IBM 5155 power supply
 

I have just completed a 21 page document on this switch-mode supply from the early 1980's. IBM did not publish data on it.

http://worldphaco.com/uploads/The_IB...WER_SUPPLY.pdf

It contains a range of different circuit innovations including surge limiters that use a PUT, current detectors with a current transformer, modulated base drive for the switching transistors with current feedback to improve efficiency.

One idea allows the use of a transformer with a 115V primary to be run from a 230V supply (for the 115/230V universal power input select circuit). It is a good method to synthesize what amounts to a low Z reference point that splits the incoming 230V mains in two. This idea would be suitable for running many small transformers in small 115V appliances from 230V without the need of a stepdown transformer and better than a series resistor or reactance as the method relatively independent of the load. It requires a bridge rectifier and two high value 200v electrolytic caps.

Re: IBM 5155 power supply
 

Hey Hugo,

That's another monumentally detailed piece of investigation, primary research and documentation. Thanks for making this contribution. I had never thought that viable, robust SMPS existed as early as this and it is high-powered and quite intricately thought out.

I looked at the overcurrent sense arrangements with particular interest. Really interesting to see the two conditions you describe:

*overcurrent AND

*final Vout, supply, > specified limits.

That's really ingenious. I also note the inrush limiting arrangement. I've never seen or heard of a programmable unijunction transistor before - really interesting to see an unfamiliar variant of a very useful circuit.

Cheers!