Hi



We had a gem last week. Converter blowing its fuse. Easy, I

thought, leaky 3055s. No load checks came up with nothing, it

behaved fine even when heated up. It only tripped when a load of a

few amps was applied. Odd. Alarm bells should have started ringing

then, but just how cunning can a mere converter be? Simple design,

741 error amp driving a TIP31 npn driving four 3055s. Replaced all

the 3055s, no change. Replaced the TIP31 (with a TIP41, didn't have

31s), no change. Checked the voltages: as the load came on, the

output from the 741 dropped as the output voltage went up, classic

symptoms of leaky output devices.



Puzzlement. Scratched collective heads. Then someone noticed that a

whistling note zipped across the workshop AM radio as the load came

on. Grabbed the cro, lo and behold there's about 5 volts p-p rf on

the output, and 12 volts on the input. The thing was a better

bloody transmitter than the radio it was feeding.



Decoupled everywhere, no change (except for a reduction of the rf

on the input side). Replaced the 741, no change. Decoupled the

bases of the driver and outputs (getting desperate by now), problem

got worse if anything. Added filtering to the feedback signal to

the 741, it went completely unstable. Disabled the current limit

transistor, nothing. Replaced a pair of protection zeners hanging

off the bases of the TIP and the 3055s (paranoid design I thought),

no change. Final cure was a teensy little .01uF across the

collector-base of the TIP driver.



Total time spent exceeded the cost of a new one. Customer got

charged a lot less, of course (we'd like to keep him...)





Slight topic change - I'm looking at stocking some transistors to

replace the 3055s in these things, but the good looking ones are

pricier enough to be noticeable in the bill when four have to be

replaced. What the heck, most of our 24 volt customers are bright

enough (or gullible enough) to believe us.



Cheers



--- PPoint 1.88