Saturday, June 21, 2014

Amiga Repair and Restoration (Part 3)

The story so far.  Amiga not yellow anymore, bad caps, started to replace caps, also did quick mod to the audio to reduce output levels...

Anyway, there are a couple of slight issues with the capacitor kit I ordered.  First of all I'd just like to say that the values are all as they should be and that there are no issues with the quantity of capacitors in the kit I received according to the pick list included.

The first issue is that they not the same physical size.  Allow me to demonstrate:

On the left is the board, pre-capacitor removal near the PCMCIA slot; on the right is the replaced capacitors from the kit:

Left - Original       Right - New 

Or how about these two capacitors near the A1200's RAM:

Top - Original
Bottom - New
The more astute amongst you may have noticed that although the capacitance values are the same the cannister sizes are actually different.  This is a bit disappointing (and I dropped my review of the kit to 4 stars because of it) as I would have expected all of the capacitors to be available with the correct physical size rather than just the majority.  I know some of you may be thinking, "Well that bit is inside and no-one will see it!" but that's not the point.  If I'm going to the effort of keeping this A1200 alive then I would much rather the board looked like it did when it came from the factory (minus the op-amp faux pas).

On the positive side, they were easier to solder in as there was lots more pad to hit with the iron...

Top tip.  Have separate flux and apply liberally to the board when de-soldering or soldering stuff on.  Don't forget to remove any excess with isopropyl alcohol.  The reason I mention it is because I don't have any and some of the caps were an absolute **** to solder without it.  Relying on the flux in the solder I have was not really very useful since it tended to burn away before I could get the iron to the pad (remember how small this stuff is!) resulting in a few suspicious looking joints.  I will have to re-visit them but not before I have obtained some flux.

As I got towards the end I realised that I had two numbers on the leaflet which accompanied the caps that did not seem to be on my motherboard.  Ah.  The good thing was I had taken lots of pictures of the board before I removed all the caps for just such an occurrence.   A quick visit to my phone to look at the pics and that was sorted.

Or was it?

In my haste I disposed of the piece of paper that accompanied the kit.  This was a mistake.  When I saw I had two caps left it looked liked I had two gaps to fill.  The values of the caps matched the values shown on the photos I had taken of the board before I started so no problem.  Except there is.  The kit contained 18 capacitors.  My Rev.1A motherboard has 19.


I have a gap at C407.  There is no gap at C460.  The vast majority of photographs of Amiga A1200 motherboards show no component at C460 but do show one at C407.

C407 - Wherefore art thou?
The first photo in this post shows clearly that I've replaced C460 with another cap.  Here, though, is what most A1200 boards look like:

C460 missing but C407 clearly visible next to the radial cap top right.
So, what the heck do I do now?  C407 is associated with Video Decoupling as per the Commodore schematics:

C407 Decoupling Capacitor
The capacitor C460 is something to do with the reference voltage at the video encoder (see bottom right of the schematic extract below).

C460 connected to Iref
Dang nabbit.  I need to try and work out whether to risk removing C460 and moving it to C407.

To be continued...

Friday, June 20, 2014

Amiga Restoration and Repair (Part 2)

The story so far...

My Amiga is no longer yellow but there are nasty things going on inside that case.

Green corrosion on the capacitor contacts
The capacitors in these machines are now over 20 years old and I don't think anyone expected that they would still be being used by anyone.  Failure of these components is inevitable and it is a bit of a lottery as to the effect when they do go.

Symptoms caused by capacitor failure include audio problems, video problems, floppy drive issues, just about anything that could go wrong could be blamed on the capacitors.  Fortunately I caught this relatively early.  Even better, because this is a well known issue with Amiga computers, particularly those with surface mounted components, kits are readily available that contain all the correct values and ratings for the A1200 (and A4000 if you're lucky enough to have one of those).

I ordered my kit from (where else?) for the bargain basement price of £6.95 + vat (+p&p).  This may seem a bit steep for 18 capacitors but if you've ever tried to navigate sites like Farnell or RS Components when all you want is a few bits then 7 quid is really a bargain.  The kit arrived a few days later and I was ready to begin the surgery.

First job was to remove the old capacitors.  I am not that experienced with surface mount stuff and so I did a bit of research (i.e. Google and YouTube) and decided that there are some crazy people in the world.  One particular Aussie guy demonstrated on YouTube how to remove the surface mount caps from his ancient games console (I forget which one) with a pair of pliers and a stiff flick of the wrist.  It looked a great way to rip the pads and tracks off the substrate of the motherboard and so I decided against this method.

The best way was to add extra solder to the pads of the caps and then using two irons, hit both pads for a few seconds and then flick the cap away from the board.  It was at this point that I discovered what people meant by the 'fishy' smell if the caps had failed.  Trying to remove C321 (pictured above) the smell was brutal.  Rotting fish mixed with week old rubbish and cat sick.  Fortunately it dissipated quite quickly but was not very nice at all.

I should say that several key things were needed to do this:

Solder (lead free with flux)
Solder wick (to clean the pads once the caps had been removed)
Two soldering irons (as mentioned above)
Tweezers (those caps are small!)
Lots of time to do it properly...

I almost had a disaster with C321 which was the first cap I tried to remove.  The edge of one of the pads lifted off the board.  It was only about a quarter of the pad so I think I got away with it.  Any more and I would have had to go and get some epoxy to stick it back down which is a very, very fiddly job.  Pretty much all of the caps came off easily and using the solder wick was straightforward to clean up the pads and remove the excess old solder.

Work started - no going back!
At this point I've taken off about a third of the caps. If you're planning on doing this yourself I would really suggest you practice on an old board that has SMD components on it before you start.  I had an old modem which after a few hours of heavy handed practice will never be the same again...

Also, take off the through-hole caps first.  It makes it a lot easier getting to the SMD caps in the big gaggle near the PCMCIA slot.  These through hole components are surprisingly tricky to get off though.  It needed a few goes to get enough heat into the solder to pull them off the board.  One cap leg in particular is soldered to a big earth plane (a big silver bit of the board in any case) which meant that all the heat went there instead of into the solder leading to several heart stopping attempts to get the thing off.

Capacitor  behind the keyboard connector
Here's another pain in the rear.  There is a capacitor right behind the plastic keyboard connector.  And the contacts are perpendicular to it meaning that you need to be a contortionist to get to it.  It's not too bad getting the old one off but this was probably the hardest one to get back on!

A Slight Diversion..

During my Googling about this I came across an interesting YouTube video from a guy in Sweden.  He had a fix for an audio fault on Rev 1A motherboards.  The eagle eyed of you may have seen that the board I have is an early 1A, so early in fact that there are NO clock pins on the board.

The issue is that the sound is really distorted when it is played back through a TV.  Think of a bad mood teenager bedroom with the stereo at full volume.  To be honest, I had noticed this problem but had assumed it was a fault with the Amiga itself rather than a design fault. But a design flaw it is. 

Op-amp with 152 resistors
In this photo there is an op-amp chip (labelled LF347M) which drives the audio output.  The gain of the op-amp is controlled by the resistor labelled '152' which means it is 1500 Ohms or 1.5k Ohms (there are actually two of them - one left channel, one right channel).  This value is too high and means that the gain of the op-amp results in clipping of the sound even at low TV volumes i.e. the Amiga audio output level is just too darn high.

The solution to this is to replace these resistors with a lower value which reduces the gain (it's the way op-amps work - google it!).  A value of 680 Ohms was suggested and proven to work by RetroGameModz on YouTube; the video is well worth a look if you're into retro gaming or computer repair. 
HOW Small??

To give you some idea of how small these things actually are, here's a photo next to a standard size paper clip..

The resistor is size 1206 imperial (approx 3.2mm x 1.6mm).  A good pair of tweezers is recommended.  In fact I  just dropped that resistor into the carpet, never to be seen again...

Anyway, see the picture below for the results of my resistor replacement....

New 681 (680 Ohm) Resistors
With my resistors successfully installed, it was time to return to the capacitors.  Except, there's a problem...

(To be continued....)

Monday, June 16, 2014

Amiga A1200 Restoration and Repair (Part 1)

My trusty A1200 was dusted off a few weeks ago and I got to wondering whether there was anything I could do about the colour of it.  When it rolled off the Commodore production lines some 22 years ago it was brilliant white.  Now, after long exposure to the sun (by it's previous owner) it looks more like someone smoked a whole lot of B&H over it..

After a bit of Googling it appears that the yellowing is caused by a reaction of the fire retardant chemicals in the plastic with UV rays from the sun.  The damage is permanent.

Or is it?

A chance discovery in Germany in 2008 lead to the development of Retr0bright (the 0 is deliberate) which is a general name for a variety of concoctions that aim to reduce the yellowing of vintage white or other light coloured ABS plastic.  It requires hydrogen peroxide with a suitable catalyst to effectively reverse the chemical reaction and remove the yellow from the plastic.

Fortunately, there is a ready made version of Retr0bright available for all those too lazy or too inept at chemistry to make it from scratch:

Find it in the haircare aisle in Boots at £1.69 a tub (3 for 2 as well!).  The haircare aisle is not necessarily a place regularly frequented by myself having been blessed with a less than generous share of occupied follicles but the 12% hydrogen peroxide cream is perfect for the task at hand.

All that is required is to paint this stuff over the yellow plastic, wrap it in cling film to stop it drying out and then expose it to a source of UV rays i.e. stick it in the sun for a bit.


Note the white square patch under the Amiga logo.  This was a label someone had stuck to the case which I removed soon after I bought it.

Even the keyboard was affected.  The photo does not really do a good job in showing how yellow all of these keys were.


Take three tubs of 40vol hair bleach stuff, add a few feet of cling film and some freezer bags for the keys and put them in front of an intense UV source (or put them outside on a nice sunny day) and wait for a few hours.

The results are startling.


The yellow has virtually gone.  It's not perfect but it's hugely better than it was.

White keys!  The yellow has gone!

In fact, the only real snag I had was that for the grey keys it looks like the 12% (40vol) hydrogen peroxide was a bit strong.  There is some white 'blooming' effect on the bigger keys which is impossible to remove.  Whatever.  It looks a whole lot better than it did.


Houston we have a problem...

During this process I noticed that there seemed to be something a little amiss with the A1200's motherboard.  Several of the surface mounted capacitors seemed to be leaning instead of sitting perpendicular to the board.  Several of them also had green corrosion on their solder pads which should have been shiny silver.  There is no battery in the stock A1200 that could have leaked to cause this so it could only mean one thing.  The capacitors had gone bad.

Note the green solder pads on the capacitor (marked U3 47 16V).  The components next to it also have green contacts.  Yuk!

The capacitor in the centre of this picture has lifted off its contacts, presumably because it is swelling up, ready to fail and leak its innards all over the precious motherboard.  Bad capacitors!

The only option to cure this is to completely replace ALL of the capacitors on the motherboard.  There are 18 in total with the majority being surface mounted.  Not a prospect I look forward to....