What's new
What's new

Fanuc 3m watchdog alarm

camaro_dan67

Cast Iron
Joined
Dec 4, 2009
Location
N.E. Pa.
Here's where it all started. I started my mill up this morning homed it and ran the simple drilling program I was running for the last week. Everything seemed fine until I switched from auto to manual and tried to move the z axis with the mpg. I was handwheeling in the Z- direction and all of the sudden it jumped like it was in a higher increment. The control displayed multiple errors one was the watchdog alarm. One of the others was a vrdy off. I was able to reset it by hitting the estop releasing it then pushing the ready button again. After getting multiple errors while trying to hand jog different axis around I put the machine back in auto and cautiously ran the same program with no parts, it had no trouble. I then had to shut the machine off and come back to it later. When I tried to restart it there was no screen and no motion. I called Fanuc and the guy walked me through a few things to check. He claims its either the power supply or the motherboard. He also had me try to start it up holding down the reset and delete buttons. I tried that and nothing came up. I did check the voltages on the power supply and they were all within .5 of a volt to what they were supposed to be. Does anyone have any suggestions on some other things to try. I don't want to spend the crazy amount thats wanted for the motherboard on this machine only to find out its something else.
Thanks everyone, Dan
 
Checking the voltage does not show what level of noise in on the DC voltages. You need to use an oscilloscope to check the quality of the power.

On an old control like the 3 it would be a good probability that the power supply is noisy due to bad electrolytic capacitors in the unit. Easy to replace and cheap. Last 11M power supply I did that to I think the caps came out to about $10 from Digikey. Shipping was more expensive than the parts.
 
I did replace most if not all the capacitors on the power supply not that long ago because the unit had lost its 24 volt output. I guess there is a chance one of them has gone bad. I have seen new components that were bad even before installing them.
The strange thing to me is that the machine was functioning normally earlier that morning and now the display won't even come up.
 
I have an older techtronics oscilloscope that I can check it with later today. Do you think I should check the power right out of the power supply or try to figure out the voltages on the motherboard and check them there.
 
On the older controls .5 volt +/- is bad.

I know the 5 volts had to be between 5.00 and 5.1 volts for the control to fire up on all the old 6M's I've had.
 
I had a chance to do a few checks on my control tonight. I turned up the voltage on the power supply and measured voltages on the motherboard. There seems to be correct voltage at the labled check points. None of the 3 leds on the board are lit up. I also still have no picture on the screen. Would a watchdog alarm shut all of this down or do I have more problems than just that one alarm.
 
Is there anyone out there that had this alarm in the past? I'm looking for some direction of what to do with it next. I would love to fix the original unit. There are a few used boards on ebay but I'm afraid if I buy a used board that there could be something externally that might damage the replacement.
 
I bought a used motherboard off of ebay. I'm wondering if I should just plug in the 3 connectors at the top of the board and leave the m1 to m19 connectors unplugged to try it. Or should I just plug everything in and try starting it up? Im afraid one of those m1 to m19 connections might have done something to the main board. Anyone have a way hook things up safely so I don't damage the replacement board? Thanks, Dan.
 
Here's an update with this whole mess. I ended up transferring all the socketed chips over to the ebay board. I connected everything up and loaded all my parameters by hand. I restarted the machine and now get a #34 servo alarm. The book says its a disconnected feedback cable. I'm thinking this might be what caused the watchdog alarm in the first place. Any of you guys ever get this alarm?
 
34 on a 3M means that no feedback was detected at power on from the Z axis.

I would not expect that to cause a watchdog alarm but anything’s possible.

Sure the cable is plugged in correctly? Set the jumpers on the board for Z correctly?
 
The watchdog alarm is gone as of now. I have not fully started the machine up because of that error. The plugs on the drives and the motherboard are labled but I will check them again. I talked to a Fanuc tech and he said it could be a problem on the circuit board for the drive or the motor encoder. He also said if that error is new it could be a problem on the motherboard where the z cable plugs in. He suggested swapping feedback cables for the drives to see if the error moves.
 
So, did you set the jumpers on the new board to match the old? Reason I’m pushing on that is because that control used jumpers to configure the encoder count and feedback type.
 
I did match the two but it is possible i missed something. I didn't know about the jumpers for the encoder count. I will have to carefully recheck. Thank you.
 
The jumper settings also determine if a tachogenerator or F/V conversion is used for velocity control. They might also select between encoder versus resolver feedback, but I'd have to revisit some documentation to recall for sure.
 
I still need to check the jumpers you mentioned. If they turn out ok is it possible to check signals at different places? I noticed there is a spot with test pins on the motherboard labled vcmd and separate pins labled x,y and z.
Or would I need to check signals from the z axis encoder?
 
It gets a bit complication to use the pins on the motherboard to kind of verify an encoder. This only works if the control is configured to use the encoder pulse frequency to determine axis velocity by F/V conversion. The test pin TSAZ will show a negative voltage when the axis is moved in a positive direction and a positive voltage when the axis is moved in the negative direction.

It is best to check the encoder signals directly at the connectors. Time to break out the o’scope.
 
I have a scope sitting next to the machine if you can give me places to check I'm up for it. I did call the Fanuc tech again today and he had me swap cables from the drives on the motherboard side of things. The error stayed as the z feedback cable being disconnected. He seemed pretty convinced that the motherboard was the problem. So after doing that test he said to try a different chip on the board. There were 3 of them with x y z marked below them. I tried another one with no change again. This is getting very frustrating.
 
Since the error did not move with the cable swap, it takes away a bad encoder as the problem. I agree with the Fanuc tech that the “new” motherboard is probably bad.

On the jumper block marked Z SET do you have a jumper on pins 1 and 16? That had to be set in all encoder configurations. Then if your board is revision 05E or later you need a jumper between 3 an14 unless the machine has scales. If scale equipped the the jumper needs to be on 2 to 15. Then there will be no or one other jumpers set on that block to define the encoder count used.
 








 
Back
Top