Wanted ROM dump and info on Fluke 8520A with option 010

[gmac34]

Hi, I have recently repaired Fluke 8520A, my unit does not have 010, and would be nice to add it, the service manual doesn't say how to convert it, but it seems I need to double the onboard ram (should be doable, sockets seems to be there) and probably replace the 3 ROMs, if anyone with this option sia able to take some pictures of its board and dump the ROM I would be grateful.
Thanks!

[picburner]

Have you checked to see if the eprom on the ko4bb site have option 010?

[gmac34]

Unfortunately there is no note attached. I'm not 100% sure the rom is different. I need to read mine out and compare

[gmac34]

a picture of the ROMs, would also be useful, just to compare part numbers.

[wn1fju]

I can provide a picture of the ROMs, but I don't have an easy way to dump the contents.  My 8520A is marked Option 010 on the back panel and I did power the unit up and verify that I can select the expanded set of math functions.

In case the attached photo is hard to read, the chips are (top to bottom):

U12:  MK36324N
U11:  MK36322N
U10:  MK36321P

Also, on my unit, U13 is not populated, and U14 has a socket, but no IC installed.

[m k]

ko4bb U11 12b8 has displayed numbers, inverted 7-segment.
When bit is off the segment is on, 03 is zero, 9f is one and so on.

If -010 machine is displaying something special you can check if ROM images are including it.

[gmac34]

Thanks
I will compare with mine in the weekends, stupidly I left home without taking a picture myself.

The service manual is quite confusing I believe there are at least two major revisions of the digital board, one with 4 ram ic U4-U7 (only 2 populated without option 010) and the other with a single 2K × 8  in the U13 position, the latter being maybe the more recent of the two.

U14 is a position for an electrically erasable/programmable rom and there is also a header for even more electrically erasable/programmable rom. (Used for instance on the 5450A resistance calibrator)

Regarding the 7 segments the only thing that comes to mind is that option 010 should be able to display stuff in degrees celsius/fahrenheit so the ° and F C characters would be relevant.
If you can search for strings, that I attach a list of what should be displayed with different option selected.

the bottom ones such s everything related to temperature, db or LFAC are all only on option 010

[gmac34]

@ m k what software did you se to explore the rom?

[m k]

Segm7, it's my own, must add something for 14-segment, it's also bit inept with scrolling.
https://www.eevblog.com/forum/microcontrollers/old-school-8080-eprom-disassembly-is-it-making-sense/msg3490764/#msg3490764

Can't figure out how those 14-segment things are coded.
Some old 7-segment staff is still there, like Err, 8520A and IEEE, so 7-segment may still be for more than numbers.

I tried to find a 14-segment 'C' but can't be sure.
Its other byte is empty but can't find anything really rational.
Simple coding has few possibilities.
Straight a = 1 is 39 00 and inverted c6 ff, swapped byte order is also a possibility.
Old style goes so that a = high bit, so 9c 00 etc.

14-segment display circuit is 2 x 7, can't find the component(DL02614) internals either, NSN has that it's from Siemens.
One German page has several different order styles and none is 2 x 7.
Maybe somebody should measure how it goes, logic analyzer is not needed for characters like 'C' where other byte is silent.

U12 046a has a string of ASCII characters, but it's not complete, single characters are missing.

[gmac34]

I bet U12 is the one that changes between instruments with and without option 010.

in this schematics from the 5450a SM it mention that U12 could be the (C or D) and in the picture from wn1fju the sequence of part numbers skips a number.
So it might be:

U10: A  MK36321
U11: B MK36322
U12: C MK36323 or D MK36324 for option 010

I bet my instrument has a MK36323....

[Uunoctium]

I have an 8520A with the extension -01 = Opt 10.
U10, U11, U12 are all Mostek MK36321

Fatal disaster: at pulling U10 (yes, with an IC puller) two of the pins remained in socket.    Zero chance to pull em out.

Due to never seen an proper working desoldering device my frankensteieous solution was to stack two additional 24p sockets together to fix that.

I'm not really interested to start an further try - all of the Mostek pins showing heavy traces of corrosion (units property label is: Varta Batteries - probably operated in corrosive environment) and looking like the will falling apart only by shaking the IC.

And second, i don't know with which equvalent type of ROM or EPROM i shall readout the content? I own only the Xecu-pro. Are the MK36321 similar fancy devices, as the TEK MK36000?
https://w140.com/tekwiki/wiki/Mostek_MK36000

[Uunoctium]

Link to the 'Mostek 1977 Memory Databook' - the MK36000 family is described at p III-36 (p 49 in PDF)

https://bitsavers.pdp-11.ru/components/mostek/_dataBooks/1977_Mostek_Memory_Products.pdf

[m k]

I bet my instrument has a MK36323....

My bet is that -010 is more than one ROM.
MK36323 as original third ROM is almost certain, but MK36324 can also be just next revision of the same.

8520A Service Manual from 1980 I found is updated 1986, seems to be the same everywhere, original version would probably has different part numbers for ROMs.
There circuit page 109/130 has C or D for U12 ROM, but it also mentions 2764 or 27128, so the whole program code can be also double.

Found error codes, including display part.

Any idea how "OPen INPUT" is displayed?

U10 1dea has "rrtoo" but can't do "FAST" from next eight bytes.

I have an 8520A with the extension -01 = Opt 10.
U10, U11, U12 are all Mostek MK36321

Can you take a picture?

ROM is just a ROM, pin and speed compatibility are enough, speed means not too slow.

[Uunoctium]

Picture is in posting above (attached)


---------
From TekWiki:

The MK36xxx have address latches, therefore the /CE pin needs to be pulsed after each address change.

and it seems it needs an adapter to perform this
https://w140.com/tekwiki/images/a/a9/MK36000_ROM_reader.pdf

[m k]

Picture has different ROM versions.

[Uunoctium]

Yes, you're right. Did'nt see that at removing U10 but now on the image it is clear.
Btw, what are the differences between MK36321N, MK36322N and MK36324N?

[m k]

Programmed data.
They are mask ROMs, so factory made ready.

For single control pin you must force the dual control pin version to act like single control pin thing.
Usually that goes so that CE is forced active and OE is used as single control pin.

[gmac34]

the rabbit hole deepens... sorry I did not think they were an odd tipe. the later revision board documented in the SM uses regular UVPROM, non pin compatible with the MK36000. These mask roms are known in other instruments to go bad, ROM rot... it would be good to preserve, but cannot find good info on how to read them properly.

[m k]

Reading MK36000 is sort of easy, you just need an adapter.
But reader and active use can be different.
For active use you may need to short CE and OE together, it depends how the chip operates.
If double control pin chip is latched you must obey that of course.

For reader you flip all ID stuff off and pick a slow enough chip.
Then just read, even size doesn't need to be correct, just big enough.

MK36000 datasheet has access time as 300ns max.
Old 2764 versions are over 400ns, so pick one of those.
My MiniPro software has not or I can't see any visible access times.

Reader is a single chip situation.
So only necessary thing is to have data available when it should, so enough time from address to data.
Means that without ID check the reader can read a thin air and be happy.

[gmac34]

I have ordered a set of adapters, will try to read off my boards with a XGecu T76, if that works. I can send an adapter to whomever is willing to read off his boards.

[gmac34]

And here we have it 321 322 and 323, the standard ROMs , only 324 needed…

[gmac34]

So, I found the right headers in my parts bin, so I put togheter an adapter based on the pdf above, and it works as a 2764, apparently, or at least I get out of it the same data that is on KOBB.

[Uunoctium]

tl;dr

Tiny difference - don't know if that is relevant anyway. My Rom's are
U10: MK..321
U11: MK..322
U12: MK..324 and your instrument has an MK..323 installed here! 

[gmac34]

Yes, that’s THE difference. It seems the only rom that is different between option 010 and normal is U12.
The other two have the same number thus they have identical code inside

[m k]

Very much so yes, if it's not an unlock stylish, not likely.

Can't find the other end of that J17 5450 S1 from schematics, not EAROM type either.

So 8520A is still using 7-segment part for text and operation is serial.
"OPEn " is U12 07d7 and the rest are completely different, can be practically what ever also.

The actual display goes so that D0-2 are a single segment address and D7 is 7-segment data, for 14-segment parts data is D6.
Both sides have one 7-segment part active at the time, 4099 driver can't do more.
I/O address 0c6 is for 7-segment and 0c5 for 14-segment stuff, single LEDs are last 7-segment part, U25 has an I/O map and again a long nose to CPU maker and their support chips.

So 7-segment ROM style being familiar is just pure luck.

Changed Segm7 scrolling buttons to be more practical.