AI Slop Video claims first 40007 GAs were not made by Ferranti but by SGS?!

What would happen if you decap an early 40007 and it turns out to be a 16RA? You still wouldn’t know if there were batches made with 20RAs or not.

What would you do in that case? Decap every 40007 you can get to confirm your theory?

It can be anything from a typo, a later change that didn’t make the documentation…

No, i think its not necesary decapp more that one.

Because we know these things.

A)the CPC 464 was launch in uk the 21 de june de 1984 with the gate array working.

B) we know that Alan sugar present the CPC 464 in 11 of april of that year, and he don’t have the ula working have the board of 50 chips inside that he show to the people present in the event

C)Ferranti made the ula pysically in uk,but the CPC is made in Korea,so the chips need go to Korea put It in the CPC and return to UK inside the 464 beetween april-june.

D) from this source: You’re NOT fired: The story of Amstrad’s amazing CPC 464

“The ULA would not be completed until February 1984, but the PCB layout was done and dusted during November 1983”.

And this

"Locomotive got to look at the the first working hardware at the end of October 1983, though at this stage the machine’s ULA chip was not ready so it had been replicated in circuit board form so the software guys knew exactly what it could and couldn’t do, and could test working hardware. This board, designed and built by Hurrey, was dubbed the “Gate Array Simulator” – or “the Gas Board”, for short.

So no we have this time line:

October 1983->gas simulator.

Novemeber 1983>design of the first board completed=z100

Febrero 1984 ula terminated.

April 1984 sugar don’t have the ula so probably its in the way of the far east?. Or in production?.

So the first batch of ULA should be a date Code february-march maybe.

With one of these date Code should be the first batch. If he have 16ra they never use 20ra comercially.

You’re not considering the possibility that the 40007 we sent for decapping was actually one of the earlier ones… :smiley:

What date have?(You never tell It).

I was wrong in the other post i try edit It but the forum ignores me jaja.

Taked this as reference:

October 1983->the gas board was made.

November 1983->the layout of first board was ended

December 1983->ROM Code finished.

February 1984 ->ULA was ended.

Note:we know that after send the final design to ferranti the first chips don’t work.

April 1994-> cpc 464 presentation sugar don’t have the ula.

June 1985-> cpc 464 begin sold with ULA working.

I don’t know if I still have the pictures of the chips before decapping, I’m looking for them.

On the other hand, nothing you are listing is any evidence of the use of 20RA rather than the mention on the service manual or specs, and you cannot trust that either as it may have been an early estimation.

Ok, at the moment in the internet pictures, i don’t view something less that 8428.

Week 28 from 1984 thats july…to old

I think that the design started with the GAS in the first place - after all, you want to make sure that your GA is working on the RTL before making a batch of Gate Arrays… prototyping. So, I think that the narrative that the GAS was ONLY made because the software guys needed it when the real GA wasn’t available yet is almost certainly false. Every rationale hardware designer would have started with a GAS in the first place IMHO, before designing and ordering a batch of GAs.

1 Like

Latest news!

I compared the 40007 and 40008 dies and they are identical. Not similar, the same.

Same 16RA023 and same routing.

I think that when they did the service manual they thought they’d need a 20RA043 but ended up fitting in a smaller 16RA023.

The 40008 in the 664 manual and the 40010 in the 6128 manual appear as HSG3130 or HSG3170, which does not match any known manufacturer numbering scheme. But they are suspiciously close to LSI Logic’s LL3130 and LL3170 part numbers.

On top of that, the two variants of the 40010 (36AA and 37AA) have the numbers 25Q13Z and 25Q17Z, which match the number of gates in the LL3130 (1332) and LL3170 (1722).

In this case, I think they were preparing to jump to LSI Logic/SGS but had to stick with Ferranti for another round.

For the grand finale, in the 6128 service manual, there’s a picture of the Z70210 MC0009A board, that has two slots for gate arrays. The top one for the 40010 and the bottom one for 40007/40008. Interestingly, the bottom one has a inscription on the board that reads “ULA FERANT”, which may indicate that previous 40007/40008 chips are indeed Ferranti’s, so the service manual could contain not-so-final information on certain parts like the gate array.

1 Like

But thats don’t made sense 40007/40008 can’t be exactly equal, something need change in the design, because the 40008 don’t need heatshink

From our wiki page on Amstrad part numbers:

40007 Gate-Array version 1 (20RA043) (uses a heat sink, same pinouts as 40008)
40008 Gate-Array version 2 (HSG3130) (uses a heat sink, same pinouts as 40007)
40010 Gate-Array version 3 (HSG3170) (no heat sink, different pinouts than 40007 and 40008)

Also, if you look at the 664 and 6128 service manuals, the gate array has the text “HSG3130 or HSG3170” on both of them. It doesn’t state that the 40008 is the HSG3130 and the 40010 is the HSG3170.

But thats is bad, we know that the HSG3130 an HSG3170 are the 40010. lsi chips.

We can edit that, not a big deal.

That said, we better get it right because if someone searches for LL3130 on Google we seem the be the main source for the results… :joy:

1 Like