CPS8256 clone, design files for KiCad

[JonB]

Hello

Following a recent thread ("You pay for the postage") where I promised to make available the design files for the CPS8256 clone "uSIO", here it is.

https://drive.google.com/file/d/1LNneVs87nMewAFnavz3xYghmJL2T_iLA/view?usp=sharing

The ReadMe.TXT file is reproduced below. I would love to hear how you get on with it!

Cheers
JonB



uSIO PCW v0.1 source files release - ReadMe - Dec 2024

0. Some info about what you just downloaded.
This is a collection of files that together contain a capture of the Amstrad CPS8256 serial / parallel adapter, plus a ready laid out board so you can build one yourself. Please note, it is designed to connect to the "Z80 Bus" that I used to connect uIDE-8 to the PCW.

More details here: http://www.cpcwiki.eu/index.php/UIDE_Universal_IDE_adapter_cards_for_Z-80_computers

So what this means is, you cannot connect this board to a PCW computer without some sort of adapter. There are two options, as I explain in the above link - a Z80 shim board or a PCW expansion port adapter (which presents a Z80 bus). Of these, the shim is the easiest. If you already have a uIDE board, you can connect this board directly to the uIDE ribbon connector on your PCW (that also, by design). The "Z80 Bus" is really just a pinout of the Z80 processor; the intention was to make boards that are universal (should work with most Z80 computers irrespective of their expension port pinouts. However, this CPS8256 "clone" will only work with the PCW computers (probably, only 8000-series machines), because there is no programmability built in to alter the Z80 I/O ports the card will respond to.

1. Terms of use

These files are released as-is, no guarantee of suitability for any particular purpose.
You may not redistribute these files without a copy of this ReadMe.
You will need KiCad to open the .PRO file, then the other files should be accessible. I cannot guarantee compatibility with the latest version of KiCad.
Board track layout was performed using FreeRouter.
Pay special attention to the pin configuration of the serial and parallel ports. My intention was for it to be used with a PC type backplane connector, but I never tested it. It is likely there are wiring errors here...
Should you find and repair defects, please let me know so that I can update my master copies of the files.

2. Copyright and authorship

I do not own copyright of the actual design. It is - essentially - a copy of the PCW8256 serial / parallel board, based on a schematic I found. I did alter the serial port so as to use a MAX222 transceiver chip, though. A lot of work went into capturing the schematic, checking it and producing the board layout (which I do claim as my own design), so you will please ensure you credit me as originator should you redistribute the files, and leave the logo / copyright text on the board layout file. This is only fair - if you want me to give it away for free at least have the decency to credit me as originator.

3. How to contact me.

I am member "JonB" of the CPCWiki Forum. If you need support or have any comments, please start a new thread on the forum here:

https://www.cpcwiki.eu/forum/nc100-nc200-pcw-pda600/

Note: Forum registration is required.

Cheers!
CPCWiki Member "JonB" (Jon Bradbury) 04/12/2024

[squelch41]

Thank you very much for making the design available

Just looking at the PCB - Resistor R6 doesnt have a value on the silk screen - it seems to be the only unlabelled part.
What value should it be?

[squelch41]

Thank you very much for making the design available

Just looking at the PCB - Resistor R6 doesnt have a value on the silk screen - it seems to be the only unlabelled part.
What value should it be?


Sorry, ignore - it is 1K, is visible on image here:
https://www.cpcwiki.eu/index.php/File:CPS8256_clone.png

[squelch41]

Have just had some of these built - so far, I've found the following 2 errors I think-

On U6, pins 22, 28 and 29 are connected directly to 5v
Think there are 2 incorrect traces connected to pin U6 pin 22 that explain this.

(Pin 11 and 14 on U8 arent directly connected to 5v and they are shown on the schematic as connecting to 22,28,29 on U6, and pin 6 of U6 isnt connected to 5v either)

Error 1 is the the connection of U6 22, 28 and 29 should go to R6 1K resistor looking at the schematic.
But looking at the board and the KiCad PCB editor, I think pin 22 goes to the 5v end of R3 which is also connected to the 5v end of R6.

R3 appears just to be a current limiting resistor to the power LED so the error looks like it is in the PCB not the schematic

As U6 28 and 29 connect together and then join pin 22, it looks like cutting the eroneous trace between pin 22 and R3, the bodging 22 to R6 should solve it.

Error 2 is that there is a tiny trace visible on KiCad and on the board whereby the 5v trace from U4, on it's way to C6 connects to U6 pin 22.

It looks like this trace needs cutting and then bodging U4 pin 14 to C6 100nF (the link on the board to U6 pin 22 is so small, it'd be very hard to reliably cut and not risk bridging again when soldering)

[squelch41]

Found 2 more issues also related to 5v line

C4 connects to U4 pin 2 as well as the designed pin 14
So c4 to pin 2 of U4 trace needs cutting as does the link between U4 pin 14 and U4 pin2
U4 pin 2 then needs to be bodged to U2 pin 1

Also, there is a connection between 5v from U2 pin 14 to U6 pin 28 (which is then connects to pin 29)
The link between U2 and U6 needs cutting.

[squelch41]

Conencted it up for the first time -- didnt blow up my PCW.

CP/M does see the SIO interface on boot.
But I can see anything on the TTL serial or MAX222 serial outputs using mail232
I can adjust the baud rate on SETSIO and in Mail232 and see the signals cchange on pin14 of the DART though

The power LED doesnt illiminate and the TX LED is illuminated all the time as well.

Further investigation when I have time!
(I havent tried tying the handshaking pins together yet - that will be #1 on the list)

[Ash]

Not sure if it could help, but there is this WIFI Modem Schematic connecting the PCW to the Dart, worth comparing?

[squelch41]

Designs are fairly similar- think they're both clones of the amstrad board.
Jon's one is nice in that it also has the parallel interface and RS232 serial as well as TTL

It's frustrating - just cant see why it doesnt work.
Linking the handshaking made no difference and there didnt seem to be any activity on the pins either

I'd forgotten to add 5v to R6 after cutting it's trace, so corrected that (power light now comes on)

But I am still getting no output on serial and the RX and TX LEDs are always lit.
Removing the MAX chip means the RX LED goes out by the TX is always lit.

Lifting the pin on the DART that goes to transmit shows no activity on the DART
The 8253 all seems normal though there is no ouput on pin 17 - again lifted the leg and is no output when it is out of the board either. But I do have 2x 8253 chips and both behave the same.

I probed the 74LS chips and they all seem to have outputs that look ok

I only have 1 DART but it is more likely that there is an error in the PCB somewhere rather than a fault on the DART

[squelch41]

the DART serial transmit output just seems to hold at about 2.7v which just seems odd?
I would have thought it would either be low or high?

[czarnikjak]

the DART serial transmit output just seems to hold at about 2.7v which just seems odd?
I would have thought it would either be low or high?

Perhaps it's just switching very quickly between 0 and 5v resulting in around 2.5v when you measure with multimeter? Are you able to check output with logic analyzer or oscilloscope?

[squelch41]

the DART serial transmit output just seems to hold at about 2.7v which just seems odd?
I would have thought it would either be low or high?

Perhaps it's just switching very quickly between 0 and 5v resulting in around 2.5v when you measure with multimeter? Are you able to check output with logic analyzer or oscilloscope?

yes, am using an osciloscope - its fixed at that voltage

[czarnikjak]

the DART serial transmit output just seems to hold at about 2.7v which just seems odd?
I would have thought it would either be low or high?

Perhaps it's just switching very quickly between 0 and 5v resulting in around 2.5v when you measure with multimeter? Are you able to check output with logic analyzer or oscilloscope?

yes, am using an osciloscope - its fixed at that voltage

Bummer, either the Dart is foobar or the output is shorted to something that it shouldn't I would guess.

[squelch41]

If anyone has a z80 DART, could they tell me what happens if they just apply 5v and ground to the power pins and measure pins 15 and 26?
Mine just outputs ~3.3v which seems suspicious.
I tried running an arduino sketch as well to test (albeit written by chart gpt) and again, just ~3.3v output.

[squelch41]

Tried replacing the DART, exactly the same behaviour

Think I've hit my skill limit troubleshooting

[squelch41]

FWIW, think I was going down the wrong route adjusting the traces as above - after re-looking at the https://github.com/VapourSoft/PCWWiFiModem schematics, the design (as opposed to the schematic) make sense.

So I thought I'd just build an 'as designed' pcb as I had spares - still exactly same outcome.

So, for me at least, this design doesn't work.