5t problems

[markjb]

we started are machine run a dozen parts. it shut down restarted and we get a alarm 3079 if we try to curse across the program or call up a line. if you start it up and let it sit it will shut down and restart. we dumped the memory and reentered the program now it is working. what would couse this it happens every couple weeks.

[markjb]

ran good for 5 hrs. shut down with error 3079

[NC Cams]

When/if it shuts down, look to see if any of the LED's on the A or B board or Memory board are lit. If so, advise as this is critical to troubleshooting any electronic gremlin that may exist on the 5T.

If you're getting a P/S Alarm, the following codes apply:

10 inadmissable G code
20 End point of circular interpolation is out of range
30 Tool offset value too big
31 T code in circular arc interpolation
60 no sequence number to be searched
61 Problems where G70, 71 72 and 73 are commanded in the address or some other screw up.
63 Another error function in the G70 thur 73 codes in association with error in P and Q
62 Big problem too complicated to type in here.
64 Tape mode problem with commands being input
65 more G70 thru 73 problems
68 more G70 thru 73 G code problems
70 amount of input exceeds capacity of tape memory
71 No address in memorized data which is designated in Edit mode.
78 no sequence number by P in memorized data for M98 and M99
79 tape memory parity error - (faulty CMOS)

If the alarm is setting off the following addresses (3079) you may specifically have an electronics issue:

TH code 3
TY code 0
P/S code 79

The above means that memory chips at position 30 are having an electrical problem. This could be due to aged/decayed solders (readily fixable) or (god forbid) bad CMOS IC's. These CMOS IC's are very hard to find and very expensive if you can find them. A simply pull and clean of the pins and reseating might do wonders.

Another thing is that the backup batteries are going south and you're seeing a momentary power drop on the memory board. Again, a complete resolder of the memory board will fix this along with a back up battery replacment. Be sure to completely erase the memory when/if you replace the batteries.

Also, it may not hurt to totally clean and reseat all the board to board connnectors. After all, this controller is now pushing 30 years of age and the fact that it still runs is somewhat amazing. Connectors on these old machines can EASILY go intermittant and this can cause all kinds of goofy ass intermittant problems - especially in light of the fact that Fanuc was too cheap to use gold plated connectors instead of the lame/cheap tin plated ones they did use.

Scared to resolder the boards yourself? Call Glenn Rowe at Billor Ent (888-4214-5567) as he specializes in resoldering and bench servicing/testing 5T and M circuit boards.

The B51695E/04 maintenance manual is something you may want to review as this manual gives troubleshooting helpf for this and other 5T or 5M problems.

[Dan Fritz]

It is a memory parity alarm on the memory & edit board, and it can be caused by a poorly seated CMOS IC or a low power condition on the board. The memory & edit board is that 1/2 size board that mounts on top of the lower "B" board. It's mounted in such a way that you can remove 2 screws on the top corners and swing it down to permit access to the B board.

The Memory & edit board has a battery pack for power backup, which is about the size of 3 "C" cells. (In fact, you can make your own battery pack with 3 C or D cells linked together).

The CNCs power supply provides 5VDC power through a set of connectors on the edge of the board. Look to see if those connectors (on both ends) are clean and not discolored from overheating. Cleaning those power supply connectors is sometimes needed.

If the 3079 alarm can't be cleared by erasing the part program memory, then I would suspect a bad CMOS IC. Since you can clear the alarm, look for a power problem or re-seat the ICs in their sockets.

You can totally eliminate the use of the memory & edit board by connecting a BTR (Behind the Tape Reader) interface. This lets you run off a nearby PC in TAPE mode instead of using the Fanuc part program memory. We sell a BTR link that connects the parallel printer port of a PC directly to the Fanuc's tape reader input on the A board. Other vendors sell BTR boards for the 5 control that have serial ports. For less than the cost of a new memory & edit board, you can go the BTR route and have a much more powerful control.

[erd39030]

Hi
I have the same problem on my Hitachi Seiki 3NE-300 lathe.
I have the 3079 alarm when I go to EDIT mode.
If I unplug the batteries then the alarm goes off, and then I get the 70 alarm.
So I clear the program memory with C+ERASE... but it gets to the start: alarm 3079.

Any ideas?
I will start to clean boards and reseat chips to see if it solves the problem (I hope that I can do it without damaging any CMOS).

Dan, how much will it cost the BTR solution that you sell?

Thank you: Everardo Rascon

[Dan Fritz]

Hi Everardo,

Our BTR solution is sold on the West coast by Refresh Your Memory, Inc. in San Jose California. The guy to talk to there is John Hosmon, (408) 224-9167 or john@rym.com

The BTR kits include the special cable and software for a standard Windows based PC. It works with all versions of Windows from 95 to XP. With a BTR connection you can forget about the Fanuc's memory, because you'll never need to use it. The control will run directly off the PC in TAPE mode, and you can see your data scroll by on the PC as the program executes. You can also run programs with decimal points, call subroutines, and edit files while the machine is running.

I'm pretty sure that RYM currently lists the kit at $1195. You may want to contact John for a quote. I do know that RYM will send you a kit on a 30-day trial basis, and if you don't like it you can send it back with the shipping being your only cost.

[NC Cams]

Prior posts 3 and 4 give a lot of insight into where to look - namely, power feed to the board.

The wiring in the machines is nearly 30 years old. Ditto that for some of the adjustment settings on the power supply.

IF the connectors are loose or corroded - easily done in a 30 year old machine - you may NOT be getting the necessary voltage and/or current to the memory board. Time spent cleaning, retensioning and reseating the power connectors is time well spent. YOu should see 5.00 vdc min with 5.1 tolerable.

The P/S has a number of voltage adjustments - some go to the A and B and some go directly to the memory board(s). First, clean/reseat ALL the connnectors. Then, using the factory adjustment procedure, check and adjust the P/S bus voltages via the pots on the power supply - there are several so check them all. Follow the factory prescribed method and test points.

Keep in mind that the A board also has a 5V voltage adjustment on it - it too should be checked and adjusted with the same GOOD voltage meter used to check the P/S.

Although the manual says to adjust and check the power on/at the P/S, it does NOT hurt to see what the 5v buss voltage is on/at the A, B and memory boards after setting/resetting the P/S buss voltages. Any flakey/irregular power is a nightmare with ANY 5v TTL logic and the 5T is pretty much ALL 5vTTL logic.

Your 3079 error is suggestive of a power problem - you now should know how to at least check to see if you have a P/S or wiring problem.

Hopefully, by taking care of the power, you will fix your issues. Hopefully, you will not have a CMOS problem (see post 3 for diagnosis method - easily done via noting any led's that light up on the various boards - advise if any do as these are important).

If, God forbid, the CMOS memory chips have gone bad, finding replacements will NOT be easy. Unless you can find some and find some that are do not deplete your life savings, your only solution will probably be to go to a BTR like Dan is offering and bypass the memory board completely.

[NC Cams]

The B51695E/04 maintenance manual may be of some help as it provides adjustment info for the P/S and A board buss voltages.

They are NOT easy to find - Fanuc doesn't offer them anymore and the ones on the internet are QUITE pricey. I have a pristine manual that can be copied and supplied for the copying expense and shipping and handling.

Payment in advance is requested - I sent copies out and didn't get paid (even got stiffed via UPS COD) so, sadly, I need to be more judicious in who I put on "open account" status.....

[erd39030]

Thank you guys,
I will call RYM today.

Have a nice day

[erd39030]

Hello!
Back in troubles again

When I use M3 or M4 the spindle starts but never send back the FIN signal, so the program hangs in that line. Any suggestion where should I start looking?

I couldn't fix the memory card (well the 3079 alarm is gone, but now I have 2 leds - D, OFX - ON, and the display 0111010 - OH, OT, SV, TH, TV, P/S - , no alarm). I'll stop to messing with it before I burn the whole electronics or do something that make me cry.

Dan's software seems to work pretty well, so I'll buy it as soon as I get some cash (my credit cards skyrocketered beyond Jupiter), along with some 5T manuals (I have none).

By the way, you guys know a lot of the Apollo era CNC machines (like mines). Have you written any book? If not, you should. If yes, please tell me the titles.

See ya!
Everardo

[Dan Fritz]

Hi Everardo.

What kind of spindle drive do you have (DC motor with continuously variable speed, or simple AC motor with a set of gears in the headstock)?

Most of those old 5T equipped machines had Fanuc DC spindle drives. Every time the spindle is turned on or off with the M03/M04, or whenever the speed is changed with an S-code, the "Spindle speed arrival" signal is supposed to notify the control that the spindle has reached it's commanded speed. That usually generates the "Fin" signal that lets the program continue. If the spindle DOES come on and wind up to speed, but the control hangs up with the cycle start light on, it may be a missing Speed Arrival signal. If the spindle doesn't come on, that's another matter alltogether.

The "Speed arrival" signal is generated by a small IC reed relay on the spindle drive board. These IC relays sometimes go bad and start to "stick" or "time delay" (click on several seconds late). Try this:

Get the control to "hang up" on an M03 or M04, then go around to the spindle drive board and look for the few IC reed relays that are on the board. You can tell the difference between an IC relay and a regular IC by the "Gap" in the pins. A regular IC has either 14 or 16 pins, evenly spaced along each edge. An IC relay has only 4 pins on each edge (two on each corner, with a "gap" in the middle). Give each relay a light "tap" with something non-conductive (like a small screwdriver handle) and see if the control suddenly takes off. If it does, you've found the culprit.

Also, if you post the number on the spindle drive board, maybe I have a schematic for it and I can tell you which relay is for the Speed Arrival signal. There may also be some "pots' on the board that can affect this signal. Usually, there is a pot that adjusts the "tollerance" range for the Speed Arrival signal. When the speed gets within, say 10 RPM of the commanded speed, the relay comes on. A mis-set pot can prevent the signal from coming on at all. To find the correct pot, you'll need to have a schematic.

The IC relays are easy to replace, but it takes some skill to de-solder the old ones from the Fanuc 4-layer boards. The relays themselves are pretty cheap.

[erd39030]

Thank's for your help, Dan,

I have a DC Driver, FSD-15 DN21 (Fuji Electric).

I tried knocking some relays but the bad one didn't show up.

I noted that it doesn't hangs when changing the speed, only hangs after M3/M4 (and the cycle start light remains ON).
In fact I start the spindle manually and then I run the whole program without M3, using different speeds (G97 & G96) and it works fine. Perhaps it is not related to the Spindle Arrival signal.