All the cheese wedges, in order

[BigEd]

In reviewing the story of Acorn's second processors, I thought it might be interesting or amusing to try to collect a chronology, or at least a catalogue, of all the cheese wedge products.

Any additional info, such as year of introduction, more variations, perhaps approximate production quantities, maybe even price at introduction, would be very welcome.

Here's what I think I know so far, or think I can guess, in order of appearance:

• ANE01 Teletext
• ANE02 Prestel Adapter
• ANC01 6502 Second Processor (3 MHz)
• 80286 Second Processor (never-released "ANC03 Second Processor - 16 bit"??)
• ANC04 Z80 Second Processor
• ANC05 32016 Second Processor
• ANC06 Cambridge Second Processor (also a 32016)
• Extended 6502 Second Processor ("Turbo" 256k)
• ANK01 IEEE 488 Interface (1983?)
• AEH20 Econet Bridge
• ANC13 ARM Eval System (1985)
• ANC21 Universal Second Processor (is this for external fitting of a Master internal device?)

I did once snap the collection at TNMoC:

I got some info from Chris's Acorns. See also BeebMaster. There are of course other second processors not made in cheese wedge format.

[hoglet]

On these BBC Micro Order Forms there is mention of an Acorn ANC03 (Second Processor - 16 bit):
http://chrisacorns.computinghistory.org ... mAug83.pdf
https://www.4corn.co.uk/archive/docs/AM ... 4)-opt.pdf

There are also prices and release dates.

I wonder if that was the unreleased 80286?

Dave

[BigEd]

Good find! That leads me to this list of Product Codes which helps fill in some gaps.

[Boydie]

There's an article in Beebug Vol 1, Iss 6 (Oct 82) about the Tube. This refers to the 6502, Z80, and a "16 bit NS 16032 based machine", which goes along with what Acorn claimed they would be producing. It would seem there was a common perception that the 16032 was 16-bit, which was why NS renamed it to 32016 in order to emphasise its 32-bitness. It would make sense, then for Acorn to have a different product code to also emphasise the 32-bit nature of their 2P and distance it from what had been announced as a 16-bit model.

Also, the original 16032 2P as announced was only supposed to have 256K of RAM. Hence the 1024K launch model may have required renaming.

Perhaps ANC03 was the 256K 16032 prototype, which was itself never released?

Also, I don't know how Acorn numbered their PCBs, but the 6502 is 0208 000, Z80 is 0209 000, and 32016 (and 16032) is 0210 000. Whereas the 80286 is 0229 000. This would imply the first 3 were designed close to each other, whereas the 80286 was some time later (I suspect it was designed for the ABC range, with the offshoot idea of a cheese wedge coming later; presumably it was then canned because it would have out-performed the Master 512).

Incidentally, has anyone else noticed the anticipated availability for the 6502 2P, according to the price list on Chris' Acorns? They must have known how late it was going to be (and in fact were only 2 quarters out)!

[BigEd]

Excellent extra info, thanks! I like the idea that the 16032 was the mystery machine. Much like the 68000, it's 32 bit on the inside and 16 bit to memory. It just happens that it got renamed to 32016, and it took very many revisions to get enough of the bugs out: it might be that 256k was never going to be enough, or that the chip just wasn't yet fit for purpose, or not yet able to be clocked fast enough to be useful.

[Boydie]

From reading old Beebugs and Acorn Users from the period, it would appear that the real reason for delayed release was that Acorn found it rather trickier to get the Tube working properly. Beebug from Sept 1982 lists all three 2Ps, giving provisional production dates of November 1982 for the 6502 and Z80, and February 1983 for the "16 bit". In reality, the 6502 and Z80 didn't hit until mid-late 1984.

So, by the time the "16032" was ready, it had been renamed, probably had all (most?) of the bugs ironed out, and memory was a lot cheaper. Plus, Acorn were working on the ABC range, with up to 4Meg in the scientific workstation version (ie the 32016), so their software ambitions had probably moved too and 256K really wouldn't cut it (which makes it all the more odd that they should suggest, and get to the point of designing, a 512K Master Scientific). The cheese wedge version only used the 6MHz part, which was the slowest one NS made, so I doubt clock speeds were an issue (so long as you discount the fact that it broke the standard ADFS Winchester access because anything less than 8 MHz couldn't keep up).

[fordp]

The second processor I had was originally a 256K 16032 prototype. A bit of a RAM swap and it was upgraded to 640K. The lid was swapped and it then had a 32016 lid.

[roland]

You'll need a very large table to connect them all at the same time
Wasn't there an official 6809 Second Processor for the Beeb? After all, there was a 6809 board for the System.

[jgharston]

You'll need a very large table to connect them all at the same time
Wasn't there an official 6809 Second Processor for the Beeb? After all, there was a 6809 board for the System.

Only unofficial systems such as Graham Toal's.

[KenLowe]

My 6502 co-processor has a '16O32' lid for some strange reason. It's discussed briefly in this post. Unfortunately I don't know the history of this co-pro, but it did also come with a very early Tube Host Code, v0.01 which I've subsequently upgraded to v0.05 to make it compatible with DNFS1.20 (with tube host code v0.01, the co pro would only work with NFS 3.34). This thread provides some more detail.

[BeebMaster]

The AUG (I have 3rd edition November 1983) says on p.434 "Currently, three second processors can be used on the Tube. These are a 6502, Z80 and 16032."

Maybe Bray, Dickens or Holmes had one!?

Very similar to the ARM Evaluation system is a (possibly unique?) board marked "2/4 Megabyte 'A' Second Processor".

The only 3rd party bit of kit to use the official Acorn cheese wedge box that I've ever seen is the RH Electronics Video Digitiser.

The IEEE488 interface is a bit unusual in that it came in cheese wedge and non-cheese wedge boxes, and is the only Acorn wedge that doesn't have an Acorn-branded board inside.

I wonder if people would count the Music 500 as a cheese wedge - granted it's not in a cheese wedge box but I reckon it could easily have been, and doesn't have an Acorn board inside, but does have Acorn branding on its box, and is an important item in Acorn's range of BBC micro expansions, so I would say to all intents and purposes it is a cheese wedge.

[BigEd]

(ARM was the result of a secret project called 'A' so that board is perhaps carefully not using the 'ARM' word in order not to leak information - if so, it would predate the Eval unit.)

[hoglet]

The AUG (I have 3rd edition November 1983) says on p.434 "Currently, three second processors can be used on the Tube. These are a 6502, Z80 and 16032."

Maybe Bray, Dickens or Holmes had one!?

Actually, I think I have one

The board layout is quite different to the other pictures on the internet.

(It's also rather grubby, sorry about that)

This is the one given to me by FordP a few years ago. The memory has been upgraded, I think to 640KB.

Dave

[sweh]

Actually, I think I have one

Funny; the board says 16032 but the CPU says 32016; according to wikipedia NS renamed the chip https://en.wikipedia.org/wiki/NS320xx#32016

[BeebMaster]

I wonder if Acorn anticipated the development of the chip, and possible pinout or other changes, and left IC4 for the 32016 and IC3 for the 16032?

[hoglet]

I wonder if Acorn anticipated the development of the chip, and possible pinout or other changes, and left IC4 for the 32016 and IC3 for the 16032?

I think the spare socket is for the NS32382 MMU.

[stuslayer]

that 16032 board state ISS. F - doesn't that indicate that it's a prototype, as Acorn used letters for proto's and numbers for releases?

[fordp]

My 6502 co-processor has a '16O32' lid for some strange reason. It's discussed briefly in this post. Unfortunately I don't know the history of this co-pro, but it did also come with a very early Tube Host Code, v0.01 which I've subsequently upgraded to v0.05 to make it compatible with DNFS1.20 (with tube host code v0.01, the co pro would only work with NFS 3.34). This thread provides some more detail.

Maybe it was on my Copro at some point. 16032 lids were only on prototypes as far as I know.

[BeebMaster]

Ring up Davina McCall, there's an episode of Long Lost Cheesewedgery in the making here!

[RobC]

I wonder if Acorn anticipated the development of the chip, and possible pinout or other changes, and left IC4 for the 32016 and IC3 for the 16032?

I think the spare socket is for the NS32382 MMU.

Yes it is. I have fitted the MMU to mine in the past but you're supposed to change some link(s) to introduce additional wait-states so it makes the co-pro slower. That, coupled with the fact that I'm not aware of any available software using the MMU, made me take it out.

[Dave C]

I have a cheese wedge Teletext Adapter. What can I do with it now analogue TV is toast?
Dave C

[BigEd]

You can do something using a Raspberry Pi
https://github.com/ali1234/raspi-telete ... spberry-pi
But I think you'll need either to figure out a modulator story to get RF to send into your teletext adapter, or will need to figure out how to inject composite video into the adapter.

[guesser]

The easiest way to set up the pi with some teletext services is to install vbit2, then connect its composite output to a UHF modulator (you can use an old VCR for this, or a standalone unit). From there just follow the Advanced Teletext System manual to tune it in.
Alternatively others have said you can pull a link inside the adapter which disconnects the tuner section and patch the composite signal in directly, but I haven't tried that myself.

[awilliams]

> ANC21 Universal Second Processor (is this for external fitting of a Master internal device?)

Yes it is.
I have used these with the internal 80188 and internal 6502 boards externally.
My work M128 had the internal 6502 and the 80188 in one of these units.
Potentially you might be able to use these with a BEEB too but I have not used one in that combination.
I am regretting consigning two to hard rubbish at some time in the past.
Alan