48k Spectrum Replacement Lower Ram Options

For experts to discuss very technical stuff and newbies to ask why the Spectrum they bought off ebay doesn't work.
Post Reply
]-[
Drutt
Posts: 7
Joined: Thu Feb 01, 2024 11:43 am

48k Spectrum Replacement Lower Ram Options

Post by ]-[ »

I have a poorly 48k Spectrum Issue 4B, it seems like a lower ram issue, I'm just waiting for some BAT85 diodes so I can do the OR gate mod. I intend to try the DiagRom written to a 27128 chip to verify this is the case. Doing the resistance check between pins 1 & 16 comes in at about 2.7kR across all of them so it doesn't look good. I have available to me some quite a number of MB8118-10 ram chips and also HM4864P-2. Would either of these be OK as replacements if the board is modded to provide only +5v to pin 8?

Thanks H
cj7hawk
Manic Miner
Posts: 361
Joined: Sat Sep 25, 2021 9:11 am

Re: 48k Spectrum Replacement Lower Ram Options

Post by cj7hawk »

4116 RAM chips are inexpensive and easy to get from China presently.

And if uncertain, a 4116 DRAM tester is also relatively cheap.

8118 should also be OK, but will require changing rails as you noted.

Though if the pins really are NC on 1 and 9, it probably won't matter much whether the board is modded or not.

You still need to make sure the power supply is fixed too.
User avatar
1024MAK
Bugaboo
Posts: 3123
Joined: Wed Nov 15, 2017 2:52 pm
Location: Sunny Somerset in the U.K. in Europe

Re: 48k Spectrum Replacement Lower Ram Options

Post by 1024MAK »

Hi :D

Yes, the MB8118-10 DRAM chips will work. Pin 1 (VBB, -5V) will need isolating.
Pin 8 which is VDD +12V on a 4116 will need changing to +5V for a MB8118-10.

For a HM4864P-2, again these DRAM chips should work. Again, isolate pin 1. And pin 8 now needs a +5V supply. You can leave the existing +5V supply to pin 9, on these chips this is A7, it needs to be a defined logic level.

Keep in mind that the video section of the Speccy still needs a +12V supply.

Mark
:!: Standby alert :!:
“There are four lights!”
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb :dance
Looking forward to summer later in the year.
]-[
Drutt
Posts: 7
Joined: Thu Feb 01, 2024 11:43 am

Re: 48k Spectrum Replacement Lower Ram Options

Post by ]-[ »

Thanks for the replies good to know what options I have. Of course if it only turns out to be 1 or 2 of the existing ram chips it probably makes more sense just to swap those rather than re-solder all 8 to change the chip type.

I did the or gate mods and got the DiagRom in. It seems to be running through the ram tests but the screen is a bit messed up and output is doubled so hard to see exactly what is going on.

I have removed tested and socketed IC3 and IC4 (both tested OK), also IC23 (tested OK) and IC24 (replaced IC) but this has not made any difference.

This is how the DiagRom splash screen looks
Image

I get the one beep at the start then a flickering yellow border with a strange accompanying buzz from the speaker, then what looks like a results screen followed by 3 beeps then a different screen of again what looks like results then a grey border with a black centre. I can post a video somewhere if it would be helpful or interesting?

I'm pretty sure the caps need replacing having tested some they seem to be well out of spec, is this likely to be contributing to the garbled display? Any suggestions of what to check next would be gladly received.

Thanks H




**** Stop press, disappointingly the issue is the ULA :( *****
cj7hawk
Manic Miner
Posts: 361
Joined: Sat Sep 25, 2021 9:11 am

Re: 48k Spectrum Replacement Lower Ram Options

Post by cj7hawk »

I'd say it's extremely unlikely to be the capacitors.

Looks like a stuck address bit... Address Bit 0/7 - Pin 5 to all the DRAM chips... Probably stuck low. On the ULA/4116 side of the resistors.

There's only 1 or 2 capacitors on most speccys that can cause trouble with age. Most of the others won't cause much worse than video noise at most.

Sorry, should add, either there's a board fault (eg, broken track/ shorted track ) or a ram chip fault ( pin held low by a ram chip ) or the ULA has failed at the A0 output.
Post Reply