low voltage drop rectifier

I was looking for a low voltage rectifier, and came across the "active bridge":

I only have NPNs handy so I haven't tried the circuit, but it looks like the transistors are reverse biased. I did some reading and found out they can work in reverse, but my testing indicates with a much higher voltage drop between the emitter and collector. I tested a 2n3904 and found Vce = 0.1@3mA compared to 1.3v reverse biased. A pn2222 was 0v (the meter I used only goes to 1 decimal) forward, and 1.1v in reverse.

So is there something I'm missing that would explain how the circuit would in fact work as shown?

It doesn’t look like it would work, but I tried simulating it using LTSpice, anyway. LTSpice agrees with me. I think the transistors are backwards. I’ve seen a similar circuit using MOSFETs (which the author mentions) and I know that the MOSFET circuit works.

I’ve attached the LTSpice file for the BJT version, if you want to fiddle with it.

bjt_rect.asc (1.34 KB)

If you turn the transistors around, it works pretty well. See attached. Lot of parts, though.

bjt_rect2.asc (1.34 KB)

The article explains that yes the transistors are used the "wrong way round", but
asserts with the right devices this will work... No idea if thats really true myself,
maybe if I find time I'll investigate (unlikely).

jremington:
If you turn the transistors around, it works pretty well. See attached. Lot of parts, though.

Thanks. I appreciate seeing the LTSpice output. I tried CoolSpice and found it rather limited. I just installed OSCad but haven't tested it out yet. If OSCad doesn't work out, LTSpice looks worth trying.