exxos's Atari, Amiga & retro forum

OK, I soldered in the "wrong" transistor temporarily and it works... after warming up, which suggests some capacitor fun too.


I can't get the smaller of the two keyboard ribbons to go into the connector after trimming, which suggests that there's a bit of old ribbon in there, which I can't see.

However, using a bit of wire to connect between different wipers on the two connectors produces a character, which shows that the machine is actually running.

Oh, and that resistor was most definitely cut for some reason.

I recently had similar issues with a Spectrum, definitely sounds like a bit of ribbon stuck in the socket. Had a job and a half getting the bits out.

Nice to see it powered on!

Given that the memory corruption gets better the longer the machine’s on it may be a capacitor issue.

As for the ribbon, I couldn’t see anything. I think that it’s corrosion on the contacts. After polishing with some paper & contact cleaner I got the ribbon in, not that it works much at all. I’m definitely going to have to get it new one, and some case screws.

This evening's Speccy fun...

Given the issue with the +12V and -5V lines I decided to desolder all the 4116 RAM chips, fit sockets and replace them with hacked 4164 chips.

The desoldering was a bit of a hell as the factory had bent the ends of the legs of the chips to hold onto the board for soldering, so the Hakko wasn't enough and I had to get the hot air station out and use a lot of flux plus a scary amount of pulling and manoeuvring with some tweezers to wriggle the chips out. Thankfully no traces, pads or vias were harmed.

Several of the holes had filled with hardened flux, so I had to use a paper clip (which has JUST the right diameter) to poke through whilst heating with the soldering iron!

Anyway, I got the sockets all in, straightened the pins on the 4116 chips and put them back in to test if I'd broken anything... Phew! Nope... as it was before.

So, the next job was to start modifying 4164 chips. Cut pin 1 and wrap pin 8 over the chip and solder a wire between it and pin 9.

Did the first one and decided to test before going onto the next, so hoiked out the first 4116, put the modded RAM in.


I then plugged in, switched on and...

Ta-dah! It worked! By sheer luck the faulty RAM chip was the first one. Obviously the random NPN transistor in location TR4 is good enough.


I need a new keyboard matrix as one of the ribbons has cracked through three lines high up so it's not repairable. (Of course, one is the one serving the return key so I can't enter any commands.)

I haz transistors!

I can put the proper component into the TR4 slot tonight and generally clean up the board, and especially the 4116 RAM chips which are caked with the horrid flux I had to use to get them out.

I've also got a load of axial capacitors which I no-longer require, but they're good stock items.

I'm not sure how long it will take for the keyboard membrane to arrive, hopefully before the weekend, similarly the case screws.

Unfortunately I can't find a reset switch for sale. The only "shop" which lists one has closed temporarily (until stocks have built up).

stephen_usher wrote: 09 Feb 2022 13:44 I haz transistors!

I can put the proper component into the TR4 slot tonight and generally clean up the board, and especially the 4116 RAM chips which are caked with the horrid flux I had to use to get them out.

I've also got a load of axial capacitors which I no-longer require, but they're good stock items.

I'm not sure how long it will take for the keyboard membrane to arrive, hopefully before the weekend, similarly the case screws.

Unfortunately I can't find a reset switch for sale. The only "shop" which lists one has closed temporarily (until stocks have built up).

Can you by any chance just use another momentary contact switch for the reset, there must be something similar out there?

Yeah, I need to get the dimensions and see if I can find something on RS.

stephen_usher wrote: 09 Feb 2022 13:59 Yeah, I need to get the dimensions and see if I can find something on RS.

Sounds like a plan :)

I've just cleaned and put all the rest of the 4116 DRAMs through the tester and found a second bad one, so I'm going to have to modify another 4164 to replace it.

Ordered some switches which will hopefully fit the Spectrum+ case.

After replacing the TR4 transistor with the correct one the voltages are far closer to spec. The +12V is spot on but the -5V is about half a volt high.

Anyway, that's all the soldering done.

The rest of the eventing was spent cleaning the upper case and keyboard. Even with a good scrub and use of a tooth brush not everything came off, but it was masked with some furniture polish, which brought up the shine. As usual, due to the poor design of the rubber mat, the Return key sticks if you press it in the wrong places. The rubber mat domes really needed to be a bit stronger to push the keys back up firmly.

All's ready for the replacement keyboard membrane now, and the replacement reset switch. (Oh, and a key stem, but that'll take a while.)