Page 9 of 13

Re: I've borked it

Posted: Mon Feb 19, 2024 5:03 pm
by msknight
OK - here's some interesting stuff.

I got another zeddy and de-soldered the ferranti, replacing it with a socket. Against all odds, the Ferranti seemes to have survived.

This was enough to prove the ferranti in the zeddy I was working on, had died.

However... and this is where the spanner got thrown in the works ... the vLA81 will not work with that vintage RAM chip. Or possibly that RAM chip is borked... or I've wired something wrong... or something. But it appears that having that RAM chip in, has been an extra fly in the ointment that messed up the other chip swapping.

When I've gathered myself, I'll look up that chip and see what gives.

Re: I've borked it

Posted: Mon Feb 19, 2024 5:15 pm
by Moggy
I notice from a picture of your board that you are using the 2764 EPROM wiring method for the 27256 chip I assume you are using?

On the 2764 pin 26 isn't connected internally so all's well there and pin 28 (Vcc) is held at 5v along with pin 27 (programming. ) which allows the chip to function. the 27256 has address lines A13-A14 at pins 26-27 which are now having 5 volts shoved through them, is this likely to damage the chip in any way? I ask in ignorance as I've never experienced the problems you are having but just a thought out loud as they say.

Another thought is have you repeated the ROM code throughout the chip? When using a 2764 to hold 4k ZX80 code I read somewhere that the code had to be repeated throughout the entirety of the chip in order to work.

Again just musings on my part.

Re: I've borked it

Posted: Mon Feb 19, 2024 5:26 pm
by msknight
Hi Moggy,

You are correct. A13 and A14 are basically held at +5v which is pinning them high and causing the addressing to read from the highest 8k in the EPROM. The image is, therefore, indeed, repeated four times through the EPROM.

My problem appears to be the RAM chip hm65256ap-15 for which I had trouble getting hold of a datasheet and indeed, the only thing I could get on it was in a Hitachi IC Memories Data Book, but that chip doesn't seem to be working with the vLA81 and I'm trying to get hold of another data sheet to confirm, before I try another memory chip.

Michelle.

Re: I've borked it

Posted: Mon Feb 19, 2024 5:40 pm
by msknight
OK - second chip of the same lot failed. Either I got ripped off with a bad batch of RAM chips, or something about these chips is not suitable for the zeddy, or else not suitable for the vLA81.

But I'm having a real problem finding a data sheet.

Mark? Help!

Re: I've borked it

Posted: Mon Feb 19, 2024 5:40 pm
by Moggy
There seems to be plenty of data sheets for that series of chips and the only difference I can see is access times.

Re: I've borked it

Posted: Mon Feb 19, 2024 5:46 pm
by msknight
Can you link me to one please.

Re: I've borked it

Posted: Mon Feb 19, 2024 5:48 pm
by Moggy

Re: I've borked it

Posted: Mon Feb 19, 2024 5:50 pm
by msknight
Oh, and on that chip, pin 1 is the program chip. Sent to a higher voltage than +5 for programming, but once it's programmed, that's it.

I suppose that makes it a PROM rather than an EPROM ... or rather an E-PROM... Electrically Programmable ROM rather than an Erasable Programmable... hmmm... that would make sense. It's the first E in EEPROM rather than the second E.

Damn this gets confusing!

Re: I've borked it

Posted: Mon Feb 19, 2024 5:52 pm
by Moggy
No just a PROM I would have thought as it's a one shot device.

I do believe ( no doubt incorrectly) that once programmed some connecting links inside the chip are blown making it read only, possibly just a factory thing rather than home programming.

Re: I've borked it

Posted: Mon Feb 19, 2024 6:54 pm
by 1024MAK
My fear is that the original Ferranti ULA and both vLA81 modules have had their pin 12 outputs damaged some how.

With R2, all the above should have no difficulty producing a logic high voltage on their pin 12. As this is a logic level output, it should either be driven high, or when the Z80 is trying to access the RAM, should go to a logic low.

The test with the resistor showed that there isn't a short of the /RAMCS line to 0V/GND.

Mark