Over the hundreds of threads I've read here, are many passing references to NVRAM installs gone wrong, specifically on WPC: Too easy to screw up, too easy to ruin a board, don't attempt it unless you know what you're doing and even if you think you do, just send it off to be absolutely sure. Better yet just stick to batteries in the first place. This is a mod for the manliest men, not the wannabes.

Well I'm a stubborn bastard, teaching myself by fire in this hobby (with great assistance from awesome folks here when needed). I've progressed from board repair with the metaphoric hair dryer and blowtorch, to better tools like a Hakko 808. Since then I've repaired several board issues with a great first-success rate, and more challenging items. On the NVRAM side I progressed from Sys6 to Sys9 and Sys11, successfully.

So last night I approached Dr.Who for the NVRAM attempt... Ho. Lee. Zeushit. I am still shaking!

WHY are those traces and thru-holes so stupidly fine and close together? WHY in seemingly this one and only case did Williams feel compelled to fuse a DIP chip against a board like that? WHY is that one location crammed between the ASIC and game ROM sockets to impede emergency/secondary topside access?

The SRAMs on my other games fell out with a quick wave of the Hakko FR300, but that WPC SRAM was a fugging surgical nightmare. The chip legs fully plugged the thru-holes so the desolderer couldn't open them, even as the "pin" part of the leg was floating centered. Adding flux and new solder didn't help.

Nippers couldn't get "behind" the legs or "under" the chip because the legs were bent flush against the chip, and the chip was fused against the board. So I had to snip them against those @^&^! hairline traces. Get one side of the chip free, then bend and break it off the other side (so much for salvage). Clip the remaining leg bits. Desolder the top side, be careful of those tiny pads!

Which I checked 6 times, and all seemed well until I installed the new socket and pins toned together. AUGH are thru-hole traces completing the paths elsewhere? Gah, but they go under the ASIC and ROMS so I can't see to verify! Schematic shows shared address and data lines but... gah, at this point there's nothing to do but put the NVRAM in and...

SUCCESS!

So I could breathe and play and all was well even after turning the game off and on, so hooray, I graduated. But damn, I don't know if I want to do another one again for a long while.

Are they all really that treacherous?

Ha! Yeah, I saw that post *this morning, after* I had my little battle last night. I knew better than to apply! But man, that would be an awesome opportunity and I love working with and learning from old-schoolers like that. I'm not so young anymore myself, but always willing to learn.

Ah well, even if I've a way to go to be master-class, at least I was careful & conservative enough to not screw mine up, and get it working right on the first try. I'm sure it gets easier with practice and technique.

Heh... more proof I'm an idiot: while doing more work last night I realized mixed up my Hakko model numbers... the desolderer is the FR300, not the 808. Oof! Woe be to anyone who tries this stuff with a basic iron (808)/ without at least a good desolderer (FR 300)! I amended my first post to clarify.

Appreciate all the info, tips, stories, and pics. This is work I enjoy but it's definitely not for the weak willed. Always appreciate the knowledge that skilled folks are willing to share so generously!

Yeah, I was fearful of exactly that. Which is why I nipped as much as I could of every leg all around the chip... I was simply unable to get the nippers to get complete full cuts on the ROM side of the chip - I guess maybe I need new ones. But they did cut enough to essentially score and create a weak relief point on the legs that was easily broken before any pad damage was done. Since the game runs afterward I'd say I got away with it... this time

Oddly enough, that revealed the next weak link in that my 5V is only outputting 4.88 at the driver board, so I have more work in the future but nothing this bad.

Man, doing an ASIC socket with a pult? You da boss! I did plenty of board work (including multiple 40-pin sockets) successfully with a bulb torch previously, but then that aggravation and time was enough to make me move on too. Even a good tool won't save you from lousy technique so I'd say you're better skilled than I... a tip o' the hat to you sir!

Curses, the Louisville Show was just a couple weeks ago! But that's not too far away by distance, so if you are generally free in the first weekend of March, you should reach out to the organizers (I can give you their info) as I'm sure they'd love to have you for 2018. I know myself and a couple others would be grateful for the chance to learn from a master on the spot!