Printable View
Hello All
I am working on a series 5 Fanuc, with Fanuc model 5 and 10 servos.I have checked the +5V precision voltage on the power supply and adjusted it to +5.026V or so,it was +5.126V. The +6V and-6V reference voltages are 6.148 and 6.030 respectively. Do these need adjusting and if so is the RV3 pot the correct method? I have moved it back and forth and no change. Also is there any other voltage checks/adjustments I could complete to eliminate a run/drive away problem? Regards Craig
Craig, if you're haveing a servo runaway problem, it's not likely to be caused by a minor power supply adjustment.
Here's a common cause of servo runaways on the 5T and 5M controllers:
On the edge of the power supply, you'll see a bunch of grey cables with Molex style connectors. These cables go to the A board, the B board, and the Memory & Edit board. Look carefully at the connectors at BOTH ends of these cables. If you see any discoloration, you may have a connector that's overheating. The 5v pins to the B board are the most likely to overheat, because the B board uses the most current.
On early model Fanuc 5 controls, the pins that mate with these connectors on the power supply were not gold plated, and would overheat and corrode over time. The later power supplies had gold plated pins, and were not as vulnerable to overheating.
Here's what happens: When the B board looses 5 volt power for any reason, the pulse coder feedback from the servos is affected. The LEDs in the pulse coders go out, and (ordinarilly) the B board would shut down the control with a pulse coder feedback alarm. Unfortunately, if the B board loses power, it can't trigger this alarm either, and the servos will just "drift" away. If the power is lost while the servo is moving, the servo will continue to move at the same speed until it overtravels. Hopefully, your E-stop circuit will stop the servo before any damage is done to the ballscrews.
If you see any discoloration on the Molex connectors, you can replace the connector or, if you don't mind a hokey solution, you can just solder the wires to the power supply pins.
There are other reasons for servo runaways, but on the Fanuc 5, this connector problem will let them run away without triggering any alarms.
Thanks Dan,
I checked all the molex conectors to the B board and the power unit, all are like new. Checked all voltages at the power unit against the printed value next to each of the pins where available, and all are within 0.2V. Traced these to their destination and checked voltages all OK. The resistance of these interconnecting cables all 0.23-0.25 Ohms. I am still concerned about the inability to adjust the +6V reference with RV3 on the A board what do you think? Any other help would be greatly appreciated.
Thanks Craig
Hi Craig,
The +6R and -6R are just reference voltages, and they are just regulated down from the +15 and -15 voltages with a pair of zener diodes near the PC07 chip on the A board. The RV3 pot adjusts the +5R voltage only. My schematics don't show any +6 or -6 voltages on the power supply itself.
These "reference" voltages do not need to be exact. They only need to be constant, which is why they use a simple zener diode circuit for regulation.
There is a pot to adjust the +5R on the A board (RV3) but there is no pot to adjust it on the Fanuc 5 power supply. That too is regulated by a zener diode ZD3.
If your servo is running away, I suggest that you pull the fuses on that servo and test the VCMD signal coming from the A board. The signal to the X axis has a checkpin marked VCMDL, and the Z axis on a lathe is marked VCMDM. If this is a mill, then X is VCMDL, Y is VCMDM, and Z is VCMDN.
The voltage you see on these checkpins should be exactly proportional to the servo's following error. If you pull the fuses on the servo and turn the control on, you should be able to "dial in" any amount of servo error by using the step-jog buttons or the handle pulse generator. A few thousanths of error in one direction should generate a + voltage on the VCMDx pin, and a bit of error in the other direction should result in a - voltage. A servo error of zero should result in zero volts (adjustable to zero with pots RV1 L, M, and N on the A board). If you turn on the control and see a big voltage on the VCMD pins, then you've got a problem on the A board.
NOTE: Be careful when pulling the fuses on a Z axis on a mill, or on a slant-bed lathe on the X axis. You don't want the slide to drift down under it's own weight and introduce servo error. Some machines use an electric brake in the motor, and turning on the servo with the fuses pulled would release the brake, but the motor won't hold the ballscrew from turning.
If your VCMD signal looks normal, then you might want to test the Tach signal also. The tach signal on most Fanuc 5s is generated by the PC06 hybrid IC on the A board. The pulse coder signals are turned into an analog tach signal by the PC06, which is basically a frequency-to-velocity converter. The faster the pulse coder frequency, the higher the voltage from the PC06. The tach signal is visible on check pins SCDVL, M, and N, and it should be zero volts when the motor is still. If you have the fuses pulled, you should be able to rotate the motor a little back & forth, and see a small analog signal on the SCDV check pins. The voltage won't be much at such a slow motor speed, but you can at least see if the signal is there.
NOTE: you will only be able to turn the motor a short distance before the control alarms out with an "excess error" alarm. This is normal. Just rotate it back & forth a short distance and you'll be OK. A zero volt signal when the motor is stationery would be normal. The offset for setting it to zero is RV2L, M, and N on the A board.
I've seen many cases where the square black hybird ICs fail and cause servo problems. A bad PC06 or PC03 can make the servo run away, but it will only go a short distance before the servo will shut down with an "Excess error" alarm. The control sees the pulse coders turning, and when it counts up to a value preset by parameter, the control should alarm out and shut of the servo for safety.
Good luck!
Dan Fritz
Dan I will check all this out today.The lathe only travels a short distance of about 3/16 of an inch before tripping out. If the control trips on "excess error" alarm how is this indicated? I don't have a list of the DGN codes, are these fault codes or something else?
Thank you once again, Craig
WE had essentially the same issue with our 5T.
After spending a small fortune on board resoldering (it helped) and harness rework/repairs (poor maintenance by prior owner that helped LOADS MORE), we ultimately found some poor/broken connections that would go intermittant out on the iron.
Point is, all your "issues" may ultimatetly NOT be in the NC cabinet - ours were there (cobbled wires) and on the iron (intermittants due to too many years of flexing).
I do hope that you have a machine schematic. What could take days could be done in hours if you have schematics, especially with regard to circuit tracing.
We have some lathe schematics and OEM manuals that we could copy and send for copying & shipping costs. These "generic helpers" proved invaluable to us in our troubleshooting efforts.
The lathe schematic clearly shows the pinouts which should be close enough to get you bye even if you machine is vastly different than ours.
I measured the voltage on VCMDL pin and found it to be 13.025V with the fuses pulled. The VCMDM pin voltage was a steady -0.0566V. The tach signal is there for both axis. Question which PC 0? hybrids do I replace to hopefully solve this issue?
Once again thanks for all the information, Craig
There are only a few ICs that can cause the VCMD signal to go off the charts like this. The L (x) axis circuits are all in a horizontal row on the top-left area of the A board. The row of circuits for the M and the N axes are below them.
The VCMDL signal comes directly from the 8-pin 741 Op-amp at coordinates I-8 on the A board. If that op-amp is blown, that could cause it. These op-amps are available from Radio Shack, but they seldom go bad.
The 741 op-amp gets it's input signal from the square black A-PCO3 hybrid IC. I've seen a lot of these go bad. The only source for these is Fanuc, because this is a Fanuc-made IC. Any Fanuc service tech should be able to get them.
The A-PC03 gets its input from the A-PC02, and the PC02 gets its input from the A-PC01. These are also Fanuc-made ICs
My bet is on the A-PC03 at location I-10
Thank you both for your help, I have purchased some APC02,APC03 and APC06 hybrid chips off Ebay. Fanuc in Australia don't want to know about this age of control. They were around US$15-25 each which I found OK. I have sourced a EX Fanuc tech to install them when they arrive. Iwill let you all know how I get on. Once again thanks for the fast answers to my questions.
Regards Craig
Dan Fritz and NC cams,
Thanks for your information. I had these PC 02, 03, 06 chips replaced on both axis. This along with a replaced PC 01, and both amps, fixed my motor control problems. I now have a P/S 79 alarm, which can be reset but after the buffer is read, it alarms again. I think, reading back through prior posts, it may be a memory board problem. Do either of you know of a fix?
regards Craig.
The 79 code is definitely a memory board problem.
There should also be a two digit code in front of this (example:3079). The first 2 digits tell you at what position on the memory board where the problem is located. Consult your owner's manual for a map as to which memory IC's are being reported as problematic.
Typically, the first thing you do is CAREFULLY pull the memory chips and clean the pins with a pencil eraster. THen CAREFULLY reinsert and reseat the IC's. You'll also want to check, clean and tighten the gray connector contacts that provide power to the left side of the board.
You WILL need to totally clear the memory stack once you have a 79 error code and when you're repowering the system after servicing it. You might want to replace the back-up battery pack t this time as this keeps the memory alive when you power down the machine. You'll need to clear the mem stack if you ever lose power to the board.
I'd install new back-up batteries, clear the memory stack, fix/tighten the gray wire power connectors, reset the IC's, reinstall all the paramters PROPERLY (NOTE: this tells the machine how much memory you have and at what positions) and this should eliminate the xx79 error message.
Hopefully, the memory IC's have not gone bad as they are VERY difficult to find and VERY expensive due to the age and rarity - definitely NOT Radio Shack or DigiKey items. IF you do have to replace a memory IC, you MUST replace it with one preferrably of the same make and DEFINITELY one having the same speed rating. Should one memory IC need replacing, it would probably be better to replace all of them and use matching brand and speed IC's.
Finally, be real careful plugging in the ribbon cable interconnect cable that goes to the B board. It is real easy to reverse the leads and this gets VERY, VERY expensive to repair.
Mark the connectors for position/polarity BEFORE you remove them - do NOT simply rely on your memory to recall how the plugs need to go. I speak from experience, very expen$ive experience on this issue.
Hi, Paul Here ,,,,On my 5T the Rv3 pot on the upper "a" board will only adjust the +5R voltage,,the lug is located just above and to the right of the Rv03 icQuote:
Originally Posted by wassa64
also the Lug for checking the +5 volts is below and on the right side of the PC07 ic ,,,,the pot on the power supply is used to adjust this.
dont mean to butt in Just helping..