exxos's Atari, Amiga & retro forum

Fit the keyboard with/in the original case?

Yes... and possibly the top metal shield too... at least with a bit of a cut-out.

Should be able to fit a couple of external switches to switch between CPUs and ROM images (or have a 512K ROM potentially).

Could someone look at my GAL code (posted earlier in the thread). There's not a lot else in the ROM decode. The 74LS244 is generating DTACK using the output from the GAL (ROM /CE). I borrowed this idea from another TOS 2.06 decoder on the 'net.

Have I got the RW line from the CPU the wrong sense? It is High for read and Low for write, correct?

At the moment the ROM /OE line is low if RW is HIGH, /AS is LOW, A18, A19 and A20 are LOW and A21, A22 and A23 are HIGH OR /ROM2 is LOW

Of course, if we want to support a 512K ROM we can ignore A18 and connect that directly to A17 on the ROM.

/CE on the ROM is connected directly to /AS.

/DTACK is generated from the ROM /OE line via a 74LS244 with the ROM /OE line controlling the output gate and the inputs connected to ground, so if ROM /OE goes low the 244 drives /DTACK low.

Found the problem! D-Oh!

The output pin should be 19 and not 13! *STUPID!*

Here are the new PLD and JED files.

P.S. This is why I like GALs... not bodge wires needed to fix the circuitry! ;-)

stephen_usher wrote: 10 Oct 2020 20:32 Found the problem! D-Oh!

The output pin should be 19 and not 13! *STUPID!*

Here are the new PLD and JED files.

P.S. This is why I like GALs... not bodge wires needed to fix the circuitry! ;-)

Well done, that's fixed it, Rom in your relocator and boots to Tos2.06 on both CPU sockets :D

And just for a bit of fun, here's what it looks like in the H4

PhilC wrote: 10 Oct 2020 21:26

stephen_usher wrote: 10 Oct 2020 20:32 Found the problem! D-Oh!

The output pin should be 19 and not 13! *STUPID!*

Here are the new PLD and JED files.

P.S. This is why I like GALs... not bodge wires needed to fix the circuitry! ;-)

Well done, that's fixed it, Rom in your relocator and boots to Tos2.06 on both CPU sockets :D

Yay! :-)

If there's a Mk.ia I can add a jumper for switching between 2x 256K ROM mode and 512K ROM mode. That'll be very simple. (Oh and fix the writing for the CPU selector!!!)

:yay: :dualthumbup: :yay2: :thumbup: :girldance:

Sounds like a plan. Just burning EmuTOS to a spare eprom to give it a proper test now too.

All it'll need is the GAL code ignoring A18, leave the jumper open and a bodge wire underneath from A18 to A17 pin of the ROM socket and Bob's your aunty.

Ok, so a further update.

Using my rock solid 128mb TF536 all is not quite as it should be.

At first gembench would crash without doing a single loop. So ive now connected up the 5v supply directly to the TF536 and its a lot more stable and will fo Gembench and the Intdiv test just fine but Frontier will crash within a second of the intro starting every time.

Its worth noting that this is on my H4 and the TF536 has a 1k pullup on the 8mhz line and a 100pf cap between ground and 8mhz. I tried it on a TF536 without the pullup etc and its not as reliable so far.