Thyristor current limitation using secondary windings on buck conv. Inductor?

[Chris56000]

Hi!

I've just completed the repair of an industrial buck converter power supply providing 5V at 8A from a 24V supply, that used a very novel, rapid and safe method of foldback current limitation.

The circuit employs a combined current/control transformer with a large centre–tapped secondary winding, across the outer end of this is a small 200V 1.5A TO–5 thyistor.

The cathode of the thyristor connects to the finish of the current transforme secondary and then to the "Current Sense" pin of a 723 regulator. The gate of the thyristor is fed with current fed from the collector of an "upside down" PNP transistor whose emitter connects to the centre–tap of the current transformer secondary and a positive power–supply point, whilst the transistor's base is connected to the wiper of a low value preset across the 12V auxiliary power supply line.

I found, in the course of testing this power unit after repair, that this whole circuitry of the unit  gives it a fast and effective "foldback" current limitation that is completely safe in operation, but I can't find the theory of it's operation documented anywhere!

Has anybody come across this technique before and where can I found an account of it's operation?

Chris Williams

[T3sl4co1l]

No idea.  I get parts of what you describe, but there's too much additional information lacking, to read a schematic from it unambiguously (tl;dr: could you draw it?).

I don't get what purpose a 723 is supposed to serve here, unless it's one of those early bastard switchers (ab)using standard components in nonstandard ways.  (There were several SMPS circuits using 723, and yes even LM317.)  These usually perform poorly.  But the fact that an SCR is fast enough to manipulate the current sense transformer... is consistent with that.

There's not usually much point to foldback current limiting in a buck converter, but it's neat from time to time.  (Hiccup limiting is very common these days, partly because it implements an aggressive foldback, but also because it's almost trivial to implement -- it's almost unavoidable when the regulator is self-powered, as most offline converters do.)

It might also be mandatory with other compromises in the circuit.  For example, a hysteretic controller won't be very well behaved at low output voltages, for which a foldback characteristic could be used to tame it.

In any case, though I don't have a feel for the particulars of this circuit, it's entirely possible that an SCR might be used as a combination comparator and sense amp, and yes, that could be used to implement limiting or foldback, when wired appropriately.

Tim

[Chris56000]

Hi!

Schematic now added! Does anybody recognise it?

Chris Williams

[Chris56000]

Hi!

Further to the above, the 723 shown on my circuit extract (I can't show it in full or mention what unit it's for as it's commercially sensitive and repaired as part of my employment) is used in tandem with a second used as a switching–regulator, but I did find that the regulation characteristics of the circuit are consummate with a good quality lab supply or a product from somebody like Coutant Lambda!

The circuit extract was drawn out by me from a scrap board of the same type, and I'm satisfied I've not got it wrong!

Chris Williams