CPC Mainboard Versions

[Terje_Norway]

Hi,


You obviously have one of the newer machines, but they didn't stop production of the ORIGINAL CPC464 until January or February 91. 


- How do I know this You might ask ? ? ?


The case is simply that I have TWO of the latest batches of machines. They have labels with 90-12 and 91-01.
These machines came from Weeske in Germany (they were official dealers in those days). Both have matching serial numbers on both the boxes and machines.
What makes these so special is that they seems to be built of "leftover" parts.


You could have a look at the following post in this thread over 7 years ago.


http://www.cpcwiki.eu/forum/amstrad-cpc-hardware/cpc-mainboard-versions/msg7334/#msg7334


Yours

Terje Grind
NORWAY

[Shaun M. Neary]

Definitely interesting, but if they were still being produced, my question is, who was selling them?

I never saw the CPC's after the Plus range came out anyway!

[dragon]

Hi,


You obviously have one of the newer machines, but they didn't stop production of the ORIGINAL CPC464 until January or February 91. 


- How do I know this You might ask ? ? ?


The case is simply that I have TWO of the latest batches of machines. They have labels with 90-12 and 91-01.
These machines came from Weeske in Germany (they were official dealers in those days). Both have matching serial numbers on both the boxes and machines.
What makes these so special is that they seems to be built of "leftover" parts.


You could have a look at the following post in this thread over 7 years ago.


http://www.cpcwiki.eu/forum/amstrad-cpc-hardware/cpc-mainboard-versions/msg7334/#msg7334


Yours

Terje Grind
NORWAY

Maybe you can take picture of it? Label/computer, please?




Well there is the oficial stop production from amstrad(alan sugar always work with orion in far east.)


And another diffent when importation stops in uk france or germany.


I never view a spanish computer >5 /89. The middle 90 is a uk model?.


In the other hand you 91 german computers can explain why the 40226 spare are out. If amstrad stop the producion in 1990  and they sell the rest of components to local german(the 40226 was made in germany). Mostly are used in this local computers.

[dragon]

I have read a interesting note in the sof968 in the wiki that suggest amstrad include a certain model of crtc to avoid a bug in the ntsc 6128. All American cpc mount the same crtc.


What variant "not have a
programmable frame flyback pulse length."?

[Bryce]

I have read a interesting note in the sof968 in the wiki that suggest amstrad include a certain model of crtc to avoid a bug in the ntsc 6128. All American cpc mount the same crtc.


What variant "not have a
programmable frame flyback pulse length."?

NTSC 6128?? The CPC is neither PAL nor NTSC, it's RGB and there was no special US version. Any difference would have been in the MP-1/2, not the CPC.

Bryce.

[pelrun]

Just read NTSC as shorthand for "the version sold in the US".


The monitor uses the line frequency as a timebase, and since that's 60Hz in the US the monitor calibration changes (to maintain the proper screen height) and the CRTC settings are different.

[arnoldemu]

I have read a interesting note in the sof968 in the wiki that suggest amstrad include a certain model of crtc to avoid a bug in the ntsc 6128. All American cpc mount the same crtc.


What variant "not have a
programmable frame flyback pulse length."?

MC6845 or UM6845R.

[arnoldemu]

NTSC 6128?? The CPC is neither PAL nor NTSC, it's RGB and there was no special US version. Any difference would have been in the MP-1/2, not the CPC.

Bryce.

It would be nice to confirm what is different.

The CRTC values used at startup are different and define a 60Hz display. The PCB link is set for 60hz. They also have a UM6845R or MC6845, but beyond that there is no absolute confirmation.
I expect some kind of 110V power supply inside the monitor.

I would also be interested to know if the monitor is the same but the vhold has been adjusted or why would the CRTC values would need to be different?

[Bryce]

It would be nice to confirm what is different.

The CRTC values used at startup are different and define a 60Hz display. The PCB link is set for 60hz. They also have a UM6845R or MC6845, but beyond that there is no absolute confirmation.
I expect some kind of 110V power supply inside the monitor.

I would also be interested to know if the monitor is the same but the vhold has been adjusted or why would the CRTC values would need to be different?

So you're saying that there's a US version of the Lower ROM? I only know of the 40025, what was the US version called?

Bryce.

[arnoldemu]

So you're saying that there's a US version of the Lower ROM? I only know of the 40025, what was the US version called?

Bryce.

No. 40025 already has the code to look at the 50/60hz link and set the crtc values. I guess it'll use 40025, soft968 doesn't indicate there is a different rom.

[Bryce]

No. 40025 already has the code to look at the 50/60hz link and set the crtc values. I guess it'll use 40025, soft968 doesn't indicate there is a different rom.

Ok, but 50/60Hz is the scan frequency, a seperate subject to PAL/NTSC which are encoding protocols.

Bryce.

[dragon]

I quote what the soft 968 tells for people not have read it


"American CPC6128: Frame Flybacks and Interrupts
This technical note discusses the relationship between frame flybacks and interrupts on the
NTSC version of the CPC464/664/6128. Currently the only production model affected is the
American version of the CPC6128 - all other markets use PAL/SECAM models and this note
does not apply to them.
There was an error in the original hardware specification for the CPC464/664/6128 in that
the value to be loaded into the Vertical Total Adjust register in the HD6845 (register 5) was
incorrectly given as 6 for the NTSC version whereas it should have been 4. Thus the various
ROMs produced for the CPC464/664/6128, when used with an NTSC system, load an
incorrect value into the 6845 whenever a full reset occurs; for example when the machine is
first powered on, or the RESET_ENTRY firmware call is made.
PAL/SECAM systems work correctly and, fortunately, the only effect of the incorrect value
on NTSC systems is to cause the interrupt associated with frame flyback to occur at exactly
the same time as the frame flyback pulse becomes true. With the correct NTSC value the
interrupt will occur 125 microseconds after frame flyback becomes true - corresponding to
PAL/SECAM systems and the description given in Section 1.
What this means in practice is that on an American CPC6128 a program which tests the PPI
Frame Flyback signal (for example by calling MC_WAIT_FLYBACK) will not see frame
flyback become true before the interrupt occurs, but must rely upon frame flyback still being
true when the processing associated with the interrupt is complete. If the interrupt processing
takes too long, the program will appear to 'Lock Up' because it never sees frame flyback set
true.
In order to minimise the possibility of 'Lock Ups' occuring Amstrad have ensured that all
American CPC6128 machines are fitted with the type of 6845 which does not have a
programmable frame flyback pulse length. Therefore the frame flyback will last for a fixed
1000 microseconds rather than the programmed time of 500 microseconds. The 500
microseconds period is in fact quite sufficient for the system routines invoked by the frame
flyback interrupt; 1000 microseconds will allow a number of user routines to also occur at
that time without any difficulty.
Clearly though, software for American CPC6128s which has much to do at frame flyback
time, or which wishes to avoid flickering effects on the top few lines of the screen, must
arrange to place the correct NTSC value into register 5 of the 6845 using code equivalent to
the following:
SET_NTSC LD B,#F5
IN A,(C) ; Read PPI port B
AND #10 ; Inspect LK4
RET NZ ; Return if not NTSC: No action required
DI ; Need exclusive access to CRTC
LD BC.#BC05
OUT (C),C ; Set CRTC address to register 5
LD BC.#BD04
OUT (C),C ; Set Vertical Total Adjust to 4
EI ; End of exclusive access
RET"


I tells its a hardware bug anyway. But not is the rom that program the crtc at start up?.

[Bryce]

So Amstrad was mis-using the PAL / NTSC terms when they actually meant 50/60Hz systems.

The ROM doesn't program anything, it's just storage for code. The CPU reads the ROM and sets the CRTC up. According to the text, they chose to use a CRTC where the frame flyback pulse length couldn't be changed so that the setup sequence wouldn't change it to a value that could cause issues on a 60Hz machine.

Bryce.

[pelrun]

So Amstrad was mis-using the PAL / NTSC terms when they actually meant 50/60Hz systems.

Yes, it was super common. Most laypeople didn't care what "PAL/NTSC/SECAM" actually meant, just that their region only ever used one of them. All NTSC regions got one hardware rev of a product, and all PAL regions got another. So the terms became synonymous with the region in general usage.

[arnoldemu]

Am I understanding this correctly?:

1. in the US, the monitor would use 110V AC at 60Hz.
2. the monitor uses this for it's native frame rate to control the vertical deflection.
3. the cpc has a 50/60hz pcb link. Setting this causes the firmware to send different data for the CRTC to match the native monitor display
4. There is a bug in the firmware - it used wrong value for CRTC R5 for 60hz, they worked around it by installing a CRT with a fixed vertical sync (UM6845R, MC6845).

Questions:
- The CPC monitor schematics show 220/240V AC input, what would need to be changed inside the monitor for 110VAC?
- Is the horizontal deflection rate in the monitor the same for 50 and 60hz power supplies?
- Is the horizontal deflection rate in the US CPC monitor the same rate as used in NTSC?

I wonder why the schematics in the service manuals don't show details for US monitors? Or can they be driven by 110V without breaking?

[Bryce]

Questions:
- The CPC monitor schematics show 220/240V AC input, what would need to be changed inside the monitor for 110VAC?
- Is the horizontal deflection rate in the monitor the same for 50 and 60hz power supplies?
- Is the horizontal deflection rate in the US CPC monitor the same rate as used in NTSC?

I wonder why the schematics in the service manuals don't show details for US monitors? Or can they be driven by 110V without breaking?

Answers:
- Just the transformer and the fuse, which is a seperate plugable module anyway. No further changes required as the rest of the
  circuitry only depends on the output voltages of the transformer which would be the same. The frequency of the supply is
  irrelevant as the circuitry also only gets the rectified DC voltage.
- The CTM doesn't use the power line frequency to determine the flyback frequency, this is done with the CSync signal which
  is determined by the CRTC settings.
- That depends on what CRTC settings are made when the 60Hz jumper is set. So you tell me, what does it send to the CRTC?

As far as using 50 or 60Hz in the US is concerned, the CPC/CTM combination is an isolated system, so it would work fine at both
50 or 60Hz refresh rates. It would only make a difference if you were using an NTSC modulator with the CPC, then the CSync
would need to be 60Hz.

Bryce.

[arnoldemu]

These are the differences:

;; CRTC data for 50Hz display 05c5 defb &3f, &28, &2e, &8e, &26, &00, &19, &1e, &00, &07, &00,&00,&30,&00,&c0,&00 
;; CRTC data for 60Hz display 05d5 defb &3f, &28, &2e, &8e, &1f, &06, &19, &1b, &00, &07, &00,&00,&30,&00,&c0,&00

* no difference in horizontal line length, horizontal displayed, hsync position or hsync length. 
This indicates a 15.625kHz line length (compared to NTSC TV which is 15.734 and PAL TV which is 15.625 kHz).
So it's using PAL line frequency.

* it's using 262 lines per frame compared to 312 for UK CPC.

* The VSYNC position has also been adjusted so the picture is centralised in the smaller frame.

So if the same monitor was used would they have turned the vhold to compensate and adjusted the vertical size to expand it over the tube vertically?

 

[Bryce]

So if the same monitor was used would they have turned the vhold to compensate and adjusted the vertical size to expand it over the tube vertically?

Yes, they'd have to, otherwise they'd have a 16:10 widescreen type of picture.

I assume they did this with the intention of bringing out a US TV modulator, which obviously never happened.

Bryce.

[00WReX]

Hi,

You obviously have one of the newer machines, but they didn't stop production of the ORIGINAL CPC464 until January or February 91. 

- How do I know this You might ask ? ? ?

The case is simply that I have TWO of the latest batches of machines. They have labels with 90-12 and 91-01.
These machines came from Weeske in Germany (they were official dealers in those days). Both have matching serial numbers on both the boxes and machines.
What makes these so special is that they seems to be built of "leftover" parts.

You could have a look at the following post in this thread over 7 years ago.

http://www.cpcwiki.eu/forum/amstrad-cpc-hardware/cpc-mainboard-versions/msg7334/#msg7334

Yours

Terje Grind
NORWAY

Hi All,
Sorry to jump back to this post, but while having a look over ebay, I stumbled across this 464 for sale in France and it has one of those serial numbers that Terje mentions.
The auction is close to ending, but will still be visible via the link for a while.
I will also copy one of the pictures here for reference...

https://www.ebay.fr/itm/Amstrad-CPC464/273064766562?hash=item3f93edac62:g:8m8AAOSwu6tafLMI

[attach=1,msg156127]

Cheers,
Shane

[leZone]

00WReX : intestesting this late 464 version ! (it's an english computer version on your link)

MC6845 or UM6845R.

My CPC 6128 Indescomp to the USA market have a : MC6845P


On european standard CTM-644-2 monitor (220v/50Hz), it works with the right colors (so PAL i mean, i dont think CTM is able to analyse NTSC color mode), and with 50Hz frequency (no need to change the vhold)

All photos here : https://acpc.me/#ACME/DOCS_TECHNIQUES/DISSECTION/USA/CPC6128/550903356_Indescomp_USA


Sorry for mother board, i'm not able to make better photos of it, because the metal shield is soldered on the board and i don't want to destroy the original state of this piece of history

I never see the "usa monitor", is some one have it ? Does it exist ?

[dragon]

Just few discoverys.


Zpto  Is a incremental number in all products in amstrad. Every little pcb of cpc pcw pc ppc etc.


They assigned it in the order they developed it. Every discover + 1.


It begins  in cpc with z70100. When it reach z70999->it change to z80000.


So when amstrad go into 90 years. They arrive to z90999. And they got out of numbers!. So they change the system to these new of the plus computers.


The number mc in the boards follow the same line of ptno. Are every board developed in the order they developed it. Cpc-pcw-pc-pcw etc....


And in 664 and 6128(except cost down). Cpcs have 3 consecutive numbers pcb+led pcb+volume pcb. 

[VincentGR]

This is a cost down 6128 (MC0100A) as we have on the wiki but, there are some chips on the right on different positions.


http://www.cpcwiki.eu/index.php/File:CPC6128_PCB_Top_(Z80330_MC0100A).jpg

[gerald]

This is a cost down 6128 (MC0100A) as we have on the wiki but, there are some chips on the right on different positions.


http://www.cpcwiki.eu/index.php/File:CPC6128_PCB_Top_(Z80330_MC0100A).jpg

The board on the wiki is a Frankenstein one.
It has memory in both possible chip packages
- 2 banks of 2 IC of 64kx4 -> the 4 chips missing on your photo)
- 2 banks of 8 IC of 64kx1 -> the 2 column of 8 chips on the right side of the board
Only one of each bank need to be assembled, and most (if not all) board only used the 64kx1 arrangement (i've seen a 464 cost down using 64kx4 ram through)

[robcfg]

@VincentGR , if you don't mind, I'll add a link your picture in the Mainboard versions page, for completeness.

[dragon]

@VincentGR , if you don't mind, I'll add a link your picture in the Mainboard versions page, for completeness.

Please add the plus prototype too.