modern RAM boards possible?

[radventure]

I'm not well versed in all things Apple // but given the prospect of sourcing multiple 4116 DRAMs for either troubleshooting, repair or upgrading to 48K, are there alternatives? 4164 DRAM ICs (modified) can work as far as I know (happy for anyone to chime in on the "gotchas" of this). I've repaired a few ZX Spectrums using complete "lower" and "upper" ram boards where a PCB covering the entirety of the bank of RAM, connected via machined pins to the emptied RAM sockets, carry an SRAM based module and some supporting components. Is there anything like this for the // and //+ ? I realise the original aesthetic is lost with this approach.

Example images, for 48K spectrum (credit to Tynemouth Software blog) - warning, may contain images offensive to Sinclair-Sensitive people

 

[Hugo Holden]

Gawd! Its a real shame the IC sockets will be damaged from the large rectangular pins on that RAM adapter board and are now unsuited to receiving a real IC pin. I'm surprised they managed to get those into the holes in the machine pin sockets.

Apart from that, it looks like a great idea to me.

Gold plated Round pins with small diameters in the range of 0.45 to 0.5mm which suit machine pin sockets, especially with some lubrication added are made by MillMax and available at Mouser.

Still 4116's are easy to get and not very expensive either, and they keep the computer authentic. I got a very large swag of purple ceramic body ones from Ukraine to run in my PET, all excellent, ex Russian mil spec types.

 

[radventure]

There's a "Land Rover" feel to some of these repair parts, cold chisel, hammer and shifting spanner are standard fitting tools . I'm pretty sure the boards I've used have been round pins.

note - not knocking Land Rovers, driven quite a few, and have had a couple in the family.

 

[Someone]

A funny soldering job on the adapter card. The SMD components seem to have far too much solder on them.
One can replace the square pins with appropriate round pins from companies like mill-max.

 

[Hugo Holden]

There's a "Land Rover" feel to some of these repair parts, cold chisel, hammer and shifting spanner are standard fitting tools . I'm pretty sure the boards I've used have been round pins.

note - not knocking Land Rovers, driven quite a few, and have had a couple in the family.

Speaking of Cold Chisel. this is one of my favorite songs, the guitar work is wonderful:

 

[radventure]

A funny soldering job on the adapter card. The SMD components seem to have far too much solder on them.
One can replace the square pins with appropriate round pins from companies like mill-max.

I definitely could have picked a nicer example of that lower RAM board! There are many variations around, most are fairly inexpensive, generally all provide better reliability and cooler operating temperatures.

Here's one for a C64



Might not appeal to some - this one is very striking. I've had a bit of a look around the few resources I know for Apple stuff, but I havent seen anything like this for the Apple // and //+ machines with large banks of DRAM. Hoping maybe I've missed something.

 

[Hugo Holden]

I definitely could have picked a nicer example of that lower RAM board! There are many variations around, most are fairly inexpensive, generally all provide better reliability and cooler operating temperatures.

Here's one for a C64

View attachment 1272009

Might not appeal to some - this one is very striking. I've had a bit of a look around the few resources I know for Apple stuff, but I havent seen anything like this for the Apple // and //+ machines with large banks of DRAM. Hoping maybe I've missed something.

That design shows that the person who did it, definitely had an interest in the physical appearance of it, as well as its electrical properties. This is always a very good sign, when people care about both factors. Often one gets missed. Though I think with this one too, the pins are not ideal, because one day you might want to fit the original memory IC's back in the sockets. For dual wipe sockets, ideally the claws should not be stretched apart more than 0.3mm, which is the thickness of a standard IC pin.

In this case though, if that was done with the correct size pins, likely there would be too few pins present to make it mechanically secure. So It might have been better with all the other pins placed to engage the socket claws, even if they connected nowhere, and had a dual rectangular pcb that would be less cosmetically interesting.

 

[epooch]

Replacing with 4164 DRAM is pretty trivial, and addressing 48k of them (replacing all 3 rows with one row of 4164) just needs a few signals from around the motherboard, and a 74ls153 on an external board instead of the the 74LS257 in J1. Also, you need to use "128cycle/2ms refresh" (like MB8264A) DRAM to comply with the Apple II refresh scheme. I modified a junk Apple II+ RFI version motherboard to do this. I came to the final conclusion that a RAM board would have been a better idea, but I never made one after trying to use all 64K of the RAM like the Apple II+ 16K memory card does. It was taking too many ICs and getting unwieldy. I might take another shot at it using a GAL.
The problem with this approach is that all this extra hardware and the MB8264A requirement probably make sourcing a few 4116 DRAMs and a ubiquitous 16K RAM expansion board a less costly solution. Using much newer DRAM with integrated refresh (or SRAM) might be the way to go.

 

[radventure]

I wish I had the skills for such an endeavor. I wasn't sure if this work had been done for a 16K board to easily populate a/each bank of a II/II+. I've seen some interesting IC leg oxidization, or I guess mild corrosion affecting just the 4116 ICs on a board. They might be all ok ...... for now, but there's quite a few to replace to replace for peace of mind (and until I can get a RAM tester unit to prove each one of what I have or any nos ones I get). I've only re-used the cleanest ones (with a couple of NOS when I got errors) in the first bank for now.