On Fri, 25 Feb 2005 10:04:41 +0200, Paul Keinanen wrote:

On 24 Feb 2005 18:49:06 -0800, "SMJ" wrote:

Anyone know where I can see a schematic of a common emitter power
supply? I've always read about using the emitter of a pass transistor
as the positive terminal output (Common collector). Taking apart this
trip lite power supply I see that they have the negative terminal
connecting to the collector of the npn pass transistors. This seems
like it would extremely complicate the regulation design. Can anyone
explain the benefits, and where I can learn more about this design
technique?

With the NPN collector as the negative output terminal, which is
usually also connected the chassis ground, you can install the TO-3
transistors directly to the chassis without insulation :-).

Seriously, the standard emitter follower configuration requires quite
a lot of base current, especially if bad transistors, such as 2N3055
are used. These have a low current gain and high Vbe even at moderate
currents. Now the question is, where does this current come from ? In
high current supplies there are usually an other emitter follower
driving the series pass transistors and this driver has also a Vbe
drop. Thus, even higher voltages are needed to drive the base current
into the driver.

The regulator needs to operate quite a few voltages higher than the
final output voltage. In addition some extra margin is needed to allow
for the voltage drop during capacitor discharging. If you take the
regulator voltage from the same capacitors as the series pass
transistor, the capacitor voltage may have to be 5-10 V higher than
the output voltage. This excessive voltage needs to be dissipated in
the series pass transistors at high currents, creating a lot of heat.

One way to avoid this, is to use a separate medium current secondary
winding to drive the regulator at a higher voltage and use a slightly
lower voltage, high current secondary winding.

To avoid this a PNP transistor is used on the positive side or a NPN
on the negative side. It may be easier to look at the situation with a
PNP in the positive side.

Just what I was going to suggest the O/P do. Apart from the earthing the neg
terminal, using a PNP pass as an afterburner makes a helluva lot of sense. I
did that on my very first home-made bench PSU where a 723 drives the external
PNP (darlington) to 15A. I chose this configuration because the available
transformer didn't provide enough headroom for cascaded emitter followers.

(and I absolutely HATE 2N3055's with a passion, for the reasons you state - low
HFE at any sort of current worth having).

In this circuit, the current needed by the regulator flows from the
positive capacitor end through the emitter and out from the base
through the regulator to the negative end of the capacitor. Now the
regulator voltage is below both the output voltage and the capacitor
voltage and the emitter needs to be only about 1-2 V above the output
voltage. You can even drive the series pass transistor into saturation
(which requires quite a lot of base current) and hence reduce the
capacitor-output voltage difference even further, since it is now easy
to sink the base current.

This is how the LDO (Low Drop Out voltage) regulators are implemented.