bid wrote:One more hardware item for now. Obviously I am keen to protect the machine and make sure it does not blow up on me, especially as using it after 20 or so years, could put a strain on the 'old girl'.
I noticed that this the PSU gets VERY VERY hot indeed, if not cooled by a fan. So I have replaced the ageing fans with new, and also added a CPU speed controller to make them quieter and checked air flow carefully. Obviously, this is a vintage machine, so I am also wondering how to protect the computer as it ages, exsecially the parts that go "bang", especially as you cant just pop down to PC World and ask for parts
The STe runs pretty cool, and apart from the PSU, is very reliable. However, the TT is quite a lot different, and examining the board, I have a number of very hot chips.
///snip///
The 68030 is "warm", but the DIL's in the slots marked U3 and U4 are running "burning hot". What are these used for, and also where is my 68881/2 co-processor??

Help me!!
I am thinking of adding some of these adhesive DIL Heatsinks
http://www.maplin.co.uk/dil-ic-heatsinks-30377 around the place, and wondering if this is a good preventative measure. Im talking where chips are quite hot indeed, is this a good idea?? Also, I am also wondering what other preventative measures we can take, if any, to preserve the board and PSU. Also if a PSU burn out can it hurt the board itself?
As I say, I am not an electronics engineer, but I would certainly consider replacing components or making modifications to preserve the life of the machine if it was advisable? For example, should I consider replacing the capacitors now, as they must be 120 years old, or should I wait for them to blow?!

Four main things kill silicon transistors and chips:1, reverse polarity
2, over-voltage
3, physical damage (including corrosion)
4, thermal damage
Taking them in turn (I will use "chip" in the text, but this applies to other silicon devices like transistors as well):-
1, reverse polarity - this is not a factor unless you add a non Atari PSU and mess up the wiring - use a multimeter, don't rely on the colours of the wires (as someone I know did with a different machine - only to have it "blow-up"!).
2, over-voltage - most of the time a PSU will fail with an under-voltage failure mode, which stops the computer from working, but does not damage it. However, sometimes the regulation circuit in the PSU fails and the output voltages rise way above safe values. Needless to say, this is not good. Most chips do not work correctly with too high a supply voltage and all will be damaged if the voltage rises too high. It is possible to build an inexpensive protection circuit. Let me know if you want details.
3, physical damage (including corrosion) - With regards to the corrosion issue, if at any time any liquid gets inside an electronic device immediately switch it off and unplug it and remove all batteries. Then if possible remove the circuit boards. Wash then under clean tap water. Make sure all the contamination / liquid is removed (but be careful not to damage any components). Then allow to air dry at an angle in a warm place for at least 24 hours. Carefully check, and if any sign of water remains leave for another 24 hours. Why do all this, well most corrosion issues result from contaminated liquids which eat the copper PCB tracks and component legs. Note that leaking batteries and leaking electrolytic capacitors also cause problems.
4, thermal damage - this comes as two types of problem. Overheating, and thermal stress (caused by temperature cycling). Silicon transistors and chips all have a maximum temperature rating. Operation above this level can lead to damage caused by thermal runaway. With digital chips the faster the logic gates in the chip switch, the great the amount of heat generated. The problem is getting the heat away from the silicon core or die (or "chip"!). To aid the dissipation of the heat, either a heatsink or a fan can be used (or both). Both draw the heat out of and away from the casing of the chip thereby reducing the rate of temperature rise within the chip casing. Important things to note are: if using a heatsink, ensure it is big enough to dissipate enough heat, and that the thermal resistance between the heatsink and the case of the chip is low. So use heatsink compound (which is a white paste that fills in all the little holes so maximising contact and keeping the thermal resistance as low as possible). When using a fan to cool chips, it is the path of the airflow that is important (more so than the temperature of the air, although cool air is much better than hot air!). So if the designer used fans in the design, operate the equipment with those fans running and with everything fitted correctly in its case. The designer should have arranged the cooling system (of which the case is a part) to draw air over all the components that need cooling.
How hot are your "hot" chips? The MPU 68030 and the two DIL chips U3 and U4 are all running at 32MHz so are always going to run warm. Are they too hot to hold your finger on for more than 20 seconds? If you can hold your finger on them for 30 seconds then I would suggest that if the cooling fans are effective, no additional heatsink is needed. But note that this is general advice as I do not have a TT myself. As long as the heatsinks actually help remove heat and don't get in the way of the case metalwork, then I can not see any problems with this.
Electrolytic CapacitorsThese are the cylindrical types in metal cans with either metal or plastic tops.
Strangely enough, they suffer from the same kind of problems, plus natural aging. The main things that kill them are:
1, reverse polarity
2, over-voltage
3, physical damage
4, thermal damage
For 1 and 2 depending on the magnitude of the applied voltage, either they do not work correctly, age quickly or go bang! shooting out of their case and attempting to make orbit leaving a nasty mess in their wake... So ensure any replacements are fitted with the correct polarity, and that they have the same or greater voltage rating than the component they are replacing.
3, If you can see any damage or if there is any deformation, replace.
4, Most capacitors have a temperature rating of 85 degress C. The higher the temperature, the quicker electrolytic capacitors age. The heat is both from the air, nearby hot components and from within the capacitor itself. For the PSU, I recommend that you fit good quality ones, buy replacements that have the same capacitance (in uF), have a voltage rating equal or greater, are 105 degree C rated and are low ESR types (which are designed for use in switch mode PSU's, which is what these Atari PSU's are). And of course they have to actually fit on the PCB!
As Atari hardware is over 20 years old, I recommend renewing all the electrolytic capacitors in the PSU at the very least. You may want to renew the ones on the main board, but failure of these normally just degrades the performance of the relevant section of circuitry rather than being a show stopper. Where ever a capacitor is near a hot or warm component fit a 105 degress C type.
Most other components have a long life and so leave them well alone. The only other issue is the solder. This ages over time. So check the solder joints on the PSU (important). If you have time, do the same on large components on the main board (less important).
Hope this helps

Falcon, Atari 520STFM, Atari 1040STE, more PC's than I care to count and an assortment of 8 bit micros (nearly forgot the Psion's)