Well I certainly can't say this evening wasn't entertaining. Got the HP 16500B out of the cupboard. Powered on to a screen of random lines. Turns out the "base" (Bank0) RAM SIMM had slipped out of the mounting. Power off, all the blades out, push it back in, power on. SYSTEM SELF TEST FAILED: SEE SERVICE MANUAL System board: FAILED Touch screen: IMPAIRED Base RAM: FAILED Oh joy. Cleaning the (rather dusty) touchscreen surround sorted the touch screen. Reseating the other SIMM sorted the Sysboard failure, but the memory test kept failing. Swapped both SIMMs for a 64MB set I had kicking around -- thus giving me a 64MB HP 16500B. Powered back up, "SELF TEST PASSED, LOADING SYSTEM SOFTWARE." Now we're getting somewhere. Or not. "Corrupted LAN Board! LAN Not Functional!" So evidently the 28C16 parallel EEPROM on the 16500L LAN card has developed a case of amnesia (that's if the Random Lines Event didn't nuke it). I got to the System screen, but the 16555D isn't triggering, and trying to configure the Ethernet adapter hard-crashes the analyser. Does anyone happen to have a dump of the 28C16 EEPROM on a 16500L LAN card, or even a spare 16500H or 16500L card? Or an E2479A 16500A/B --> 16500C upgrade kit... or the bits of a 16500C needed to upgrade a 16500B... I think I'll stop there :P Apparently you need HP-specific Factory Test software to reprogram the EEPROM, which was AIUI never released outside of HP. I figure it should be possible to reprogram the chip with a couple of wires, some solder, and a 28pin turned-pin DIP socket, but I need a known-good image first. So anyway, I figured I'd try and soldier on and figure out what was wrong with the 386 motherboard. In with the POST card... yep, still dying at checkpoint code 06, which basically means the BIOS couldn't read the shutdown register in the RTC. Off with the old RTC socket, a quick clean up with Flux-off, and under the microscope it goes. I started tracing the under-chip tracks out, following them with the tip of a mechanical pencil. Until, that is, one of them crumbled... Hmm. Copper isn't supposed to do that. Out with the Xacto. Scrape off the solder resist... hmm, black fuzz, but no copper?! Apparently the battery did more damage than I thought! Eventually I found a total of three broken or damaged tracks, all on the data bus. I bridged them with Roadrunner wire (really thin enamelled copper wire, like transformer wire). In goes the new socket and.... 0d : 0c Hmm, still no go. So I pulled the RTC chip, put it dead-bug style on the desk, and started looking at the tracks again. And then I saw it, out of the corner of my eye... The VCC pin had broken off the RTC chip. So what did I do? I fixed it. I took a fresh turned-pin IC socket, and inserted a piece of 22SWG tinned copper wire into the top-right corner socket (i.e. VCC). Clipped it back to ~5mm, then put the socket to one side. Next, I used a triangular needle file to cut a notch into the plastic of the chip, right above the VCC pin stub. A bit of scratching with the Xacto got me enough Vcc pin stub to solder a wire to. Lightly tinned the VCC stub, just for good measure. RTC chip goes into the socket, and the stub was soldered to the 5mm VCC wire. Into the motherboard it goes. POST card in, speaker (on POST card) connected, power on. 0201... 0605... 0d0c... 1F1E... lots of hex numbers... 403F... *BEEP BEEP* dit-dit-dit-dit-dit-dit-dit-dit So it's just passed the chipset and base hardware tests, proven that the mainboard and chipset are alive... and emitted two beeps for a fatal error, and eight error ID beeps. Quick look up... that's a Video RAM R/W Test error, which makes sense because there kinda-sorta isn't a video card plugged in... or anything besides the POST card for that matter. It's at about this point I shouted "King of the lab!". My mother poked her head around the door and gave me a very strange look.... So yeah, the 386 board is basically fixed. I need to hunt down a case for it (the K6 board is inhabiting the 386's old case). I might chuck the thing in the dishwasher to clean off the last bits of dirt and fluff, but at this point it's working and I'm happy. Or I would be if Ye Olde HP Boat Anchor hadn't picked tonight to turn its toes up... Guess you can't win all the time :( And if you question why I call it a boat-anchor, you've obviously never tried to move a 16500B... :) Thanks, -- Phil. classiccmp at philpem.me.uk http://www.philpem.me.uk/