Where’s the line form? I was just discovering how hard 16MB IIfx SIMMs are to find, now that I have one, and that even 4MB SIMMs are pricey. I’d love to max it my IIfx, too, it’s a great system especially with an 8•24GC installed.
Despite having had everything in hand for an absurdly long time, I haven't assembled more than a test build that didn't work, and I haven't gone over it to see what's wrong (could be solder bridges that are under the chip, could be counterfeit DRAM, I dunno). I'm trying to make 2018 my year to get off my rear and actually finish things for once in my life, so hopefully soon! I'll make an announcement once I've made progress.
OK, well. I finally kicked myself in the pants and built up a dozen of the SIMMs. I cannot seem to find a set of 4 that will boot in the IIfx, though the soldering job looks OK; I guess it's time to build up a test rig. Fortunately, I took Doug's advice and got a rail of about 20 64-pin SIMM slots from the NOS dealer he recommended, and I also got some ZIF test sockets for the SOJ chips so I can also test them before populating.
Now I just need to build a board and figure out what I'm going to lash it up to... I'm inclined to use ribbon cable to hook it up to my Cyclone II dev kit, which has some convenient ports, so that I can actually test it at full speed. The PSOC5LP might also be a decent candidate, and I have a dev board or two for that as well.
Will anyone have interest in such a test board? It should be suitable for verifying operation of any IIfx SIMMs. I might add a logic analyzer port or two for my HP logic analyzer, though the Mictor board connectors are surprisingly expensive (and a PITA to route around).
Also, I finally figured out why my IIfx wasn't making sounds: the speaker itself was dead. Curiously, it's a 32 ohm speaker, which seems to be pretty hard to find, but I had an 8 ohm PC speaker handy that was the same dimensions and had the same plug. It sounds OK for now, I'll have to do some more testing later to see if the sound gets really distorted at high volumes.
I have no doubts that your board design is sound! I don't know about the chips, thus the desire for a tester, but I do notice that when I load the bigger SIMMs in, the RAM test takes correspondingly longer. That basically tells me that I don't have useless counterfeit slugs, but it doesn't say whether they pass QC; The tester I have in mind should be able to test individual chips with the test socket (so I can test before assembly) as well as testing SIMMs for normal failure modes (bad bits, bridges, etc.).