Troed's H4 build

[rubber_jonnie]

Yep, I did mine too

[troed]

Lightning ST added. Not really tailor made for the H4 - blocks all other expansion sockets

... now to go flash a new dual TOS since the Lightning includes a 256k decoder.

IMG_20200702_110303_2674.JPG (98.3 KiB) Viewed 1078 times

/Troed

[troed]

Hmm, actually I have issues with the Lightning (or the expansion slot). While I've successfully had it recognized by its drivers, it has never shown me attached USB drives (while they work fine if I connect the Unicorn to the H4 - same USB and STORAGE drivers). Also, quite often the Lightning isn't recognized at all. Wondering if it's due to the hole-pin solution not fitting well in the socket (I used the ones supplied with the Lightning for that expansion port).

[czietz]

Can you please check that the Lightning ST is really pushed all the way into the socket? If I look at your photo, this doesn't seem to be the case:

pins_not_in_socket.jpg (23.72 KiB) Viewed 1069 times

[czietz]

For comparison: This is how it looks in one of my STs. (On the photo is the Storm ST alt-RAM board, but it uses the same type / length of pins as the Storm ST.) The pins should go all the way into the socket.

pins_in_socket.JPG (55.2 KiB) Viewed 1068 times

[troed]

Can you please check that the Lightning ST is really pushed all the way into the socket? If I look at your photo, this doesn't seem to be the case:

Cheers - that was indeed the reason for both the Lightning not being detected as well as the USB flash drive not working. I would never have guessed the amount of force needed if not for your photo This is, by far, the most pressure I've ever put on a PCB. Had to remove the H4 from the case to make sure it wouldn't flex too much.

In any case - now I have a like for like comparison between Lightning and Unicorn, made on the same H4 board with the same flashdrive.

Lightning: 340-350kb/s
Unicorn: 60kb/s

/Troed

[troed]

Replaced my original Atari Falcon 4MB board with the SIMM holder and 60ns SIMM. The original memory would have no chance at 16MHz.

Also got my Ricoh RTC chips (same as Mega ST). Will start looking into schematics to fit it to the H4 with GAL decoding for the address etc.

/Troed

[troed]

TOS 2.06 decoding now working. Opted for not using the Lightning but fulfilling a want I've felt for a long time, to get that working myself with a good old GAL.

Also found what looks to be a complete schematic and GAL code for the Mega ST Ricoh RTC, written by Elmar Hilgart ages ago. I'll put some effort replacing the Dallas RTC on the H4 with this, just because it plays nicely with TOS as is.

Ricoh_RTC.zip

(40.43 KiB) Downloaded 39 times

/Troed

[troed]

Finally getting to try to implement doubleST on the H4. Studying the schematics for the "jumper rows" made for the mod board, they seem to lack the ability to run the Shifter off the GAL clock, as well as DTACK_in and DTACK_out (in higher speed modes my results were that DTACK needed delaying from CPU to GLUE).

The H4 schematics actually has a set of jumpers down by flashy clock that would have the DTACK signals, but those don't seem to have made it to the actual H4 PCB.

This is of course possible to get around so I hope to have a 22V10 wired in with the latest code I did back then (not on Github) within the next few days. Exciting

Code: Select all

Clock to use is either 64MHz or 32.008 MHz depending on state of the switches. One switch combination invalid. Basically, Switch0 decided between 64 and 32MHz, S1 between doubleST video modes or not.

This clock (output) is then wired to pin1 as input clock for the GAL itself 
 
  S0 *  S1   = 64MHz Shifter clock, 32MHz MMU, 16MHz CPU = doubleST
  S0 * !S1   = 32MHz Shifter clock, 32MHz MMU, 16MHz CPU = stefan16Mhz
 !S0 * !S1   = 32MHz Shifter clock, 16MHz MMU,  8MHz CPU = stock 
 

/Troed

[exxos]

DTACK was removed just before production.. It was hard to design a motherboard to fit a addon which hadn't been designed yet But also this is why it ended up with a bodge wire on CPU DTACK, so technically CPU DTACK can still be broken