🚀 go-pugleaf

In article <BuE4p8.5Bq@acsu.buffalo.edu>, v064mb9k@ubvmsb.cc.buffalo.edu (NEIL GANDLER) writes:
> 
>    I am a sophomore engineering student and have been researching and
> trying to develop a new type of car amplifier that runs cooler and
> the taim was at the power supply. 


>            So I tried a mechanical method using 2 high
> speed quality relays. The design amazingly worked with little problems,
> like noise from the 60-hz modulated relays, a small amount of magnetic
> intereference which can be solved and slightly warm relays. The only
> major problem I can see is reliabilty. How many cycles can a relay take
> before they burn out, thats the only question. I must do more work and
> improve on this.


That technique was actually widely used in the pre-transistor days, when vibrator invertors were used to step up 12 & 24v supplies to a level which
would drive valves (tubes). It seems to have fallen into disuse with
the arrival of power transistors.

I don't mean that to sound discouraging, it often happens that an old
technology revisited with modern techniques can turn up surprises. A
quick scan through a catalogue shows that low-power relays seem rated
at about 150,000 operations, high-power ones at 20,000,000. That is
probably limited by a combination of metal fatigue & contact burn.

Perhaps trying to find some old vibrator supplies would give parts better
suited to experimentation.

I am a bit surprised that transistors would get too hot. In a pure switching
supply they are either fully on or fully off, and even putting 20A through
a fully on transistor should only cause it to dissipate 16W or so, which
a smallish heatsink should be able to cope with.

Steve