Harlequin 128 repair

I bought this on ebay as 'non working' with the owner suggesting that they didn't have the experience to effect a repair, I saw this as a good challenge as I never buy anything that is working!
First job always remove all IC's to protect them should I need to power up the board and also to allow for testing before reinsertion.

Topside
All IC's correct.
Pulled all IC's for later testing
U17, 24, 36 had badly bent pins, they could have still been making contact with the socket though.
U43 had pin 7 bent completely underneath so was not making contact with the socket.
Many of the IC's showed signs of tarnishing
According to the diagram U47 is a resistor pack, however a 74HC257 was fitted, this IC should have gone in as U41 at the top of the board where a socket had not been fitted. P7 of the assembly instructions say that a socket in U47 was not required, as a guess the original builder fitted the socket in U47 first then when they got to U41 had ran out of sockets (and IC's) U41 is pivotal to the working of the machine so wouldn't have worked without it.
All resistor vales checked against the BOM OK
All diodes tested OK
Underside
I noticed a significant number of unsoldered or dry joints (circled in red) so I reflowed the entire board. While doing this it became obvious that some of the blame was to go to the 'tarnishing' of both pads and component legs, it took some aggressive rosin flux to get any solder to take. I suspect that both the board and components were stored somewhere damp before assembly.
One concern is that some of the pads had been lifted during assembly, particularly with some passive components (resistors, diodes, transistors), as though the component had pressure applied from the component side during soldering. eg, D4 had both pads lifted and separated from the tracks so these needed remaking, also Q7 & Q11 emitters pads were lifted and broken away from the tracks. Oddly enough Q11 collector pin was snapped off at the transistor (not sure how this could have happened, so I replaced it.
 Picture
​As an aside; I am not a fan when builders bend the component legs over, this makes it very hard to extract them as necessary but also I did find a few bridges where the leg had made contact with an adjacent component!
​First powerup (with all IC's removed.)
(V in, 5.1v out of VR, however, half the board was unpowered, traced to a lifted track on the other side of D4, surprisingly this very thin track fed an entire half a board. Once repaired I measured the corner to corner Vcc resistance at 900mOhms, so there was either another high resistance track somewhere or this was part of the design decision to use 74HCxxx IC's to lower the current requirements. Not being happy with this I added a corner-corner Vcc strap.
​Before replacing the IC's I passed them through the Retro Chip Tester Pro and they all passed. I also dumped the ROM (27C256) and it matched the version I had previously downloaded. I also tested the Z80 at 8MHz.
All IC's installed, except IC47 which was meant to be a 470R*8 resistor pack, this was missing so I made my own (circled in red). I replaced all the jumper links with bright green or blue ones as they were hard to see when black, sandwiched between black IC's on a black PCB, get the idea?
​
Second power up, Spectrum startup screen showing, installed the Diagnostic card and it ran through the basic diags with no issues, just need to build a case/keyboard from recently acquired parts now. I did measure the current draw and it was quite low at 150mA, a result of using HC components again. Finally treated the underside of the board with Silicone Conformal Coating.











​Installed in its nice new case, with a new keyboard membrane.
Did need to make a new hole in the back for the RGB port
And not forgetting that the cassette port is now combined into one socket so they needed relabelling.