exxos's Atari, Amiga & retro forum

I intend to do the 4164 upgrade and power supply replacement at some point in the future.

stephen_usher wrote: 26 Mar 2026 09:54

rubber_jonnie wrote: 26 Mar 2026 09:41 4116 RAM - Shudders...

Thankfully the 128K/192K version uses 4164.

Phew!

Well, ran out of 14 pin and 16 pin dual wipe sockets... RS order placed.

stephen_usher wrote: 26 Mar 2026 12:40 Well, ran out of 14 pin and 16 pin dual wipe sockets... RS order placed.


IMG_5015.jpeg

Had exactly the same problem with my Atom...

I really need to see what I'm missing socket wise and just order some.

Resistors, diodes (except for a couple I don't need at the moment) and transistors (except for one I don't need) are on the board.


Ran a power test as I had to use some equivalent transistors, some had different pinouts, and all is fine. Pulling 32mA currently.

Looks like my RS order is the last item in the DHL delivery van today, stop 93. So a long wait.

stephen_usher wrote: 26 Mar 2026 22:19 Ran a power test as I had to use some equivalent transistors, some had different pinouts, and all is fine. Pulling 32mA currently.

Yeah I know the feeling about equivalents...

I replaced some n one of my Speccys and though direct equivalents the pinouts were reversed and it didn't work so well initially.

I've been studying the schematic today in KiCAD.

Well, I wonder why they used two odd values of decoupling cap, 47nF and 22nF. They're just decoupling the TTL chips and RAM power rails. Half the ones on the RAM aren't needed as they're there for if you use 4116 chips as they're on the -5V rail.

I think I'll just throw my standard decoupling caps on the board and be done with it.

stephen_usher wrote: 27 Mar 2026 12:43 Well, I wonder why they used two odd values of decoupling cap, 47nF and 22nF. They're just decoupling the TTL chips and RAM power rails.

Likely targeting different frequencies for decoupling.(but depends on package size) So for example..

0603:
100nF ~100 kHz – 15 MHz
47nF ~300 kHz – 23 MHz
22nF ~500 kHz – 34 MHz

Leaded:
100nF ~5 MHz
47nF ~7.3 MHz
22nF ~10.7 MHz

I used 1uF, 100nF, 4.7nF on my updated ST536 rails :P But in my defence, there's a lot of HF 100mhz noise because of the clock.

The Lynx is 6MHz, so I guess that the 47nF leaded is fine. Given how far away from the chips most of the caps are I'm not sure how much use they are anyway.