exxos's Atari, Amiga & retro forum

HigashiJun wrote: 18 Sep 2026 10:55

rubber_jonnie wrote: 17 Sep 2026 20:58 You've got as many projects on the go as me!!!

Nobody has more projects on the go than rubber_jonnie...

:lol:

:lol: :lol: :lol:

Indeed. I only have one software one and one hardware one at the moment, and I've not really started the software one.

Anyway, on the lynx 96K front, other than the missing resistors all the passives and transistors are on the board except for most of the decoupling caps, over 50 of the blighters! I've modified the video RAM board so that it's compatible with 4164/41256 chips and hence doesn't need -5V and +12V.

There is one ceramic cap in the RF/composite circuit which is an unobtainium value, 56nF, which I've left out. I'm not going to use RF or composite out anyway as that needs +12V and if it's like the 128K I'll not be able to get a stable chroma signal anyway.

I'm also leaving out the UART as I don't need the serial port.

Once the resistors turn up and I've got enough decoupling caps in place I can start scavenging parts from the 128K board, such as the main crystal, and start getting the board ready for a first test.

stephen_usher wrote: 18 Sep 2026 13:19 Indeed. I only have one software one and one hardware one at the moment, and I've not really started the software one.

Anyway, on the lynx 96K front, other than the missing resistors all the passives and transistors are on the board except for most of the decoupling caps, over 50 of the blighters! I've modified the video RAM board so that it's compatible with 4164/41256 chips and hence doesn't need -5V and +12V.

There is one ceramic cap in the RF/composite circuit which is an unobtainium value, 56nF, which I've left out. I'm not going to use RF or composite out anyway as that needs +12V and if it's like the 128K I'll not be able to get a stable chroma signal anyway.

I'm also leaving out the UART as I don't need the serial port.

Once the resistors turn up and I've got enough decoupling caps in place I can start scavenging parts from the 128K board, such as the main crystal, and start getting the board ready for a first test.

Sounds promising, and getting rid of the 4116 is always a good thing.

Bother! Missing some chips: 74LS12, 74LS26, 74LS243 and 74LS393.

Managed to get all but the LS12 from Mouser, though some of those stocks are low. The LS12 ordered from Cricklewood.

Oh, and I think I've worked out why the 128K machine had issues with memory refresh... I blindly installed the chip that Mouser sold me as a 74LS393. When I removed it from the board today to transfer into the new board I discovered that it was a 74LS11! No wonder it didn't work. It's amazing the machine functioned at all.

stephen_usher wrote: 18 Sep 2026 17:43 Bother! Missing some chips: 74LS12, 74LS26, 74LS243 and 74LS393.

Managed to get all but the LS12 from Mouser, though some of those stocks are low. The LS12 ordered from Cricklewood.

I hate it when that happens...

Good old Cricklewood though, they're pretty handy :)

The chips haven't arrived yet but I've got on with other jobs.

I've scavenged DIN sockets and the expansion connector from the 128K board, along with the video crystal and the variable resistor and variable capacitor. Even without the 56pF cap there's a clock of sorts getting to the TEA socket.

I've also created the new ROMs for the machine and fitted those. I've put a buck converter with big filter caps on the 12V rail to power the TEA and I've fitted a modulator.

Powering the machine up now with a bodged CPU clock, if I hold it in reset and release I'm getting activity. Code is running. It's only the lower RAM bank which are getting the correct signals as the logic to drive the upper RAM and video RAM just aren't there. The lower RAM is getting RAS and CAS, however. The problem is that although the main reset circuit is complete it's obviously not holding the CPU and everything else in reset long enough. I can look at that later though.

Well, feeding a proper 4MHz clock to the CPU sometimes after reset it's giving the start-up beep, which is encouraging.

stephen_usher wrote: 21 Sep 2026 17:37 Well, feeding a proper 4MHz clock to the CPU sometimes after reset it's giving the start-up beep, which is encouraging.

You'll soon be the Lynx expert :D mines upgraded to 96K now.

Found the reset problem. I had the diode the wrong way around. The board reliably starts up and beeps now.