Page 7 of 8

Re: Oh no! Not another ULA.

Posted: Tue Sep 26, 2023 6:30 pm
by Andy Rea
1024MAK wrote: Mon Sep 25, 2023 11:54 pm Would now be a bad time to suggest 6X :lol:

Mark
I've got one of those.... the 20Mhz Zeddy... normal speed is slightly fast at 3.33Mhz

It is anything but a standard zeddy, ask Moggy...

Re: Oh no! Not another ULA.

Posted: Tue Sep 26, 2023 6:36 pm
by Andy Rea
Moggy wrote: Tue Sep 26, 2023 1:34 pm @Andy

You say 70ns ROM for Big Bang, I assume ,after seeing your video that, this is a custom chip?
its just a at28hc64b-70su eeprom but SOIC on a SOIC to DIP adapter board.

a regular DIP eeprom of the same speed should work..

slower chips should work but because all that stupid resistors bridges that Sinclair used to allow overriding data and address lines make a right bloody mess of the signal integrity. teh uber ( 20 Mhz ) zeddy splits the data and address busses up so no overriding signals

Andy

Re: Oh no! Not another ULA.

Posted: Tue Sep 26, 2023 7:39 pm
by Moggy
Thanks for that Andy but as I only have the Spectrum EPROM burner for the old 21/25v stuff then programming an EEPROM is a bit of an obstacle for a dunce such as myself sadly. :oops:

As for the 20mhz jobby then yes I can confirm that it's a bit of a super beast! :lol:

Re: Oh no! Not another ULA.

Posted: Tue Oct 03, 2023 6:11 pm
by Andy Rea
A question for the technical minded...

does anybody know of any external device that uses the ROMCS signal at the edge connector in any way other than to override the internal ROM?

reason I ask, with fasr enough cpu, ram and rom in the zx81 13Mhz is possible. ( without zxpand connected )

with zxpand+ (mine has winbond 27c512-45) plenty fast enough 13mhz is possible.

now do a config "x" and double speed works but quad speed crashes.

I want to remove the romcs resistor, and put a logic gate instead, this should result in a cleaner signal at the internal rom.

regards Andy

Re: Oh no! Not another ULA.

Posted: Tue Oct 03, 2023 9:34 pm
by Paul
I am not Shure how the memotech devices work.
Was it the centronics or the hrg module that replaces part of the ROM with the internal 1k RAM?
And how does it do it?
Apart from that I only know of simple disable ROM

Re: Oh no! Not another ULA.

Posted: Tue Oct 03, 2023 9:57 pm
by Moggy
The HRG unit uses the internal 1k RAM, don't think the centronics one does though.

Re: Oh no! Not another ULA.

Posted: Tue Oct 03, 2023 10:38 pm
by 1024MAK
There’s four ways that /ROMCS or /RAMCS is used.

Firstly, a wire link to +5V to prevent the internal ROM or RAM being used. Common with RAM packs, but not many devices do this for ROMs.

Secondly, use of a transistor to do the switching, as in switching the +5V supply to /ROMCS or /RAMCS.

Thirdly, driving /ROMCS or /RAMCS via a diode which in turn is driven via a logic chip.

Fourthly, just using a logic chip (this completely overrides the internal address decoding).

The last three are used mostly when overriding the partial address decoding to add either ROM, RAM or IO devices into the Z80s memory map.

Mark

Re: Oh no! Not another ULA.

Posted: Wed Oct 04, 2023 4:13 pm
by Andy Rea
OK I think I was right the first time lol

if I take the romcs at the edge connector ( with a pull down resistor) to one input of an OR gate and the other input from the ula generated ROMCS and the output directly to the internal ROM.

that will work right?

Re: Oh no! Not another ULA.

Posted: Wed Oct 04, 2023 10:16 pm
by 1024MAK
You mean either side of R28…
One side is connected to the ULA.
The other side is connected to the edge-connector and (on a standard machine) to the ROM.

But I presume on a modified machine, the output of the OR gate will drive the ROM instead.

Because the OR gate has one input directly connected to the ‘ULA’, the thinking is that it will switch fast enough?

Why do you think you need a pull-down resistor? A high value would not make much difference as if I understand you correctly, it’s still getting a signal via R28 (680Ω), admittedly very slightly delayed due to stray capacitance. But a low value would draw rather a lot of current.

Mark

Re: Oh no! Not another ULA.

Posted: Thu Oct 05, 2023 6:16 am
by Paul
My guess is he wants to short R28