Ferranti PC31 - help needed

[gridrunner]

Hi everyone,

This is a bit of an obscure one, but I know we have some really knowledgeable people in the group and perhaps someone can help.

I recently got hold of a Ferranti XT Clone machine, which is the PC-31. My main reason for buying was the fact that it was built in Wythenshawe, just down the road from me - a fact proudly displayed on the rear of the case. I was a bit taken aback by this and it sent me down a bit a side quest to learn a little bit more of the history of Ferranti in the area. The factory was pulled down in 2011, but a number of notable computing firsts were achieved at the site. The scale of the losses to our manufacturing and technical base in the last 30 years is truly shocking. Anyway I digress.

The machine arrived with a note from the previous owner that the power had cut out when he tried to power it on and there didn't appear to be a CGA monitor with the machine to do any really testing.

The first odd thing was that Ferranti had installed two separate but identical PSUs, Antec 50 watt units, that put out 5,12,-5 and -12v across two identical connectors that plug in to the mainboard. Was there not a single, higher wattage PSU available at the time and is this a cost saving exercise? One PSU additionally drives the Rodime ESDI hard disk and the other a floppy drive and tape streamer. The thought has crossed my mind to retrofit a micro ATX PSU to take the place of these older PSUs.

I disconnected the PSUs from the mainboard and connected an old IDE hard disk in place of the Rodime to provide some kind of load. I then checked the mainboard power rails at the connectors for any sign of a dead short and there was none.

Both power supplies appeared to come up powering only the floppy disks and hard drive but the upper power supply then released magic smoke. A Rifa X2 cap. No surprises there and I caught it before it stank the place out. I clipped this out and went on to test the power, with both PSUs testing OK. I removed the expansion cards and then started adding them back one by one after re-connecting the mainboard.

What I get to now is a set of POST beeps - two long and one short. Luckily I still have the manual and this tells me that this is a video ram fault.

Judging by the dip switch setting this machine has a 640KB complement of RAM, a 8087 math co-processor that weirdly seems to have a heat sink attached and the video card - a ferranti job - that was set to 80 x 25 Colour.

I have tried using a 8 bit compatible VGA card instead of the CGA original but when I put this in and set the switches to EGA (with the VGA card using an emulation mode for EGA) I get no post beeps at all.

My question really was concerning the video ram. Would a fault like this be confined to the video card or could it be the mainboard? The are two 4416 DRAMs on the card itself.

Any points here would be greatly appreciated as I'd love to get this machine working. The scary part is the sheer number of Ferranti ULAs used everywhere which are no doubt unobtainium these days. Any anecdotes about this machine or Ferranti would also be interesting to read!

thanks

Stu

[Bill]

Hi Stu,
I worked for Ferranti at Gem Mill in Oldham,1969to74. I don't know anything about your PC but I was involved with assembly of prototype ULAs from 1972. The early devices would arrive in wafer form and we had a small group packaging the devices into ,usually, 24 or 40 pin DiL lead frames. Some were ceramic with a soldered lid and some moulded in plastic . Some wafers arrived with us with only 1 or 2 working chips out of maybe 100 on the wafer, after they had been probe tested, so management were not pleased if we wrecked a working device at assembly! Yields got better with time of course and mass production took over. I remember that we made ULAs for Ford Motor Co(some kind of auto gearbox control), Rollei cameras and other famous co's. After I left, ULAs appeared in all sorts of equipment, it was eventually a successful technology.
Bill

[acollins22]

Hi Stu,

I remember the Advance 86A&B but not the PC31.

I those days the video cards tended to have their own RAM. Pinching RAM from the processor was a later trend.

Have you tried the MOSI museum for information. They have a lot of Ferranti info I believe because of the Manchester connection.


Cheers,

Andy.

[bobbyball]

I worked for Ferranti at the time of the "Advance" series of PC's and later but do not remember the PC31. The Advance series had a RF modulator so could be used with ordinary TV sets or a CGA monitor. Happy days... There was an "XT" clone machine that came after the Advance series and had a metal case, yours looks possibly one of these,

My father was General Manager at Wythenshawe for a while...

I'm sorry I can't really help you but I hope you get it sorted.

[gridrunner]

Hi Robert,

thanks for the reply. The machine is an XT Clone, I think Ferranti's first machine got in to a bit of legal trouble because the BIOS couldn't be shown to have been developed in a 'clean' manner in the way Compaq managed to do it.

I have refurbished just about the power supplies to refit in the machine. I took a copy of the two BIOS EPROMs which are 16k chips. Given the manual was pointing to a video ram fault with the two long, one short POST beeps, I've removed the two RAM chips from the Ferranti CGA card ready for two socketed replacements.

The other card featured is an Orchid Prodesigner II ISA card that is a VGA card that also features XT compatibility. Rather interestingly, the card also has a CGA and EGA emulation mode in software on the card that is enabled by a DIP switch. However it also needs an optional 3rd ROM with that code. I didn't have this, which might explain why the machine failed to boot with it. Not all XT BIOS's support VGA operation. Luckily I found the image for that option ROM and was able to burn one and fit it to the card.

So once I can finish the PSUs I'll put it all together and restest!

[gridrunner]

Hi Andy, Bill,

thanks for replying to my post. I did find a couple of interesting links to the building and the work carried out there.

This PDF document from the computinghistory.org.uk site, provides a lot of history on the computing work of Ferranti as it relates the development of the Argus Process Control computer for the Bloodhound Missile.

http://www.computinghistory.org.uk/u...shed_2012_.pdf

There's also a photo album taken on the last day of business in March 2010 of the Atlas House site.

https://www.***********/photos/487522...7623709591704/

[Nuvistor]

Can’t help with the PC31.
I worked for a supplier and frequently visited Gem Mill, we installed and serviced a computer in a clean room.
Also computer installations in another Oldham site, can’t remember the name and also installations in Wythenshaw.
This would the 1980’s and early 90’s.

[gridrunner]

Hi everyone,

I'm pleased to say that the Ferranti PC31 is now up and running. It was a little bit of a challenge to get there, so here's a bit of diary of what's been done to it thus far.

The machine is powered by x2 Astec 50 watt PSUs. Both had Rifa mains filter capacitors that blew, otherwise they tested fine. I've replaced the majority of the capacitors in both including the X2 rated filter caps.

The machine wouldn't boot owing to a video card fault. The BIOS beep codes (two long, one short) referred to faulty video RAM in the manual and this proved to be the case. I replaced the RAM with socketed new chips (x2 64KB DIL) and the machine booted.

However the quest for a CGA monitor left me a little bit short. I found locally on ebay what I thought was the answer, a Thomson 450G monitor. I'd read online that Commodore used this chassis re-badged for the 1401 CGA and Composite monitor for the C128 range of machines. However the one I picked up felt suspiciously light and lacking in external ports. On opening up, the PCB looked very bare and I was dismayed to find I'd bought a green screen CGA monitor. If anyone would like this, let me know!

However a spare ISA video card came to the rescue, an Orchid Prodesigner II, which happens to emulate CGA and EGA modes with the right EPROM software. This card is now driving as close as I can get to a period LCD monitor. There's a link to the firmware here:

http://www.vcfed.org/forum/showthrea...ranslation-ROM

A Gotek drive with flash floppy firmware is sitting as drive B on the machine. There were two Shugart SU455 drives which both work fine and the drive A unit has been moved to one of my other old PCs, a more modern P200. I did notice that one of the drives was not allowing writes to floppy disks, seeing all disks as write protected. The IR photodiode that peers through the write protect notch appeared to be defective. Fortunately, a jumper on the board (WP/WP+) appears to disable write protect on the driver itself, which was handy to find.

The hard disk is surprisingly not MFM, it's a SCSI Rodime drive with 20MB of storage. It's working fine so far, but threw out one read write error trying to access one of the games on the disk. It was last used in 1994, looking at some of the files on the drive.

One final tweak was to replace the AMD 8086 CPU with an NEC V30 CPU, just to provide a bit more welly. There is a co-pro fitted, which is also AMD I think, but oddly it has a massive heatsink stuck on to it.