Core store chips

[woodchips]

I have exploded an old Ferranti core store, no alternative since it was all soldered together. But the Ferranti chips don't tun up anything on the web, possibly someone here has details please?

EXP441, EXP559, CN126, CN118, CN120, CN124, CN666, FRB1037, the last is difficult to read.

Going to unsolder them and try testing to see, can't be too complicated, high/low drivers, switching diodes, differential amps etc.

Photos attached, the core plane, each rectangle is 128x32, 13 bits seems to be 8 plus 5 ECC bits, plane total 16KB.

 

[Chuck(G)]

Are you certain that those are CNxxx ICs, and not ZNxxx? Just asking because that was Ferranti's line of standard logic.

Also, if this was anything like Ferranti's Pegasus, memory would be 12+parity bits.

 

[g4ugm]

The Pegasus had no core, only magneto delay lines, and no IC's only valves. Mercury had core.... but was discrete transistors. These are most likley Argus or Atlas boards.

Might be worth e-mail the Computer Conservation Society not sure who though!

http://www.cs.man.ac.uk/CCS/res/res61.htm#n

ALso MOSI (www.mosi.org.uk) hold the Ferranti Archives. There may be some data in there..

http://www.mosi.org.uk/collections/explore-the-collections/ferranti-online.aspx

 

[Chuck(G)]

Sorry, I had the Argus 400 docs in front of me, but I had Pegasus on the brain.

Some of those DIPs may not be ICs at all, but diode arrays and pulse transformers.

 

[woodchips]

Thanks for the replies.

They are definitely CN prefix, dated from mid 1970s and come from some sort of military computer.

All the chips are 14 pin DIL except the EXP659 which have 12 pins in a 16 pin DIL, these look like either steering diodes or high/low side drivers.

The PCBs seem to have lots of layers, possibly 8 or more, making track tracing hard.

There are a few pulse transformers, but doubt these parts are because not tall enough, like a DIL relay just a couple of mm taller.

Can't make any progress with the Ferranti archives.

 

[daver2]

Woodchips,

If they are from a Ferranti Argus 500 I may be able to help you when I get back from holiday and make a couple of telephone calls.

Dave

 

[g4ugm]

If they are Ferranti Military definitely try the CCS. Several folks in there from the military side of Ferranti. I believe that Argus 400 boards would be 24 bit. I know http://www.jb.man.ac.uk/ Jodrell Bank made AMD2900 clones of the 400 because their telescope driver software depended on 24 bit words....

 

[woodchips]

Thanks.

I have contacted CCS, but still can make no headway with MOSI. It seems strange that all the photos are of a complete machine, never the bits, so you can have a go at trying to identify things.

Bob

 

[g4ugm]

Try the contact form here:-

http://www.mosi.org.uk/contact-us

 

[g4ugm]

I spoke to Jan in Archives @ MOSI today and he said much of the technical stuff was retrained by companies that bought the IPR from the liquidators. A list of the current owners are here:-

http://en.wikipedia.org/wiki/Ferranti#Current_ownership_of_former_Ferranti_businesses

He also said MOSI would be interested in any technical documentation from Ferranti that folks had...

 

[woodchips]

Limited progress, traced the voltage regulators in the store but yet to power them up to see exactly what voltages they give, look like +5 and +/-20V. Finding it very hard to decipher what the chips actually are. The transistor high and low drivers were easy, but the others aren't. It isn't helped by the complexity of the core store design. A high driver wire goes through the plane, is joined to the adjacent wire and then goes onto the identical plane next door where the two wires are separated and go to two separate low drivers. Why? It doesn't seem to use steering diodes for the X and Y wires at all, every wire has a high and low driver. Possibly the design is because they simply ran out of room for all the chips?

Amazingly managed to buy some new old stock TI core driver and differential receiver chips. Came from Quarndon in the UK, but they also have a selection of ancient ALU chips like the 7451 etc AOI, ALUs, adders, carry lookahead etc in various technologies, if you want to make a real ALU then this is a chance.

 

[daver2]

Hi,

Back from holiday now and all rubbish e-mails deleted!

Part of Ferranti became Thales in Manchester.

For all things Ferranti computer related you may wish to contact:

Cevn Vibert
Solutions Architect
Thales UK
Mountbatten House, Basing View, Basingstoke, Hants. RG21 4HU
www.thalesgroup.com/uk
+44 (0)7791 080623.
cevn.vibert@uk.thalesgroup.com
I am also contactable on a range of Business Networking sites
such as LinkedIn http://uk.linkedin.com/in/vibertprofile

Cevn is particularly interested in Argus 500 and/or Argus 700 equipment. I have contacted him today and he is happy for me to pass his details on. He has posted on a few other sites (see for example http://www.theqe2story.com/forum/index.php?topic=2488.0).

If anyone has stocks of processors, core store or the like for the Ferranti Argus 500 or 700 he would be particularly interested in you contacting him...

Regards,

Dave

 

[George Baker]

Hi
I am new here but interested in getting an Argus up and running, 100 or 200 probably.

I can go in to the archive at MOSI and look stuff up for you. Jan is busy but does his best to help. I have seen 100, 300 and 350 circuits in the archive, there may be more

George