The Sinclair QL Forum

Derek_Stewart wrote:Hi,
Here is a copy of the CST QDISK v1.18 jumbled.
CSTQD118.zip

Thank you so much.
I will try to complete a clone of this interface as a summer project, I will keep you informed of the progress

anyone managed to get 1.19 firmware ??

So finally, thanks to the pictures of Derek_Stewart and techy, I was able to complete the capture of the schematic of the V4 board.

I ordered a pair of PCB, and construct one interface, and it's working!!! (with Version 1.18 of the Rom).

I put all the information here: https://github.com/alvaroalea/QL_CST_FL ... troller_V4

I only have one BIG PROBLEM, when I connect it with my RAM expansion, the interface KILL (completely) the GAL chip of the ram expansion ( I burn 4 expansion ), if someone want to made a copy of it, be aware (or if you are using the internal expansion from chr$ )

I double check the schematic, and I can't found what is the problem, the schematic is very similar to previous version, that work perfectly with V3 version (the one from nikosdoykas ), so I'm out of ideas, if some one can take a look, I will apreciate.

One thing that looks strange is that RW is connected to PGM pin of the eprom, I think that PGM since eprom is read only in the circuit should have been connected to 5V not to RW.

lliont wrote: ↑Sat Jun 17, 2023 8:37 am One thing that looks strange is that RW is connected to PGM pin of the eprom, I think that PGM since eprom is read only in the circuit should have been connected to 5V not to RW.

Yes, I already notice it, But this track is clear in all the pictures I got of the interface, and in this board I try to make it 1:1 This is not clear to me, perhaps CST use some kind of flash eeprom that allow direct update during development?

In any case, I use a W27C512, so the pin 27 is A14, and I mirror the 16Kb ROM 4 times, so this shall not affect, neither the reading of the ROM, or the GAL in the other interface.

Hi!

Out of interest, what is the RAM expansion who's GAL keeps getting blown by your CST remake?

Perhaps the fault lies in the RAM expansion design, rather than your disk interface...

In my experience, GALs are pretty hardy devices and aside from enabling DTACK and DSMC, a RAM GALs outputs are usually all facing in towards the RAM itself, rather than out towards the shared bus and the disk interface.

The GALs inputs on the other hand are typically all CPU driven signals, such as the address lines, which your CST interface also only reads/sinks current, rather than driving.

Anyhow, might be worth extending the scope of your investigation beyond the disk interface.

The RAM expansion is this one:

https://github.com/alvaroalea/QL_512kb_ ... _Expansion

The circuit is the same you can found in other SRAM interfaces, the only diference are two small jumper to allow daisy chain to get the 896kB maximun of the bare QL, In any case I test with only 512kB, because the Q-Disc use the 74LS266 with the SP0-SP3, more is not possible.

The expansion has been tested with clones of Q-Disc V3, Trump Disc (0kB), QSound and Qbide without problem (I think on reflections, but I'm able to configure conbinations of this ones, very large, without it)

Yes, DTACK and DSMC, was my main target, I'm also check the 74LS03 of the Qdisk, thinking perhaps is a counterfeit 74LS03 and do not has open-colector outputs, and cause a short circuit in the DTACK line, but passed the test of High-Impedance as expected.

Finally I found the problem, I used a black 2N2222 from aliexpress, because it's similar to the original, after reading the Qsound thread, I replaced the transistor with a genuine from mouser and voila, the interface work well with other cards.

So, it's finish, and now we have another disk interface well documented if anyone want to build it.

the q disk can take more than 512k ram.... sp3 removed an go hi in address .E0000H

https://www.qlforum.co.uk/viewtopic.php?t=3422&start=30