So, when did Amstrad decide the world had all the CPC's it needed?
Surely not 1990, on this very forum someone has a 1991 build classic 464
Did they only do one production batch of plus's or keep going for a while I wonder
On a related note, anyone know when the PCW (not PCW16) was made till?
This factiod is of interest as it was the last CP/M machine of all. At least 1993 for sure
I also wonder how much the CP/M licenses cost per machine, nice to think DR made something out of it...
Regarding the PCW series, the 8000 series seems to go until 1988-89, the 9000 series up to 1991, the PcW10 up to 1992 and the PcW16 up to 1995, according to PCB markings and chip manufacturing dates.
The standard CPC series seems to go as far as 1989 and the Plus series, seems to be 1990 only (probably some 1991).
QuoteSurely not 1990, on this very forum someone has a 1991 build classic 464
Can you point me to it? Would like to take a look at any picture of it.
If you're referring to Markus' board, his is a Schneider from what I remember. Their deal with Amstrad fell through in 1988 but they still had a ton of stock that was sent back to Amstrad which were rebranded as Amstrad in 1990 (with the grey keys and grey record button).
In the UK and most of the European market, the classic CPC was finished in most shops by Summer 1990. There may have been a few left over old stock to clear however.
Quote from: robcfg on 11:59, 16 January 23Regarding the PCW series, the 8000 series seems to go until 1988-89, the 9000 series up to 1991, the PcW10 up to 1992 and the PcW16 up to 1995, according to PCB markings and chip manufacturing dates.
The standard CPC series seems to go as far as 1989 and the Plus series, seems to be 1990 only (probably some 1991).
QuoteSurely not 1990, on this very forum someone has a 1991 build classic 464
Can you point me to it? Would like to take a look at any picture of it.
Here you go:
https://www.cpcwiki.eu/forum/index.php?msg=7355
I do not doubt that he bought it in 1991, but almost all the chips in Markus' board are from 1987, so at least the board was made around that time and not 1991.
Quote from: robcfg on 12:58, 16 January 23I do not doubt that he bought it in 1991, but almost all the chips in Markus' board are from 1987, so at least the board was made around that time and not 1991.
I'd say the chips are around 86-87, but that's a proper cost down board, I don't think those were being manufactured in 87. Definitely between 88-90 though. Like yourself, I'd highly doubt the board came out in 91.
I find it difficult to believe that all MC004x boards we have scanned so far, all have ICs from 1987 and 1988 for it to be leftovers used on cost-down PCBs.
Add in that they date of the ICs is quite consistent on every board.
I think that they designed/started manufacturing the cost-down boards around 1986-7 so you'd still see a lot of regular boards those years before the cost-down ones were the norm.
The pre-asic cost-down boards have mixed 1988-89 ICs and are labelled (C)1988, so I think it makes sense for the cost-down version to have started around 1986-7.
Quote from: Shaun M. Neary on 14:28, 16 January 23Quote from: robcfg on 12:58, 16 January 23I do not doubt that he bought it in 1991, but almost all the chips in Markus' board are from 1987, so at least the board was made around that time and not 1991.
I'd say the chips are around 86-87, but that's a proper cost down board, I don't think those were being manufactured in 87. Definitely between 88-90 though. Like yourself, I'd highly doubt the board came out in 91.
IC's have a commercial shelf life of around 8 to 10 months at most. So if the chip says 87 then the board was built at very latest mid 88.
Bryce.
It really wouldn't make sense for Amstrad to still be manufacturing old CPCs right up until (or even after) the launch of the Plus models (I mean maybe Commodore would have but they had crazy release schedules).
They, or retailers, may well have been trying to shift old stock by then but actual new machines seems very unlikely.
where did the information "1991" come from? In the original thread there is not much information around that. Date of purchase? something on the label?
I pretty much doubt that you could find a real Schneider with manufacturing date of 1991. While I have no doubts that you could find a dusty Schneider box as "new old stock" in a weird shop in 1991, I doubt it was built in 1990 or 1991.
Also I would expect Amstrad would not want to compete with itself, so I guess they did not build any more CPCs once the Plus was out. But I have never read anything about that.
Quote from: eto on 16:20, 16 January 23Also I would expect Amstrad would not want to compete with itself, so I guess they did not build any more CPCs once the Plus was out. But I have never read anything about that.
Pretty much this, from what I understood, they didn't do too much building of CPCs in 89 and 90, but Schneider had sent back a big bunch of machines after the 88 fallout, and they were rebranded as Amstrad machines during that period of time. With all the stock sent back, there wasn't any real need to produce more of a product they were phasing out in favour of the Plus series.
Like this one!
(https://www.cpcwiki.eu/imgs/f/f1/WalgenbachAmstrad_CPC464_grey_Large.jpg)
ah, good replies.
I found a site that lists CP/M bios revisions, there is a 1992 PCW10 and a 1993 one as well, suspect they would be producing PCW's till 1993 at least going by that. Not sure if the PCW10 replaced or augmented the previous range
I'm not convinced by the chip dates, what if they bought in bulk and gradually used up, the date on the case label seems like the only way to really know for sure. The chips may hang around in warehouses in between the manufacturer and where they end up, who knows what strange purchasing hacks they used to get cheapest prices, I bet that was the orders. I mean I buy stuff and often find the date of manufacturer is over a year after I bought it
I got a 464 from summer 1990 going by the label..
QuoteIC's have a commercial shelf life of around 8 to 10 months at most. So if the chip says 87 then the board was built at very latest mid 88.
You can watch many YouTubers guesstimate manufacturing dates with this method.
Labels can be treacherous as you can mix and match pcbs, labels and cases so that's no guarantee.
Plus, no company likes to have a big stock of nothing as it costs a fortune.
Last, the pcb estimated dates match pretty well the IC dates.
QuotePlus, no company likes to have a big stock of nothing as it costs a fortune
I'm not so sure, the manufacturers were contractors, made other products. Any really common chips could come from a larger pool of stock, the CPC specific ones maybe more reliable for dating. There's scenarios like bulk buying based on forecasts then they turn out less and there's years worth of stock to use up. I seen manufacturers with 10 million+ inventory and not even that big factories, in manufacturing purchasing nothing seems to go to plan!
Quote from: martin464 on 22:45, 16 January 23I'm not so sure, the manufacturers were contractors, made other products. Any really common chips could come from a larger pool of stock, the CPC specific ones maybe more reliable for dating. There's scenarios like bulk buying based on forecasts then they turn out less and there's years worth of stock to use up. I seen manufacturers with 10 million+ inventory and not even that big factories, in manufacturing purchasing nothing seems to go to plan!
We have a thread where a 1991 German Schneider CPC is mentioned. We only have pictures from the motherboard with ICs from 1987 but non from the case or the label and serial number on the case. We also know for sure that Schneider and Amstrad parted ways in 1988.
There are two possible explanations
1) This is really a 1991 Schneider CPC. The manufacturer bought so huge amounts of all required ICs for the CPC in 1987, that even 4 years his stock was so huge that CPC motherboards were built without a single more recent IC made it on the motherboard. For some reason (or accidentally), Amstrad or the manufacturer put a Schneider label on this machine and shipped it to a German shop.
2) The computer was built in 1987/1988 as one of the last Schneider CPCs but it was only sold in 1991
Seriously, unless there is more evidence of this 1991 Schneider CPC I would prefer the second explanation and focus on other sources to find evidence for the last produced CPC.
Well, after looking at all our PCB scans, 94% of them have all ICs from the same year or from the same year and the previous one.
Specially on the Plus and GX4000 boards, only one of them has chips from previous year (Board is (C)1991 and has mostly 1991 ICs with some 1990 ones).
The rest of them have at most a difference of 3 years and are mostly 74LS logic chips, usually being 4-5 chips and having a range of manufacturing dates inside the 3 year difference.
In my eyes, it is clear how Amstrad worked with IC suppliers and could confirm that the last CPCs, being Plus and GX4000 were manufactured until 1991. The last standard CPCs were manufactured until 1990.
I'd like to point out that there's one chip, the Gate Array, that was custom made for Amstrad, and that follows the same rule as the other chips, so they had no big stock of chips piled somewhere.
Tomorrow I can take a look at the PCW board scans, but I think we'll be seeing the same picture.
1991 makes total sense, thank you! I remember seeing +'s in the high street in that year for sale
My guess for the PCW is 1994 from the BIOS revisions but this guys dates are guesses as well I think, cool site though https://www.seasip.info/Cpm/xbiosint.html (https://www.seasip.info/Cpm/xbiosint.html)
PCW 1.12 / 2.12: 1992Supplied with the PcW10.- Support added for the standalone Centronics port accessory.
PCW 1.13 / 2.13: 1993- Support for loading drive parameters from *.FIB files.
PCW 1.14 / 2.14: 1993- Improved Centronics port detection.
- If the memory test reports 256k, reset the memory read/write register and try again.
- SHIFT+EXTRA+RELAY now works on first-generation PCWs as well.
PCW 1.15 / 2.15: 1993- A single change: When doing a SHIFT+EXTRA+RELAY reboot, takes all devices out of interrupt mode so that they don't interrupt at an awkward moment.
So, checked the PCW boards, and as expected they follow the same pattern as CPC ones even more closely.
They all have ICs with date codes of the same year and/or the previous one. I found only two of them with one or two chips from two years earlier.
The NC computers were manufactured by Nakajima and customized for Amstrad, and they seem to have the same ic date pattern.
There was some discussion some years ago.
https://www.cpcwiki.eu/forum/index.php?msg=155028
https://www.cpcwiki.eu/forum/index.php?msg=156127
@martin464 , regarding PCWs, it looks like the 8000 series was going strong up to 1987-88.
Then, the 9512 was introduced in 1987 which is confirmed by the markings on the board and superseded by the 9256 and the 9512+ on 1991.
The PcW10 came in 1992 and the PcW16 in 1995, but my board has 1996 date-coded chips, so my guess is that it was released towards the end of 1995 and into the beginning of 1996 as they sold very few units.
Quote from: robcfg on 01:04, 18 January 23@martin464 , regarding PCWs, it looks like the 8000 series was going strong up to 1987-88.
Then, the 9512 was introduced in 1987 which is confirmed by the markings on the board and superseded by the 9256 and the 9512+ on 1991.
The PcW10 came in 1992 and the PcW16 in 1995, but my board has 1996 date-coded chips, so my guess is that it was released towards the end of 1995 and into the beginning of 1996 as they sold very few units.
Thankyou Rob, so would the PcW10 have been made till 1995 (or perhaps along with the 9256/9512+ not sure if the 10 replaced these or was an extra model in the line up). Curious as in terms of computer history it would be the last CP/M machine machine made it kind of should get a special mention when CP/M history comes up i think
I thought the serial numbers on the CPC have the build date buried somewhere in them? There used to be a serial number decoding thread on here somewhere.