What's new
A

Apocalypse

Recently I found motivation to have a go at repairing CPS1 hardware. As most of you know, the most common failure are the custom chips, only 2 on the system: the CPS-A-01 on the A-board and the CPS-B-xx on the C-board.
I think I had the discussion before with several members about how I make sure a chip is bad without having to resolder it on a known working board. Answer is below:

20181123-121751.jpg


FYI ALL the chips I've tested during this recent repair attempt were faulty...
 
So you have a known working A board in which you've installed a socket, drop in the suspected chip and verify it straight away.

Only problem is there's no replacement for the customs, not that I know of anyhow. I do recall a stash of NOS customs being sold on Yahoo a few years ago, can't remember the details though.
 
So you have a known working A board in which you've installed a socket, drop in the suspected chip and verify it straight away.
Exactly that.
Only problem is there's no replacement for the customs, not that I know of anyhow. I do recall a stash of NOS customs being sold on Yahoo a few years ago, can't remember the details though.
You can find one CPS-A-01 and one CPS-B-21 chip in every CPS2 A-board.
But definitely if there's one custom chip to be reproduced it's the CPS-A-01, I know some argue about modern FPGA chips incompatible pinout but it's easy to design a flexible PCB to convert whatever the FPGA pinout is to QFP160.
 
That's a lot of faulty chips! :D

Cool to see the test rig, looks very clean!

Totally agree on the need for a replacement. So many boards are dead due to this single issue, and fear for your future CPS1 multi setups, people, because they will die to it too...
 
So you have a known working A board in which you've installed a socket, drop in the suspected chip and verify it straight away.
Exactly that.
Only problem is there's no replacement for the customs, not that I know of anyhow. I do recall a stash of NOS customs being sold on Yahoo a few years ago, can't remember the details though.
You can find one CPS-A-01 and one CPS-B-21 chip in every CPS2 A-board.But definitely if there's one custom chip to be reproduced it's the CPS-A-01, I know some argue about modern FPGA chips incompatible pinout but it's easy to design a flexible PCB to convert whatever the FPGA pinout is to QFP160.
have there been any developments for a replacement FPGA chip yet? As I'm sure you are aware Jotego has reverse engineered the CPS1 in all its glory. This could in theory bring back to life many CPS1 a boards.
 
107091736_724874905014658_3436302258691629607_n.jpg
Would something like this programed to simulate the CPS-A-01 custom chip work?
 
So you have a known working A board in which you've installed a socket, drop in the suspected chip and verify it straight away.
Exactly that.
Only problem is there's no replacement for the customs, not that I know of anyhow. I do recall a stash of NOS customs being sold on Yahoo a few years ago, can't remember the details though.
You can find one CPS-A-01 and one CPS-B-21 chip in every CPS2 A-board.But definitely if there's one custom chip to be reproduced it's the CPS-A-01, I know some argue about modern FPGA chips incompatible pinout but it's easy to design a flexible PCB to convert whatever the FPGA pinout is to QFP160.
have there been any developments for a replacement FPGA chip yet? As I'm sure you are aware Jotego has reverse engineered the CPS1 in all its glory. This could in theory bring back to life many CPS1 a boards.
I don't think @jotego did a total reverse engineering, but a RE-engineering of the system.
 
So you have a known working A board in which you've installed a socket, drop in the suspected chip and verify it straight away.
Exactly that.
Only problem is there's no replacement for the customs, not that I know of anyhow. I do recall a stash of NOS customs being sold on Yahoo a few years ago, can't remember the details though.
You can find one CPS-A-01 and one CPS-B-21 chip in every CPS2 A-board.But definitely if there's one custom chip to be reproduced it's the CPS-A-01, I know some argue about modern FPGA chips incompatible pinout but it's easy to design a flexible PCB to convert whatever the FPGA pinout is to QFP160.
have there been any developments for a replacement FPGA chip yet? As I'm sure you are aware Jotego has reverse engineered the CPS1 in all its glory. This could in theory bring back to life many CPS1 a boards.
I don't think @jotego did a total reverse engineering, but a RE-engineering of the system.
Well he certainly did reverse engineer the CPS-A-01 and how the CPS1 board works to begin with or us very proud MiSTer owners would not have the core we have today. I have spoken with Jose and he claims it is possible, some adjustments would need to be made to his source code as his example uses MiSTer memory DE10-Nano on board DDR3 and not CPS1 memory and while the end result may be exactly the same, they are very different memory.
 
How has he reversed the CPS-A-01? Decapping? That's the only way to get a faithful reproduction at gate level.
Having it RE as a black box isn't helpful.
 
@Apocalypse, just curious which socket did you install on the board for testing the CPS-A-01 chips?
 
If somebody was to create a reproduction of the CPS-A-01 they would get a lot of thanks from CPS1 owners and, as a bonus, possibly make a nice profit; there are a lot of CPS1 A boards out there with a faulty CPS-A-01, all waiting for the day when they can be resurrected (I have about a dozen here for example).

@'Apocalypse - if you ever recall any information of the socket you used for testing please post it here, think I'll have a hunt around online too, maybe I can track it down. Been meaning to test the CPS-A-01 chips for some time now.

Edit: not found one yet but generally speaking QFP sockets aren't exactly cheap .........

Edit2: still not found one ..........
 
Last edited:
What are the symptoms of a dead CPS-A-01? I have a Final Fight with working B/C boards (tested on a working A board), but on the dead A board, I have black screen but can hear audio.
 
If somebody was to create a reproduction of the CPS-A-01 they would get a lot of thanks from CPS1 owners and, as a bonus, possibly make a nice profit
This has been discussed here before, I think the general consensus was that the costs of R&D (decapping, engineering, etc.) and then manufacturing the chip would far outweigh any incoming money.

Unfortunately the things that we think as a community would be big money spinners, really aren't. However I'd definitely buy a handful of CPS-A-01 chips if they ever were repro'd. It's a pity the deadstock that was around dried up (did anyone here get any?).
 
What are the symptoms of a dead CPS-A-01? I have a Final Fight with working B/C boards (tested on a working A board), but on the dead A board, I have black screen but can hear audio.
It's extremely variable - I've seen everything from a black screen to the image shrunk and multiplied on the display (therefore four small images of the game, one in each corner). Also remember that other faults can also cause display problems, invariably one of the RAMs or their support chips develops a problem, even the PALs can fail (although in my experience that's rare).
 
If somebody was to create a reproduction of the CPS-A-01 they would get a lot of thanks from CPS1 owners and, as a bonus, possibly make a nice profit
This has been discussed here before, I think the general consensus was that the costs of R&D (decapping, engineering, etc.) and then manufacturing the chip would far outweigh any incoming money.
Unfortunately the things that we think as a community would be big money spinners, really aren't. However I'd definitely buy a handful of CPS-A-01 chips if they ever were repro'd. It's a pity the deadstock that was around dried up (did anyone here get any?).
It's a pity that it's so expensive to make a modern reproduction of the CPS-A-01 - more and more A boards are being consigned to the parts bin because of A-01 failures.

I think that all of those NOS A-01 chips dried up, or am I thinking of the CPS-B-21 chips (which were listed as DL-0921)?
 
Good thing they’re on “SPECAIL” for just $108.33 :(
 
It could also mean they’ll actively track them down or harvest them from boards if they know someone will pay?
 
Back
Top