QIMI hardware question(s)

Nagging hardware related question? Post here!
Post Reply
prime
Trump Card
Posts: 192
Joined: Fri Feb 18, 2011 8:58 pm

QIMI hardware question(s)

Post by prime »

Hi all,

I have a couple of questions about the published QIMI schematic :

The ZX8302 enabe signal PCENL is gated by an LS32 OR gate so that it doesn't go low if A10 is high so that the ZX8302 is not enabled and the mouse ports can be read. But won't this also disable the generation of DSMCL to signal the end of the data transfer ?

What is the function of the LS32 gating the CPU clock onto the RAW1 and RAW2 lines which seem to be something to do with microdrive data ?

Cheers.

Phill.


User avatar
tofro
Font of All Knowledge
Posts: 3132
Joined: Sun Feb 13, 2011 10:53 pm
Location: SW Germany

Re: QIMI hardware question(s)

Post by tofro »

Phill,

which schematic are you looking at? The one I have doesn't have RAW1 and RAW2 connected, which would be kind of nonsense as you rightly say. Maybe you confused that with RESI and RESO, which is reset circuitry - And QIMI does loop in here to help the clock survive a RESET.

Tobias


ʎɐqǝ ɯoɹɟ ǝq oʇ ƃuᴉoƃ ʇou sᴉ pɹɐoqʎǝʞ ʇxǝu ʎɯ 'ɹɐǝp ɥO
prime
Trump Card
Posts: 192
Joined: Fri Feb 18, 2011 8:58 pm

Re: QIMI hardware question(s)

Post by prime »

tofro wrote:Phill,

which schematic are you looking at? The one I have doesn't have RAW1 and RAW2 connected, which would be kind of nonsense as you rightly say. Maybe you confused that with RESI and RESO, which is reset circuitry - And QIMI does loop in here to help the clock survive a RESET.
This is the schematic on Dilwyn's site : http://www.dilwyn.me.uk/docs/manuals/qimi.png

Do you have a link to the schematic you have ?

Cheers.

Phill.


User avatar
Dave
SandySuperQDave
Posts: 2808
Joined: Sat Jan 22, 2011 6:52 am
Location: Austin, TX
Contact:

Re: QIMI hardware question(s)

Post by Dave »

This is of interest to me as I would be reproducing this circuit on UltraQ. If anyone would like to contribute an improved circuit, I will happily incorporate it so many can benefit.


prime
Trump Card
Posts: 192
Joined: Fri Feb 18, 2011 8:58 pm

Re: QIMI hardware question(s)

Post by prime »

The lots of logic is why I've used a CPLD in my design as it reduces all the ls packages down to a single surface mounted PLCC 44 pin package (in this version, next version will have a VQFP44, which is only 10mmx10mm).

So the design is essentially 2 extra chips XC9536XL CPLD and ATmega8 (or mega88, mega168) for PS/2 to quadrature conversion & an LE33 3.3V reg for the CPLD.

I can't of course do anything about the space taken up by the ZX8302 & the ROM.

Back to the original question, there seem to be multiple versions of the QIMI schematic floating around does anyone have links for them, as I only have the one from Dilwyn's site.

Cheers.

Phill.


prime
Trump Card
Posts: 192
Joined: Fri Feb 18, 2011 8:58 pm

Re: QIMI hardware question(s)

Post by prime »

Brane2 wrote:Isn't this totally roundabout way of doing this ?

You will convert PS2 to quad and then use that crappy circuit to have CPU interrupted for every step ?

With a HW that could do much more ?
It's the COMPATIBLE way of doing things. Yes there are better ways of doing it, but they require things to be re-written, things that have been standard for a number of years (prolly about 20ish), so the way that required the least things to be changed and therefore less likely to break anything. It also has the advantage that (at least in theory, the current circuit doesn't have the connection points), that it can support a PS/2 mouse or an Atari / Amiga quadrature mouse.

Polite request, can we move the thread back to what I was asking, about how the current QIMI works rather than changes that could be made to it.

Cheers.

Phill.


prime
Trump Card
Posts: 192
Joined: Fri Feb 18, 2011 8:58 pm

Re: QIMI hardware question(s)

Post by prime »

prime wrote:Hi all,
The ZX8302 enabe signal PCENL is gated by an LS32 OR gate so that it doesn't go low if A10 is high so that the ZX8302 is not enabled and the mouse ports can be read. But won't this also disable the generation of DSMCL to signal the end of the data transfer ?
Apparently the Service guide answers this, the DTACK is generated by the ZX8301 for internal accesses so the read will still work as the 8301 will generate the needed signal for the QIMI's address range.

Still interested in seeing any other schematics other than the one previously linked tho.

Cheers.

Phill.


User avatar
tofro
Font of All Knowledge
Posts: 3132
Joined: Sun Feb 13, 2011 10:53 pm
Location: SW Germany

Re: QIMI hardware question(s)

Post by tofro »

Phill,

see here:
http://forum.tlienhard.com/phpBB3/viewt ... imi#p12468

That is a QIMI replacement built by a member of the German ZX forum.
He's round here as well every now and then.

Regards,
Tobias


ʎɐqǝ ɯoɹɟ ǝq oʇ ƃuᴉoƃ ʇou sᴉ pɹɐoqʎǝʞ ʇxǝu ʎɯ 'ɹɐǝp ɥO
QLObi
Chuggy Microdrive
Posts: 50
Joined: Wed Mar 20, 2013 9:15 pm
Location: Germany Köln

Re: QIMI hardware question(s)

Post by QLObi »

Hi all,
why you invent the wheel 2 times. There is already a valid schematic and layout around.
http://forum.tlienhard.com/phpBB3/viewt ... 271#p12271
Just use google translate, if you do not understand german. The schematic and the layout is in open eagle format.
There is already a PS/2 adapter integrated. In opposition to the version already published here (miCe thread) you need version 1.04, because Qimi likes the Atari(Amiga) signals. The basocm source code is also open.
By the way, the strange curcuit with the clock is the replacement for the popular resistor mod on top of the ZX8302 (2 resistors to -12V). In this case it is just -5V and what it helps? I don't know !


QLobi

If i find some time i will make an english project folder of it and publish here.


Post Reply