[Camfather]

Details, bought a used machine, running on single phase power via a used (from a different vendor) 20 hp rotary phase converter.

Machine was as is, supposed to work, saw it under power at the shop.

Day 1 - wired everything up, turned machine on and "played" with it for 6 to 8 hours.

Day 2 - system 930 alarm, several times, 6th or 7th try it turned on, "played" with it for 9 or 10 hours, got dnc working.

Day 3, system 930 alarm, appears the machine will start when it's warmer out. I'm in socal, it's cold for us but not so much as cold as it could be. Once machine powers up, appears to function fine.

Day 4 - have a local service guy spend 2 hours, swapped the power supply, same thing, recommended a new main board and or bubble memory card. $1600 to $2200 but not certain that is the problem.......... (it was raining that day)

Day 8 (today) - warm out, powered up with 930 alarm, after 3 tries, machine started fine.... let it idle and cycled power 3 times in an hour. seemed to work as it should each time.

Tonight - Got home @ 11pm, it's cold, machine 930 alarm 10 times, never started up......... now looking for troubling shooting ideas or a repair guy in so cal that can fix it based on facts.

[memoryman]

First: make sure your voltages (+5,+15,-15,+24) are close to spec.
+5,+15,-15 are used on the bmu.
What is your motherboard part#? We may have a spare one.
Buy a bottle of cold spray and selectively spray a small part of the motherboard to localise the problem The outside temperature is likely not an issue.

[tc429]

I agree with memoryman, a noisy/weak supply can cause issues...but also worry about your phase converter- do you have a good solid ground? what is your incoming voltage? the power supply on a Fanuc is built for 200vac, they have a 'multi tap' transformer usually so they can run on about any countries power...you should have two big transformers in the cabinet, a big three pot thing for the servos, and a small single phase one with 100a/b and 200 a/b output terminals...they like to be right around 200, but 190~220 will usually be ok...if youre out of that range, your supply might be struggling a bit- if so you can move the primary tap up or down to lower or raise the secondary voltage.

if the boards are dirty, might not hurt to clean them up- I generally wouldnt touch the bubble (marked BMU) though, Ive had a couple fail after cleaning.

if you need any used 6B/2 parts, pm me- think weve got about 30 of them stashed away that I doubt we'll ever use again...sure the boss would be willing to sell a few of them- theyre all ugly dirty, but I could clean one up/wire for testing no problem...I'm still trying to talk them into reretrofitting some sixes...they were like a old ford pickup to me...dead reliable, great work trucks I had wrote a ladder for the 6 that would allow using the new zero style op panel (long as analog spindle was used- had to use the transistors for s12 bit to run the speed needed for LED sequence), and we have probably a literal ton of old R/C robot mdi/crt panels that were the same as the 6...

[Dan Fritz]

I suggest that when you do get the control to come up ready, the first thing you should do is back up your parameters. You said that your DNC system is working, so get the DNC ready to receive, then put the control into EDIT mode, then press the PARAM key, then press P-9999 then PUNCH. That's letter "P", then a minus sign, then "9999", then PUNCH. The parameters should then be saved on the DNC system. Also, press the PARAM key twice and see if "PC PARAMS" appears on the screen. If it does, write down everything you see on the PC PARAM screen. You might have timer values, counter values, and keep relay values. These are parameters that are NOT SAVED using the usual process. If you don't have a PC PARAMS screen, then you don't have a Fanuc programmable controller.. You didn't say what kind of machine this was, but most machines used the Fanuc programmable controller. Not having these timer, counter, and keep relay values would make it very difficult to restore things if you ever lost your bubble memory board.

Speaking of which .... Your bubble memory board may be failing, causing these alarms. If the control is a "B" model, then you can re-initialize it using the keyboard. If it's an "A" Model, you might have to find someone else to re-initialize it. This is kind of like re-formatting the hard drive on a PC. There may be bad "loops" (sectors) on the bubble board that is causing trouble. The bubble initialize process is outlined in the Fanuc 6-B maintenance manual

You can tell you have a "B" Model by the fact that you have more parameters. If your parameters go up beyond #160 into the 300s, then you have a "B" model. B models also have two plugs under the flapper door where the serial port is. A models only had one plug.

Backup now, while you still can!

[MysticMonkey]

Backup now, while you still can!

!!!!!! What he said !!!!!!!

[Camfather]

thanks guys, I work night shift so I'm just getting home. We were able to back up the parameters, and I think the pc parms? some we couldn't download so we took a photo of the screen. I believe the voltages are in check, but I will double check them myself tomorrow. It's a 6M B, has 2 ports. A16B-1000-0030/05C main board. Machine is an Enshu VMC 330 circa 1985. I did not get any manuals with the machine. Enshu upon paying a $750 registration fee will provide me manuals for the machine "IF" they are still available. Still debating on that option, $750 for manuals could be worth it in the long run IF they are available :-|

I really "think" it's the phase converter, but I have no idea what I am doing other then willing to learn. So I will take the advice you all have gave me and report back if I find out anything. Thanks again!

[Camfather]

New info that I was not aware of until just now. (I have a partner) Holding Reset + Delete, while powering up, it turns on every time.

[fordav11]

I don't know about 6B but on newer controls doing that will clear the SRAM and kill all of the parameters. if you read other posts on this forum you will see many people who were given this advice wrongly, did that and totally screwed their machine. only do that when absolutely necessary if you get alarm 910 or 911 (SRAM PARITY)

[Camfather]

reset + plus delete only worked yesterday. this morning 930 again. I was able to punch out p-9999 and p-9998, only had 1 pc parmeter. voltages 5V=5.01, 24=24.00, +15=14.93 and -15=-15.00 Back to checking out connections.

[Dan Fritz]

It's NOT the phase converter! Your control has a switching type power suppy, and the phase converter will only affect the servos, not the CNC itself.

I just looked up the alarm # 930 (bad news). It says "CPU Error (0, 3, 4, type interrupt generation) Change master PCB"

You may have a bad master board, although it might be something simpler. Try removing the ROM board and make sure that all the EPROM chips are fully seated. Pry each one up a little (without removing it), then press it back into it's socket.

Funny that it should run OK for a while, then fail. look for any ICs that are plugged into sockets on the master PCB and re-seat them as well.

[Camfather]

Thanks Dan, More into it this weekend, my day job is kicking my butt at the moment. Someone suggested I try checking the 5V will powering up to see if it drops. It's weird some days it works fine all day, once it's ON, its good to go until you cycle it completely off, then the next day 930 alarm all day..... That's why I was thinking dirty power. And as I stated I know nothing about what I am doing. lol

To fordav11, I was told reset + delete clears the programs out of memory, and 0 + del deletes all the parameters on my control. Using reset + delete didn't appear to have any lost of data other then the loaded G-code files I had. Again this is on my control, don't know about others. Thanks for looking out for me though.
Disclaimer - Don't try this on your control as it may wipe parameters.

[Dan Fritz]

You're not running with the CNC cabinet doors open, are you? Some people think that heat is causing all their troubles, so they leave the CNC doors open for more ventilation. That's a serious mistake. For proper cooling, make sure that the fans in the CNC cabinet's heat exchangers are running, and that the filters on the outside if the heat exchangers are clean. Then CLOSE THE DOOR!. If the heat exchanger is full of goo, see what you can do to clean it out (this is a difficult and messy job).

The CNC cabinet is an RF shield, which protects the electronics inside from EMI (Electro-Magnetic Interferrence). A bit of EMI from a flourescent light ballast, a spot welder, or even the motor starter on your air compressor motor can create a "pop" that would shut down a CNC control with the kind of alarm that you describe. If you have an old AM radio, tune it between stations and turn up the volume. You'll be able to hear all the "pops" and "hisses" from the EMI sources in your shop.

Another potential problem is the gasket on the CNC cabinet door. Many people replace old worn-out gaskets with automotive gaskets or weather stripping. That's a mistake also. The gaskets on a CNC cabinet door are made with a high carbon content, which makes for a good RF "seal" If the original gaskets are in place, wipe them clean, then close the door. Also, be sure the CNC is grounded to a ground stake through the floor. This is intended to be a LOW IMPEDANCE path to ground, so the wire from the CNC cabinet ground to the ground stake should be a large-diameter stranded wire (not just a thin piece of solid copper wire). It's not just an electrical safety ground, it's a way to attenuate all that EMI that the cabinet picks up. Look on the inside of the cabinet door and you'll see a stranded ground wire attaching the door to the cabinet. Make sure that ground wire is attached. It grounds the door to the cabinet for RF shielding purposes.

Ever wonder why the CNC control is in a different cabinet from the servo drives and all those relays? That's because the high voltage AC stuff in the servo cabinet creates too much EMI for the CNC control. Almost anything that creates a "spark", like a motor starter relay, a welder, an EDM machine, etc. can cause EMI. Also, anything with a coil that turns on and off, like a solonoid coil or a transformer can cause EMI as well.

Unlike a bad board or a mis-set parameter, an EMI problem will occur randomly. You might find that, when the rest of the shop is shut down and quiet, your CNC runs fine, but when other things are running, your CNC shuts down at random. That's almost surely an EMI problem.

[Camfather]

Tried it with the doors open, doors closed, killed all the breakers to the shop and the house (wife got mad) and left only power to the cnc, same results. power supply stays 5.07 during boot under all condition's. Tried having a beer, knocking on the oak tree before powering it up, still 930 alarm. Yesterday I was leaving for an xmas party in an hour, walked over to the machine, it turned right on.......

Yet to try reseating the rom chips, that's next on my list.

Thanks again!

[Camfather]

"you checked the ground? I checked the ground"............weeks later.......

"you guys gonna hook up this green wire over here on the other size of the shop or can I cut it off?"

"wat"

Installed a new 8' grounding rod right in the earth next to the machine, hooked it up, turns on flawsless so far. Cross your fingers. Will report back if System 930 alarm comes back. Now I get to play with my new toy. Merry Christmas to me! Thanks for all your help guys!.

[tc429]

glad you found the culprit...the older 5 series were VERY noise succeptible, the 6 pretty solid, long as its got a good solid ground. often another issue that can point to a grounding issue is intermittent 416/426/436 pulsecoder disconnection alarms whenever a solenoid shuts off... cheaper machines with very thin gage encoder cables are the worst, if the building ground is not great, and if you cant add a ground stake(we have some sitting over basement areas) you can get to where you need to replace the often old/dead rf suppressors on all your ac coils- Ive found the little red 150v MOVs seem to last forever and work great as ac spark killers...
be certain also that if you add any features requiring 24vdc relays or solenoids ALWAYS put a reverse blocking diode across it AS CLOSE TO THE COIL as possible...the sixes Achille's heel is their reed relay I/O card, turning off a dc coil on even a little icecube relay one time without a diode will often arc the reed/weld the contact...strumming a screwdriver handle across the i/o board almost always worked to dislodge them, but still, try not to go there.

interesting little tidbit on fanuc encoders- Ive seen encoder cables 150 feet long work UNSHIELDED. it was a accidental 'someone forgot to connect the shield' thing, but noticed it a year after the fact- the reason it worked: 18 gage wire. if you look in the schematics it says something like '6 conductors or more' for the zero and 5 volt lines- if small gage wires used, you must use a lot of them, as ANY glitch in the 5 or 0 lines will make the A/*A,B/*B, Z/*Z signals strobe briefly...smaller gage on the actual signal leads dont seem to hurt, but we use 18 gage for a long time(till serials came out). if you ever get intermittent disconnects, always look to see the shielding is clamped in the cabinet, and INSULATED at the motor end...seen many times(especially dc servos with their brush dust) where the encoder shield is touching inside the amphenol plug and any 'ground fault leakage' in the servo causes a noisy shield- taping up the shield almost always takes care of it...

[gbowne1]

On my 6MB2 project, I'm still trying to document everything I've done so far, but yes, make sure all your grounds are also not leaking voltage.

Using the right type and size of wire helps as well. Make sure your sheilds are done as well.

The only problem I've had so far with it, is that the FANUC spindle servo motor is not set exactly right. I'm working on setting up the I/O and the 3 Axes amps.

[Camfather]

Well I thought we had it figured out. the grounding rod appeared to cure the problem but only for a few days. Back to the same problem. System 930 Alarm sometimes.....

Looks like I will save up a few more bucks and try a new mainboard. Wish me luck and thanks for all your help so far!!! Much appreciated.

[hrh]

Try to clean the contacts on the large relay for the "input unit". The relay has a clear plastic cover and is mounted upright on a metal bracket next to a spark killer on the "input unit" metal base.

The "input unit" is a small PCB which controls the ON/OFF circuit and the supply of power to the PSU.

If there is excessive sparking on the contacs at power on, a "brown-out" failure may cause momentary (milliseconds) voltage drops and the sequence for the PF and ENB signals cause problems for the CPU to access the EEPROMS.

[Camfather]

Try to clean the contacts on the large relay for the "input unit". The relay has a clear plastic cover and is mounted upright on a metal bracket next to a spark killer on the "input unit" metal base.

The "input unit" is a small PCB which controls the ON/OFF circuit and the supply of power to the PSU.

If there is excessive sparking on the contacs at power on, a "brown-out" failure may cause momentary (milliseconds) voltage drops and the sequence for the PF and ENB signals cause problems for the CPU to access the EEPROMS.

What would be the best way to "try and clean" these? A File, Solvents?

Thanks

[Camfather]

Replacing the large relay on the input unit ($18.00) seems to have fixed the the System 930 alarms, Have not had one in weeks. Thanks hrh