Tuesday, September 19, 2017

I Killed it....

Well, that was unexpected.

So, I was playing with my current system which comprises an A500+ motherboard in the case of the 'Scarlet Amiga'. It also has two external floppies (featured here) and an external hard disk, an AlfaData AlfaPower IDE drive. And you know what IDE means don't you? Yep, Compact Flash cards.

My current setup. Small but perfectly formed.
This one has (had) a 4Gb card running Workbench 2.1 with Kickstart 2.04. No problemo. I had to install a driver on the RDB using Oktapussy which worked fine to allow me to use the card properly and I split the card into 2 x 2Gb partitions without any problem.

So far, so good.

Back to the story. I was faffing under the desk and when I sat up I somehow managed to nudge the corner of the A500, hard (it was in a slightly different position to the photo). I dislodged the external drive and sent the thing into meltdown... Screen flashed purple, lines at the top and a distinct sense of 'something's going badly wrong!'. Immediately I unplugged it and then tried not to panic.

External hard disks for the Amiga are not that common and can be quite expensive when they appear on places like eBay (pah).

I carefully reinserted the drive, checked it wasn't overhanging the edge again and then turned it on. The LEDs on the Amiga and on the drive came on as expected. Then it went to the purple boot screen. The good news is this meant that the Amiga was OK. The bad news, something went wrong with the drive. Oh....

Flailing around like a lunatic I managed to find a test disk to see if the RAM was OK in the external drive (it comes with 4Mb). It booted OK but then it would only show me the trapdoor memory - it was too early a version of the test disk. More flailing. I found a copy of Install 2.1 and slammed it into the drive. The Amiga booted and told me I had 4Mb. Phew! Then promptly crashed. Then went red screen (kickstart failure). AAARGGHH!! Turned it off then on. Yellow screen (CPU error). AAARRRGRGHHH!!

Turned it off. Went for a brief lie down and a cuppa.

When I came back I removed the CF card from the drive. The cable is just long enough for the card to sit outside the drive in the CF adaptor allowing easy swap out when needed. Turn on. Purple boot screen. OK, back to square one.

By this time I had located a later version of the test disk which would look at all RAM on the system. It found the 4Mb. That was a big relief because the memory in this drive is in ZIP format - ZIP stands for Zigzag Inline Package. This type of RAM is now more difficult to find than Unobtainium, Youllneverfindthatinamillionyearsium and Sellakidneyforebaypricesium. The test disk also had a RAM test which I started and left running for several hours. No faults found. Phew.

4Mb Memory Found!
So without the CF card everything seems OK. Wondering if the problem was with the card I plugged it into my PC. This normally prompts Windows 10 to say it needs formatting because it's not intelligent enough to know what a Rigid Disk Block type disk is. But now, nothing happens. Running fdisk just confirmed that the PC was not detecting the CF card.

After puzzling over this for a while I decided to take the cover off the card and have a look inside. I've never seen the inside of a CF card so it would be interesting whatever happened. A couple of minutes with a small screwdriver and I found this:

A burn mark - hmmmm.
Well, there's a clue. There's a great big burn mark on the inside cover. Maybe if I look closely I can see what may have caused this...

Well, there's your problem.
Oh. That component looks:

a) like a voltage regulator
b) dead

Here's wide angle shot to give some idea of where this bit is located.

Wide angle...
It's that slightly lumpy mess underneath the square chip.

So I grabbed a spare 8Mb (yes, megabyte) CF card that, by coincidence, I had installed Workbench 2.1 onto as a test some months ago. I plugged it into the CF adaptor and switched on the A500. And it just worked. It booted into workbench with no problem at all. Thank goodness* for that.

*Substitute the word 'goodness' for any dockyard expletive you feel appropriate.

The question is, what the heck happened? Well, one theory from the FB group is that the sudden disconnection of the edge connector upset the grounding of the CF and allowed a large voltage/current through it (apparently that's why hot-swappable stuff has longer ground pins so there's no confusion). I suppose it's also possible that the edge connector pins may have shorted 12v to a rail that shouldn't have 12v on it. I just don't know. Either way, I am extremely lucky to have just destroyed a CF card, that is cheap and easily replaceable, and not my external A500 hard disk controller, which isn't....