Please Check legacy ICs before use .

Post Reply
lezanderson
Posts: 174
Joined: 14 Aug 2012 15:31

Please Check legacy ICs before use .

Post by lezanderson » 24 Nov 2014 15:27

Hi.

Some of you may of received legacy ICs from me, Z80s, V9958 etc etc . These are untested.. so best test them and verify them .. don't just take it that they will work ..

Dave has tested some Z80s and found most where Z80Hs 8MHz type or so we think .. a few where DUFF and only a small number true 20MHz Z84C0020s ??

Hope this helps clear things up ?

User avatar
Dave
Posts: 747
Joined: 11 Aug 2012 18:16
Contact:

Re: Please Check legacy ICs before use .

Post by Dave » 26 Nov 2014 03:27

Hi folks,

some background to the tests that Martin and I ran on the "Z84C0020PEC"s when we started to question whether the chips were really 20MHz Z80s.

As described on my counterfeit chips page :-http://www.primrosebank.net/computers/m ... _fakes.htm, I did a bit of "Googling" and found a method of determining programmatically whether a Z80 is an NMOS or CMOS version.

On an NMOS Z80, an OUT (C),0 instruction outputs 0, on a CMOS Z80, OUT (C),0 outputs FFh. Martin used this information to run tests on a number of Z80 CPUs that he had to hand, a mixture of various NMOS, CMOS and "questionable" types. I tested number of the "Z84C0020PEC"s as well as a couple of NMOS Z80s that I had.

In short, very few of the "Z84C0020PEC"s appear to be 20MHz CMOS.

Knowing how to test them may save you a bit of wasted time in the future should you use one of these chips and run into problems. As Lez has advised, it would be better to test them before use,

regards
Dave

User avatar
1024MAK
Posts: 556
Joined: 24 Dec 2012 03:01
Location: Looking forward to summer, in Somerset, UK

Re: Please Check legacy ICs before use .

Post by 1024MAK » 26 Nov 2014 03:33

Yes, the undocumented OUT (C),0 has been known about for a while. This and this are two documents that give details :mrgreen:

You may also want to have a read of this PDF: Z80 Family Questions and Answers

Mark

lezanderson
Posts: 174
Joined: 14 Aug 2012 15:31

Re: Please Check legacy ICs before use .

Post by lezanderson » 27 Nov 2014 12:19

If anyone else finds some incorrectly marked ICs or DUFF ICs .. please leave details on the Forum for others to read and be aware of..

I've come across CPLDs that have been pre-programmed.. This reduces the number of I/O lines by all accounts.. ?? Though I don't have full details about this.

Found a DUFF AM9511 maths co-processor.. someone in ITALY tested it and found not working.

MC68008P10 was discovered to not Work and got Rather HOT !! I believe this was tested in a Siclair QL.

User avatar
Dave
Posts: 747
Joined: 11 Aug 2012 18:16
Contact:

Re: Please Check legacy ICs before use .

Post by Dave » 27 Nov 2014 14:45

Hi Lez,

Thanks a lot for the info.

The number of I/O lines should not be reduced by programming one of the EPM7128s, unless perhaps the number of writes has exceeded the design limit (100), I guess anything could happen then, but, without more details, it sounds more like a case of "pilot error".

Regards
Dave

Martin A
Posts: 409
Joined: 09 Nov 2013 21:03

Re: Please Check legacy ICs before use .

Post by Martin A » 27 Nov 2014 21:17

I kind of skipped over the issues with the re-marked Z80's when I posted the tread about the High speed Z80 board.

In a way it's been useful as I've found just how over-clockable the real CMOS Z80's are. A rock steady 32mhz operation from a 10Mhz CPU, and 16Mhz operation from an 8Mhz CTC.

The Z80's marked as Z84C0020PEC's might not be the CMOS parts they claim, but the are Z80's and do run at the speeds you would expect of NMOS parts and will come in useful when I finally get round to repairing the ZX81.

I've used a number of the other chips Lez supplied, and no problems with any of them, they all react exactly as the labels would suggest..,

SO FAR it's only be the "high speed" Z80's that have been a problem.

Now I know some things to look for, I checked the as yet un-used chips. One of the DS12887 chips has Philippines printed on the top, and China moulded into the bottom, so may be a bit suspect, but it's passed every test I've thrown at it.

None of the other parts look like they've been sanded and re-marked.

Post Reply