RCA WP-702A Dual PSU Teardown / Design Question

[pjhenley]

My old RCA WP-702A power supply (c. 1971) has crapped out. The voltages on each half of the supply measure fine until you put a load on them. A load on the left-hand unit causes it to oscillate, while a load on the right-hand unit causes it to drop to 0V. So've I've opened it up to see how the circuit was put together. So far as I can tell, the basics of the circuit, minus the regulation, is as in the diagram below.

I haven't looked at a lot of power supplies, but I was struck with how the AC from the transformer is being rectified. They've taken the center tap as the positive and they are half-wave rectifying the top and bottom halves of the secondary to form the negative. This gives them the same pulses they would get from a full-wave rectifier attached to the ends of the secondary (at half voltage), but saves them two diodes.

So my question is why do this? Were diodes that expensive when this was designed? Does it reduce manufacturing costs on the transformer somehow? Or are there other advantages?

[viperidae]

Less heat generated in the diodes if the current only goes through one instead of two in a full bridge.
Maybe the transformers are cheaper that way too? More turns of thinner copper.

[tooki]

Take a look at the various topologies as described here: http://www.hammondmfg.com/pdf/5c007.pdf

[floobydust]

I thought the prevalence of PNP power transistors in the late 1960's and early 1970's is the reason for the swapped location.
Are you sure you have NPN parts there? Maybe post pictures

[pjhenley]

The transistor is labeled 2N5294. Data sheets indicate it is an NPN.

[pjhenley]

Take a look at the various topologies as described here: http://www.hammondmfg.com/pdf/5c007.pdf

That is a useful chart, thanks! Looking at it, I see the similar topology, but using the diodes in the reverse direction. Most regulation seems to happen on the positive side these days, so I guess that makes sense.

[pjhenley]

I'll answer my own question: If you add a third diode, then you get split rails for your regulator circuit while retaining full-wave rectification for the power output. Otherwise, this is an inefficient design so far as the transformer heat is concerned.

[drussell]

It isn't necessarily "inefficient" but the transformer has to be designed for that type of full wave rectification.  The two-diode, double-voltage (to what you would normally think of today) centre-tapped transformer style full wave was the normal way of doing things in the tube era because you then only needed one rectifier filament winding.  When things first went solid state the transformer topology often remained the same but used silicon diodes instead of rectifier tubes.  Very common.

[Kleinstein]

It depends on the power and voltage wether the full bridge or center tapped winding and 2 diodes is more efficient. With the 2 diode solution one saves the diodes (and there power loss), but needs a little more copper on the transformer and might have a little more loss there. So low voltage and high power favor the 2 diode solution. It's also easier with 60 Hz than 50 Hz as the transformer can be smaller.

Diodes where more expensive in the 1970s and the design might be a little older.