Fanuc 11 series question

I'm working on a VMC with an 11M at a small start-up shop. It has a RAM error on boot up. Initializing the control does not clear it. My documentation shows RAM is on the main board and on the ROM/RAM board. I have swapped the ROM/RAM board from another control and the problem stays the same so pretty well convinced it has to be the RAM on the main board. There is no obvious RAM chip(s) on the main board. I suspect that the RAM is in one or both of the Fanuc LSI chips on the board. Anyone out there have info that may answer that?

All the typical pre-requisite troubleshooting has already been done. Power supply scoped, re-seating socketed chips and connections, etc.

Wanted to update this thread.

I was not able to find out if the mainboard RAM was part of one or both of the LSI chips. Fanuc understandably was not willing to put much effort into digging up information for a 30+ year old control. They did have tested exchange boards ready to ship if needed. Price of course is quite high and likely more than the true value of the machine.

Machine owner sourced a reasonably priced replacement main board from a machine being scrapped. I installed it and all is good again.

Glad to hear.

Generally, my opinion is that if the machine is still viable, it's almost always "worth it" to fix it, even if 30+ years old. I have 30 year old machines in my shop that still make good parts, so it's "worth it" for me.

Good to hear, and thanks for following up

Vancbiker said:

Wanted to update this thread.

I was not able to find out if the mainboard RAM was part of one or both of the LSI chips. Fanuc understandably was not willing to put much effort into digging up information for a 30+ year old control. They did have tested exchange boards ready to ship if needed. Price of course is quite high and likely more than the true value of the machine.

Machine owner sourced a reasonably priced replacement main board from a machine being scrapped. I installed it and all is good again.

Click to expand...

Do you have a photo of the "Junked board " ?

I could take a stab at that... Been going through the FANUG 10/100 series, FANUC 11 /11O series FANUC 12/120 series... (Very interesting to see how that has all evolved especially the transition from 11 series to 12 series.).

(Just for future reference ?).

Later on those LSI chips are more geared towards servo control (on main board and additional axis cards.).

In the sections on the FANUC 11 series documentation there are a couple of "cryptic" diagrams that seem to indicate there could indeed be a separate memory array (not integrated with LSI chips.).

_________________________________________________________________________________________________


It is interesting that FANUC's approach is that if there's anything "squiffy " RAM or ROM or RAM/ROM (separate board) the manual advises "replace the whole board " kind of thing but I wonder if that's partially in the interest of safety/ product reliability but also provides FANUC the "Opportunity" to supply an improved board ;-) ?


@Vancbiker ---> Just kinda interested in your thoughts on that … If there is a board that is prone to memory failures let's say over a ten year period does FANUC engineer better replacement boards … OR is it just the same board more or less and will probably fail in the same way again ?

_______________________________________________________

Roughly speaking how much would a replacement master PCB / board be 11 series (do you reckon) ?

Ta...

cameraman said:

Do you have a photo of the "Junked board " ?

Click to expand...

No, but I can probably get one in the near future. I've been helping this place with some setup training and macro development to speed and ease setups so I'll be out there again sometime fairly soon.

cameraman said:

I could take a stab at that... Been going through the FANUG 10/100 series, FANUC 11 /11O series FANUC 12/120 series... (Very interesting to see how that has all evolved especially the transition from 11 series to 12 series.).

(Just for future reference ?).

Later on those LSI chips are more geared towards servo control (on main board and additional axis cards.).

In the sections on the FANUC 11 series documentation there are a couple of "cryptic" diagrams that seem to indicate there could indeed be a separate memory array (not integrated with LSI chips.).

Click to expand...

I'd be interested in what you might find. I did not check all the chip numbers while trying to identify what may be RAM. The most likely culprit based on package size and manufacturer turns out to an A/D converter. Part of axis and spindle control. The board in question is the analog version of 11 series. Later in the production of the 11 Fanuc released a digital version main board.

cameraman said:

It is interesting that FANUC's approach is that if there's anything "squiffy " RAM or ROM or RAM/ROM (separate board) the manual advises "replace the whole board " kind of thing but I wonder if that's partially in the interest of safety/ product reliability but also provides FANUC the "Opportunity" to supply an improved board ;-) ?

Click to expand...

Well it certainly provides Fanuc the market to sell new or repaired boards. It also reduces their cost in having staff try to help customers with repairs at a lower level. The range of knowledge and ability of folks trying their own repairs goes from "tell me what a multimeter is?" to "is the mainboard RAM a component of your proprietary LSI?". Far easier to just work at a macro level and replace the entire board. For some users reducing the downtime is far more valuable than saving a couple thousand dollars by taking time to troubleshoot and repair to a component level.

cameraman said:

@Vancbiker ---> Just kinda interested in your thoughts on that … If there is a board that is prone to memory failures let's say over a ten year period does FANUC engineer better replacement boards … OR is it just the same board more or less and will probably fail in the same way again ?

Click to expand...

10 year failure of any Fanuc board is very rare IME. Board revisions typically occur when components are NLA and a sub must be incorporated or when production/reliability enhancements are realized. An example, the bad board was a version 12C and the replacement was a version 19C. The only apparent change was a couple of resistor modules of a slightly different form factor and the LSIs were direct soldered, instead of socketed as the older version was. I can think of one case where an entire replacement board was engineered for out of production series. When Hitachi stopped making bubble memory as used on the 6, 11, and 12 series controls, Fanuc designed and built a SRAM based replacement for the bubble memory board so that customers could continue using machines with old controls.

cameraman said:

Roughly speaking how much would a replacement master PCB / board be 11 series (do you reckon) ?

Ta...

Click to expand...

An exchange repaired board from Fanuc is about $4k with a core refund of about $1500 on return of a repairable replacement. From TIE, an exchange repaired board is ~$2500 and the core refund is about $500. The board from a scrapped "known working machine" was about $500. Of course there is a gamble that the board will actually be good.

Vancbiker said:

<snip>

I'd be interested in what you might find. I did not check all the chip numbers while trying to identify what may be RAM. The most likely culprit based on package size and manufacturer turns out to an A/D converter. Part of axis and spindle control. The board in question is the analog version of 11 series. Later in the production of the 11 Fanuc released a digital version main board.

<snip>

Click to expand...

The Fanuc architecture is very interesting in terms of failure vs. self detection and the different strategies they have employed that even if various critical display and various CPU boards fritz out the rest of the machine still has the capability to "communicate" what's wrong with it by at least one or two or three different methods. (In line with what you are saying about practical commercial service environment, time vs money vs training for diagnostics.).

Also interesting how the error codes / diagnostics can go from being very specific to in later generations being much more broad and generic fitted or tailored to the most likely "actionable " "Next steps" within a FANUC service ecosystem.


So I wandering how the system of self diagnostics and error codes would delineate a specific fault with A/D converter versus something else.. Will dig deeper /scratch around on that. (time willing).

That's a useful comment about the 11 series analog versus digital … Was wondering whether those ran concurrently or whether there was preference for analog for different types of machine / work ?


Good to know that in this case that you are dealing with the analog version of the series 11 (FANUC) … narrows things down by a 1/3rd to a 1/6 th.

Vancbiker said:

<snip>

10 year failure of any Fanuc board is very rare IME. Board revisions typically occur when components are NLA and a sub must be incorporated or when production/reliability enhancements are realized. An example, the bad board was a version 12C and the replacement was a version 19C. The only apparent change was a couple of resistor modules of a slightly different form factor and the LSIs were direct soldered, instead of socketed as the older version was. I can think of one case where an entire replacement board was engineered for out of production series. When Hitachi stopped making bubble memory as used on the 6, 11, and 12 series controls, Fanuc designed and built a SRAM based replacement for the bubble memory board so that customers could continue using machines with old controls.

Click to expand...

That's super !

There seems to be a lot of GOOD up front design on the FANUC control (mission critical well engineered system by design/ architecture. )


I don't want to give off the impression I'm "bagging" on the internal tech of early 80's to early 90's FANUC controls...


There are some very good reasons / mission critical for the way they are at that time*. And was also wondering if there are more modern digital interfaces to the older style sub systems like 'Bubble memory" ---> to SRAM compatible replacement (as you mention); that's very helpful / useful.

___________________________________________________________________________________________________________


* During the cold war just when the West was going gangbusters with digital electronics and sophisticated ICs (integrated circuits / chips.) a Russian Mig went down and was captured in JAPAN (need to look up the exact year). But what they found was that the Mig was using apparently archaic electronics and even a lot of vacuum tubes. It was later discovered that if the Russians explode a nuclear warhead fairly high above the ground / in the air it creates a massive EM pulse (Electromagnetic pulse) that basically destroys and wipes out the DATA / coded instructions on ICs. The Russian modern "retro" tech was immune to such EM bursts. So having a Fanuc control with a paper tape reader + bubble memory which is much more resistant to EM pulses, means you can still cut parts (like at GE etc.) for defense / weapon systems … Still make aircraft/ aerospace parts. [With redundant systems that can still 'Dooooo" stuff.] Interestingly the Russians carried on manufacturing bubble memory units... (probably for EM resistance / nuclear "Combat".). Since then the Berlin wall came down (and the Russians don't have contingencies for nuclear war) maybe not such a belt and braces and stapler, super glue and scotch tape system is needed for mission critical industrial systems. More modern yet EM vulnerable control systems would seem / be "O.K." after the 2000's.


Today with industry 4.0 "Foreign actors" would just hack a system from thousands of miles away... Paper tape is not so easy to hack remotely, needs someone to scrumple it up in the bin, set fire to it or poke holes in the wrong places in the tape (lol) ;-) And that would only temporarily impede things.

IME, a viable main board for just about any old Fanuc is $400-$600 and usually worth it to solve these kinds of problems.

I have a noticed a big uptick recently in shops that have run 80's machines forever finally deciding to scrap them out and upgrade. Probably due to lack of OEM support and techs that can keep them going.

Garwood said:

IME, a viable main board for just about any old Fanuc is $400-$600 and usually worth it to solve these kinds of problems.

I have a noticed a big uptick recently in shops that have run 80's machines forever finally deciding to scrap them out and upgrade. Probably due to lack of OEM support and techs that can keep them going.

Click to expand...

Yep, Availability of old Fanuc stuff is really good. At the same time they were getting the main board, they picked up a bubble board that gives them 256k of program memory for ~$80. That board would have been over $2k 25 years ago. Nice upgrade from the original 32k.