what part you actually need and what your doing though when replacing stuff.

[DerHammer]

Thanks for the info Dpuch, I actually have considered that alternative but as you put it, its always a calculated risk when the parts are coming anywhere other than the manufacture. Fanuc has quoted me $1804 to mail off my board and have it repaired. It would come back with a 1yr warranty against the WDA alarm going off again. Also, I did try the delete/reset combo at boot with no success. I think the WDA alarm is triggering before the machine even sees the keyboard.

There is still a slim chance the issue could be my Graphics card, but more often than not from what I understand.. Its usually the main board when dealing with this alarm according the fanucs service department. Especially considering that the GPU is among the last to boot in the startup sequence.

I have a Fanuc tech coming out tomorrow with parts in the Van, so ill keep everyone posted who's watching. Thanks again

[DerHammer]

Quick update for those still watching this thread. The Fanuc tech came out today (for a hefty fee of course) and confirmed my fears. It would appear that My Master controller board is indeed....toast.

So, now I am in the process of trying to recover my diagnostic parameters from the machine builder and weighing my options financially. It looks as though I can either have Fanuc repair it for $1804 with a one year warranty or I can get a replacement board from an outfit I found on google for $1200. They too offer a one year warranty but I have mixed feelings about going outside the manufacture. Again, if I don't recover those parameter settings, Its a 4000lb paper weight anyways but I don't expect this to be the case because luckily the machine builder is still in business.

The only other option I can think of is an in-expensive retrofit kit if there is such a thing. Does anyone know if this is this a realistic avenue to take while trying to stay in the price range above? Also, what kinda accuracy is possible with say something like Mach-3 (that's about the only retrofit I know of that's cheap). Would this be better than an old Fanuc 0m-a control? My current system uses a few older Siemens servo motors for those wondering. Just trying to throw some ideas around.

As always, any advice offered is appreciated. Thx

[dpuch]

I don't think you can get anywhere close to retrofitting the control for $1200 to $1800
Well just maybe out of pocket costs would be that IF you can use the rest of the hardware you already have. But you will have a large changeover and setup cost in manpower. The upside is a lot more versatile control when your done.

You could sell the unused control parts to recoup a little more.

You will probably need a bunch of those settings your missing to setup mach3 as well.

I'd call it a last resort if you can't get the info you need.

[DerHammer]

yeah, that is what I was thinking too. Being that I do make a little money with this piece of iron, the wisest decision would probably be to just get it turning chips again asap vs. turning it into a science project. Especially considering that I can get a used board for $1200, I think this will be the route I go as long as I can get those diagnostic parameters from the machine manufacture.

On a side note though, I don't think it ever hurts to have a plan B. I cant help but get curious on what would be required to complete a retrofit with a system like mine. I already have ball screws, servo motors, limit switches, etc but I have never done one before, so I would have no idea how to determine what is re-usable and what isn't. where would I start to find out?

[dpuch]

In hind sight I should have mentioned my reply was form a production point of view. I figured this was the case, but should have qualified it.

I too am curious, I just never had a machine that it's use was low enough to chance finding out.

I guess the place to ask is on the Mach3 and EMC forums to see if it has been done, and what was actually involved. Ideally it would replace the control, but use all the drives, motors ect. Then sell off the control that was replaced to pay for it

This may be a good time to find old machines with obsolete or non-functional controls for <$5000 and slap a PC control on it to update the machine. Sadly as always the indirect costs are the biggest part of a project like this. Machine transport and your time for the retrofit. Some day I'll get to play around with something like that.

[John_B]

I was given a super old early 80's Fadal about 2-1/2 years ago that had been struck by lightning (the building anyway) and converted it to Mach3 and got some pretty good use out of it. I started out with a finicky control module that couldn't handle spindle sync that slowed me down a lot, but after I gathered my senses and used a worthwhile control card it worked pretty good. Never did get spindle sync working even though the controlled I sold it with supported it, but I used tension/compression tap holders and it was fine.

Mainly it was an automatic drill press for me, but I sold it a while back for $3500. Probably only broke even on that. The guy who has it now does some neat parts and has connected a 4th axis to it as well.

I would not suggest you do this unless you had a similar situation as I had with no possible salvation of the original control - or the functionality would be a major improvement. The project took me almost 9 months on the side, and probably got me a bad rap as a nuisance with a couple of companies.

Regards,
John

[dpuch]

John, About what I expected. What was required to hook up the old to the new?

Did you have to replace the drives, or just bet an I/O board that interfaced with them?
Assuming that what you ended up doing wasn't due to the original lightning damage.

Oh, and congratulation on the conversion.

Ahh, here is a recent post on a EMC retrofit that is underway. http://www.cnczone.com/forums/showthread.php?t=73943

[John_B]

The only electronics that I managed to save were 2 of the 3 axis drive motors, the VFD and spindle motor and a couple of the I/O relay banks.
I did find a match to the original motor for the burnt one so I didn't have an odd one to work around, and wound up scrapping the original I/O stuff anyway.
The main problem was many of the cables were toast as well, so I just pulled every last one out and ran new. You shouldn't have that issue though.
EMC2 may be the way to go for home conversion - a couple of machines I seen since I started mine had a huge amount of functionality - and didn't use Windows!!! Centroid is also a good package, this works real well on Cincinnati machines. Still don't know if I'd do this to a machine that is pretty near running though.
As I recall the Model A of this control was analog, but it still could function pretty smoothly. I have a 0M Model C on an clunker I bought a few months ago, and have been finding spare boards pretty inexpensively. You may try to find another board entirely also to get back to making chips, then get the original repair after the bank account heals a bit and put it on a shelf.
Regards,
John

[DerHammer]

Seems like we all tend to agree on going retrofit vs repair. In fact, the machine is actually a retrofit to start with. Its a 4000lb knee mill, outfitted for cnc in the late 80s made by a company called Webb who luckily is still in business. The particular model is called a "Webmatic 800".

I only paid 5000 bucks for this machine running a couple years ago, so the budget I have to invest is limited before the whole thing gets silly and I should just start looking for another machine. I was torn on it though because nobody wants a 4000lb chunk of iron just sitting around almost running. Its been a good little machine until this recent hick up, so I imagine if I were to fix it, it will have a few more years in it hopefully. After getting this issue resolved, I may just sell it and graduate into something with a tool changer anyways. First thing is first though, I need to get those parameters.

The cheapest replacement board I have been able to find so far is $1200, so if anyone has a lead there, it would be greatly appreciated. I too am under the impression it is an analog system. The machine seems like a sort of "Frankenstein's creation" to me anyways because the controller is Fanuc and the servos/servo drives are Siemens. My plan is to find a used board somewhere, once I get those parameter codes and go from there. If I cant, my options are to retrofit and sell the old parts somewhere like ebay I suppose. As always, Thanks goes out for the advice and links.