For the past year or so, Swim and I have been talking about using switching regulators in place of the linear 7805s. The primary benefits are reduced heat output and increased efficiency. The suggestion originally came from ckoba,a past AtariAge member.
- Before starting any work I did some baseline heat scans. After 15 minutes, the three primary 5 volt, 1.5A regulators had reached a temperature of 170F degrees (approx 78C). The memory chips directly above the regulators were being warmed to 120-140F (50-60C).
- I first replaced all of the old electrolytic capacitors with newer, higher rated components. A resistance check confirmed no shorts were introduced by my efforts. I then removed the three primary 5v regulators and the associated input/output capacitors.
- Next I installed ceramic capacitors per the application notes. This consisted of a 10uF capacitor on the input side and a 22uF capacitor on the output side. For this exercise, Swim recommended adding a 0.1uF ceramic capacitor in parallel with the 22uF on the output side.
- I reviewed my solder work, tested continuity again for any mistakes, checked for input/output orientation.
- Next - the smoke test. I removed all cards from the PEB and inserted the Geneve. I turned on the PEB and was greeted with the standard bootup swan. Next, I added a floppy controller and proceeded to boot the OS. I left the system running for an hour. There were no visible signs that the switching regulators were causing any problems.
- A heat scan showed the temperature of the regulators holding below 90F (32C). I then scanned the chips and found that most were close to the same temperature with exception of the V9938. It was running at about 120F (48C). I expect this is normal as I did not check its temp before the regulator change.
There are a few more things I want to update, such as removal of the 12v regulator (no need for it in my opinion) and testing whether or not the case will still close properly with the slightly higher regulator form factor.
For a longer burn-in I may replace the heatwave Geneve with this modified Geneve. Doing so will help me to confirm whether there is a noticeable difference in heat and operation between a PEB power supply and a PC ATX power supply (with 12v tied to the 5v line). I suppose I should pull out my KillaWatt to compare power draw, to see if there is any difference in power use with the more efficient regulators.
Swim and I still need to do some more testing before feeling comfortable with this modification. However, early results are positive. Here is a copy of the regulator data sheet. We'd be happy for anyone to look things over to make sure there aren't any "gotchas" that we haven't considered. I've included pictures of the updated Geneve for reference.