This #Amiga 600 is broken. My favourite kind! I also look at a new method of removing surface mounted electrolytic capacitors. You may be triggered…

The ugly duckling of the retro computer world. 

Early in 1992 when I was sporting my beloved, but starting to creak, Amiga 500, Yes, that’s me. Commodore in all their wisdom, decided what the home micro computer market needed was pretty much what we already had, only smaller, and worser. The scorn us 500 owners poured on the new Amiga 600 was shameful, but fully deserved. I mean, it had the worst compatibility of the available machines without any meaningful upgrades to the silicon. All of the fancy expansions you could get for your A500 wouldn’t fit, and the num pad was missing! It felt, at the time, like a big backwards step and I, and everyone I knew, ignored it completely. In fact we went a step further and loathed that little Amiga.

So what about today? Did the ugly duckling turn into a beautiful swan?

Well, no, not really. It’s still a duck, but it’s grown into its face now, and in the right light is quite a handsome little wedge. 

This one here is a very nice specimen. Both on the outside, and on the inside. I was sent this computer by friend of the channel Colcuz, who donated it for the charity event I’m running later this year. You’ll be able to buy it and many other very special Commodore machines (and one lone mystery non-Commodore machine) in a huge auction to be run on the 14th of September. I’ll be live streaming with lots of special guests from all around the world of retro tech and gaming youtube for a marathon (for me) 12 hours. Together we can watch the auctions end and see how much money we can raise for the Befriending Scheme. Last year the final total was over £15k and we’re looking to beat that colossal amount this time around. I don’t want to start up the playground wars again, but can the Commodore fans really let those smelly Speccy oiks win this epic battle? 

Inside this lovely machine you can see it’s been upgraded with a few, er, upgrades. Here is an IDE to Compact Flash adapter taking care of hard drive duties. In the trapdoor is a ram expansion, bringing the onboard Chip Memory up from 1mb to 2mb. And here, clamped over the 68000 CPU is a clever board that adds fast memory to the system. And in the ROM socket is an upgraded kickstart 3.1. 

The expansions are not extravagant, but they do turn this into something much more useful in as much as an Amiga can be useful in the modern world. Basically it’s going to be a lot easier to play games on this than your bog standard A500. 

That older, and much larger, machine can be a bit of a faff to use. You need the right Agnus to allow enough chip memory to let you run games from hard drive. Oh, and connecting a hard drive usually entails purchasing something period correct like an A590 to plug into the side expansion connector. Not cheap or ideal. The alternative there is using a Pistorm and mounting hard drive images with that, which isn’t everyone’s cup of tea, or buying an expensive real hardware accelerator with on board IDE. And whilst it wasn’t a problem back in the late 80’s and early 90’s the A500 takes up a huge amount of desk space now. The 600 just fits nicely. Its cute. 

I would go so far as to say that the 600 is one of the best real Amiga’s to own now, here in the future. 

The other best Amiga to own now is of course the Raspberry Pi. 

Now I’ve had a chance to get to know this little dude I really want one for myself. And luckily, a while back, a very nice man called Neil, or Neildo, sent this one for me to add to my own collection. I like to think of this as a blank canvas. There are no upgrades fitted, it needs a good clean, and best of all its properly broken!

So where do we start?

Well I want to start with replacing all of the leaky electrolytic capacitors in here, and I want to try a new method of removing those pesky things. My preferred method is two soldering irons, but last time I melted the plastic on the audio ports which was quite upsetting. So this time I think to reduce the chance of melting anything I am going to use a combination of hammer and chisel. I have some lovely sharp chisels here in different sizes which should make getting into tight spaces quite easy. And various hammers to give me the perfect amount of control. 

Right. Let’s pop the top off and get the motherboard on the workbench. 

Wait a minute! All the caps have already been removed! Oh That’s a shame. Now we may never know how brilliant this hammer and chisel method could be!

Now we have the motherboard free I can show you the problem with this machine. On the back side there are two ceramic capacitors that have burnt up. Properly burned. The board substrate beneath is blackened and charred. These are decoupling caps which sit across the ground and 5v rails. If these have burned then it’s safe to say something bad happened to this machine. Now it is possible it’s just one of them went short, maybe this one here where all the burning is, and that took out the other one. But what’s really worrying is that there is still a 3 ohm short across ground and the 5v rail. So the likely thing is that more of these caps are damaged and causing the high resistance. Another possibility is that whatever happened with these caps also took out some of the precious chips on the other side. The very burned one, for instance, is directly underneath the 68000 CPU. For now I have to work on the assumption it’s just the caps and and with that in mind I took some resistance readings at various places across the board and compared the results, looking to see which side was more shorted than the other. This pointed to something over here on the right side of the back of the board, near to where the burned caps were. My next idea was to inject some low current, low voltage into the board and see what glowed. I took a spare power supply lead, worked out which wires were the ground and 5v, clipped my bench power supply to those, and then set the supply to 1v and 300mA. 

With my thermal camera ready I switched it on… And just down here this cap started to heat up. Not much, I wasn’t putting much power into it, but enough to show there was some unwanted resistance there. The simplest way to sort this out is to start pulling the decoupling caps off the board one at a time until the short disappears. I started with the caps close to where the resistance readings indicated. Ill hand you over to bench Lee to explain what I found.

Right this is where I’m up to I’ve removed three of these decoupling caps these three are off of the other side of these logic chips u47 u46 and u45 and c47 was a 30 ohm short and c46 was fine c45 was 2.8k this one up here which is behind the ROM C6 was 28 ohms and if I

measure across the rails now between ground and 5 volts I’m getting 61 ohms so my guess is that there’s well there are some more decoupling caps dotted around I reckon they’ve all been damaged so I’m going to take all of them off and replace all of them hopefully I’ve got enough 

Have a shave man. You’re almost as hairy as your mic.

Now, I’m not really used to working on this type of board, and I’m still very much learning how to use my hot air station, and so there follows a bit of a mistake.

You see, I started off with the air set to 350 degrees, as that’s usually enough for most things. But the caps were very reluctant to come off. So I assumed that was because the ground points on these caps were sucking away all the heat. But no, what was happening was the caps were hard to remove because Commodore had glued them to the board during the manufacturing process. This is actually quite common with surface mounted components. And it does make removing them a bit of a gamble. You don’t really want to pull too hard in case it’s not really ready to go and you rip a pad off, but you have to break the glue. 

And in the end I managed to delaminate the area around this cap. I wasn’t very happy with myself. That was the point where I worked out what was happening with the glue. 

After this I started taking the caps off with a soldering iron with a large blade tip. That was far easier, quicker, and safer. You live and learn. Please feel free to beat me up in the comments.

I worked my way through all of the decoupling caps on the front and back of the board, around 30 of them, testing the resistance as I went. And eventually with just about all of them off the board the resistance returned to something more in the expected range of around 130 ohms.

The caps in the middle are all still within spec, the 5 on the right are properly shorted at various resistance values. The pile on the left are still capacitors rather than resistors, but the capacitance is now either too high or too low. With this in mind I decided to replace the lot. After all, if these ones had failed, possibly for no reason other than they’re a bit old and rubbish, then the others were likely to follow suit. I ordered a 100 of them and waited for the postie. It’s possible this board is still beyond my ability to repair. If it suffered an overvoltage from a bad power supply at some point, causing those caps to fail, then it’s likely that overvoltage not only took out the caps, it probably took out some of the silicon also connected to the 5v rail. My best chance is it wasn’t an overvoltage, but maybe just one of the caps failed short and the others were just on their way too. The only way to know for sure is to install all the missing electrolytic caps, replace all the ceramic ones I removed, and switch it on to see what happens. So while I wait for the ceramics to arrive I’ll get these electrolytics fitted.

First job is to clean up all the pads. Neil did the right thing removing all these caps, saving this machine from further damage. There are signs of light electrolyte corrosion in a few places. Especially here around the reset circuit. From experience I will need to remove the 555 timer and these capacitors to clean underneath them.
The postie brought me some new 330nF caps and I set about putting them on the board.

And the moment of truth has arrived. Either this board is toast and good only for spares, or replacing with a raspberry pi, or I got really lucky and the only thing wrong was the caps had all gone bad. 

Woo hoo! It works!
Well. Mostly. There’s some glitching going on here with, it looks like, the red signal disappearing from the RGB image. I will just ignore it for now and hope it’s just a settling in thing. 

Some dyna blaster to test. A firm favourite from my old amiga days when I used to take my own A500 to the pub for multiplayer mayhem with my friends. Happy weird days. 

Next job on the list is to get this machine all cleaned up. It’s in a really bad state and I spent an evening stripping all the keys off the keyboard and giving everything a good scrub in soapy water. And underneath all that filth and dust it turned out there was a fantastic little machine in excellent condition! A few of the keys have slightly yellowed but a day out in the sun, with a leafcutter bee for company, evened that out enough for me. 

Now I have a nice condition, working Amiga 600, I need to put a few of those upgrades we saw earlier in the charity machine in this one to make it usable. 

First up is the trap door expansion. On my discord server I have a channel for swapping PCB’s, and I took Nick up on his offer of an Amiga 600 ram expansion to double the chip memory from 1 to 2mb. He also sent some of the smd passives and the clock chip he had as spares. Cheers Nick!
I still needed to source a connector, the battery holder, and a few resistors. Oh, and some memory. I asked again for help from my discord members and Fuzzy Lee offered me an old PC graphics card with the required memory chips on. Thank you Lee!

Getting the chips off this board shouldn’t be too tricky. And whilst it’s likely this card won’t ever be used again I think it’s sensible to do it non destructively, so I’m taping aluminium foil over the plastic bits to protect them.

It’s not a thick board so just 350 degrees and off they both pop. I’ll leave the other two here for safe keeping.

I’ve never worked with this type of chip before, and I hear they are a pain to solder, so this should be a fun learning experience!

I start off by cleaning the pads with IPA. And then adding some liquid flux. This isn’t a good move. Liquid flux tends to boil away very quickly and as I add some solder to the pads it’s not flowing. 

I replace the liquid flux with some better quality paste from a syringe and try again. This time the solder flows perfectly onto each pad.
Offering the chip up it doesn’t want to sit on the now raised pads. This feels like it might be tricky already. My plan is to reflow the pads with hot air and hope the chip settles in place. After a go with the hot air I can see it’s not going to work that way. And so I return to my trusty chisel tip iron and try a bit of drag soldering instead. It seems to work much better. A quick test with a pair of tweezers to see if any of the legs are moving, and I find there are one or two that are still loose. I think this just wants much more solder than I’m offering. Adding more flux and much more solder I go again.

Ok, that went much better. For the other chip I don’t bother with tinning the pads and just tac it in place and then drag along the legs with plenty of solder. Worked a treat. 

It’s nice and simple to see if the legs are properly connected by just testing for continuity between the pad and the top of the leg. Being careful not to push too hard and get a false positive reading. And I do find one leg that isn’t connected. A quick dab with the iron fixes it and all the others test good. A close look at all those lovely solder joints. No bridges and all perfectly attached.

Now, with the chips in place I can switch to the microscope to fit all the SMD passives. 

First up is this relatively large tantalum capacitor. I’m used to silkscreen markings indicating the orientation of components’ polarity, but there isn’t anything obvious on this board to show me which way round this cap goes. Unless that extra bit on the right side counts? I had to go look it up and I found a video that my pal Glen over at CRG had made building this exact same card. He also had to work out which way round this went. I’m still learning all these things and will probably forget quickly anyway, but thank goodness for youtube and amazing channels like Glens. He also showed which way the LED needed to be installed and saved me having to look that up/
Look! I even get my name on the components! How cool is that?

There were a couple of components not listed on the bill of materials, which meant I needed to examine the schematics to get their values. This resistor here for instance at R6 is a 1k. 

I do like SMD soldering, but that’s all done now. Time to fit the other bits.

Edge connector, battery holder, pin header and clock chip all fitted. Will it work? Lets plug it in and find out!

I recycled an old floppy pirate disc into an Amiga test kit disc and booted from that. 

The image is still a bit flickery. I don’t have time to fix that right now so I will probably look at that in another video. Possible on my second channel. But there is now 2mb of chip ram and it all tests good!

The last thing I need to do is fit the fast ram I purchased for this machine. This uses a plcc type socket press fitted over the 68000 CPU. It’s a bit hit and miss with these as to how well they sit on these chips. And this one just doesn’t go on very securely. This other one inside the charity Amiga 600 looks a much better fit and seems very secure. It also just about fits under the hard drive caddy. It might need a bit of metal filing off to make a better fit though. The other one will not fit under the caddy. I’m not sure I will leave this one inside my own 600. If I can live without the expanded memory I won’t use it. Another problem with these fast ram expansions for the 600 is fitting them means the PCMCIA slot doesn’t work. Which makes transferring files too and from the Amiga a lot less convenient.
The only other thing I need to install is a hard drive solution. I have plenty of these around and my friend Matt from the brilliant Tech Made Easy youtube channel kindly sent me a 4gb Compact flash card I can use with this machine. And I have this compact flash adapter which has been sitting in my spares bin waiting to be used for far too long. 

I can see myself using this amiga far more than my A500, 1200 or A500 plus. It’s just the perfect size to plonk on your desk and have a blast of sensible soccer or dune 2. It’s a cracking little machine. 

Thank you Neildo for sending it to me. I will treasure it forever. 

If you want your own amiga 600 you can bid on this one in September. And if you don’t manage to win, I have another one here you can bid on too!! This one was very recently given to me by Andrew Searle who popped round with a box of amazing donations for the charity and the channel. Andrew said this one was for me if I wanted it, but I said I already had a 600 and if he didn’t mind I would put this extra one in the charity auctions for someone else to own and hopefully raise a big lump of money for the befriending scheme!
This one has the notorious reset problem that happens when the caps leak over the reset circuit. I don’t know if it will make it into a video as I not long ago repaired a very similarly broken 600 in a video you can see in the corner of your screen. But the other thing with this machine is it’s very yellow, and I recently bought a big load of peroxide and I might give retrobrighting a go!  Keep an eye out here and on the second channel for updates about some of the amazing things I will have for sale on September the 14th. Put it in your diary! And if you just want to donate you can do so on the Just Giving page which you will find linked in the description. 

Lots more to come getting all these machines ready for you! See you soon!