[awemawson]

Is there a way of running buble memory tests on a Fanuc 6M?

I'm trying to install a larger bubble memory board in my Fanuc 6M: a few numbers:

Main board is: A16B-1000-0030 / 05C (is this a 6M B2 ???)
(System come up saying its a 6M Series M47 version 10)

Original BMU is: A87L-0001-0016 / 09H (40 metre 16Kb?)

Repacement BMU is: A87L-0001-0084 / 06C (320 metre 128Kb?)

I've installed the bigger card, loaded the defect tables for each device, and been able to re-load the parameters from a file generated using the original bmu. The machine basically works (it's a four axis wire edm machine), but I get one or two oddities. For instance if I do a ref return in X&Y all is well, but U&V can give servo errors or over travel alarms. If I revert to the original BMU it behaves itself.

Admittedly I've not checked every bit of every parameter, but I've spot checked all over the place and found no errors. As well as making a file of the parameters I also photographed each screen on my mobile, which has proved very useful for checking!

So is there a way of running a bubble memory test? I assume that the RAM test on the IL screen is the ram on the main board being tested not the BMU?

[fordav11]

did you initialize the new BMU?
The procedure is shown here (read the entire thread)
http://www.cnczone.com/forums/fanuc/...tem_error.html

[awemawson]

did you initialize the new BMU?
The procedure is shown here (read the entire thread)
http://www.cnczone.com/forums/fanuc/...tem_error.html

Yes I followed this procedure:

http://fanuc.pennineuk.com/library/pdf/FANUCBUBBLE.pdf

[tc429]

that is a odd problem, sure sounds like a parameter typo though...suggest downloading/reloading rather than typing them in if thats what was done.

easiest fix might be get a RAM board from Tulip, their card plugs right in, you piggyback your BMU onto theirs, cycle power and it transfers everything in one step- really nice setup. As a aside, I see Fanuc is now making (unless rebadging tulip) SRAM cards that replace BMUs the same way, we got one in a little while ago on a old mchine we were rebuilding...

[awemawson]

No budget for that! This is a hobby machine in my garage at home.

[awemawson]

OK panic over! I re-initialised the BMU, re-entered the faulty loop data, re-read my parameter file back into the BMU along with my pich error corrections - in other words as far as I know I exactly repeated what I did before, and now she works. Not tried every function, but she's certainly doing what she wouldn't do before - very odd! I suppose it must have been a flaky file read.

But the question remains, can you run a test on the bubble memory in a Fanuc 6M?

[memoryman]

As the founder, owner, designer, janitor, shipper etc. of Tulip Electronics Inc., I can state with certainty that Fanuc's SRAM replacement board for the BMU is not ours. They came out with their board in the early 90s, with a 512k limit.
There is no BMU test built into the software of any control.

[awemawson]

Presumably the bubble chain includes a parity bit that is checked by the hardware and would flag errors to the controller?

[tc429]

As the founder, owner, designer, janitor, shipper etc. of Tulip Electronics Inc., I can state with certainty that Fanuc's SRAM replacement board for the BMU is not ours. They came out with their board in the early 90s, with a 512k limit.
There is no BMU test built into the software of any control.

Cool- good job on those boards

sorry the fanuc one I saw I *assumed* was newer/a copy of your idea or something.

as you are probably more familiar with bubble than anyone out there, have a off topic question- any idea if a Fanuc/Karel R-H BMU can be 'defragged' or condensed like on the old 6B-2? te Karels get slower and slower to boot after any edits, reloads quickens them back up, but dont last long...

[tc429]

No budget for that! This is a hobby machine in my garage at home.

I can relate- got a Fanuc controlled bridgeport here to play with

[memoryman]

The Karel BMUs are the same as the 6,9,11,12 ones. The condense function is a software matter; I don't have much experience with the robots, just enough to make our boards work with the robots.