My sx100 has been modified by PO to include a product detector using a 3 transistor circuit. Anyone have info on this mod? Tnx

I can't find any reference to a 3 transistor circuit, but I did find references to a 2 transistor product detector that was used in several receivers. https://www.qsl.net/w/wb4kdi//Heathkit/ ... ector.html

Now this requires a negative 6 to 12 V DC supply. So to use this circuit you might add a rectifier diode off of the 6 VAC filament supply, a couple of filter caps, a zener and a transistor as a voltage regulator. That would add up to 3 transistors.

Curtis Eickerman

Actually, it shouldn't need to be regulated at all, since it has nothing to do with frequency stability and small gain shifts will be all but unnoticeable. But Curt is right: an included regulator might be the third transistor's purpose.

SteveH

Hmmm. Might try one of these in my HQ-145A. It would be easy enough to add - and remove.

And if you used NPN transistors, you could tap a few ma. of B+ through a dropping resistor and a zener.

John

The third transistor could be a low gain amplifier to equalize output between the product and stock detectors.

Rodger WQ9E

Looks like a solid state
version of the Pullen mixer.

I was just going to say that. A couple of 2N3904's should do it.

Or 2N2222As. I have a few of both. They are like 1N4007s. You just never know when you will need a couple.

I wonder if it would be worthwhile to try to match a pair.

John

Or, an RCA CA 3028A should be easy to configure as a mixer as well... but I suspect they are scarce these days.

Lots of circuits like that got by without matched transistors. Or at least nobody fussed.

If you're going to use a CA3028, might as well just find an existing circuit for a product detector that used it, they were common at one time.

There was a time when you could get two transistors in one package.

Or RCA, at the least, had ICs that were just a few transistors. I think even one that had two differential amplifiers, well the transistors in that arrangement. CA3046 comes to mind but I may be wrong.

But most of that is long in the past.

If the PO's circuit is actually the 2-transistor PD or similar shown from the referenced web page, matching most likely will have no advantage at all, since the gain of one side is varied. And besides, the BFO and IF signals aren't even the same level, let alone the same source impedance. A pair of NPNs that can see lightning and hear thunder should suffice

SteveH

That’s the funniest thing I’ve heard all morning, Steve.

Circuit consists of 2 JFETs both N channel and P channel and a bipolar NPN Q. Two have inputs from IF but none from bfo. Strange. Output to s meter not audio. Will post circuit later

Since AGC is used in the SSB/CW mode, and since the BFO signal doesn't affect the AGC, I find that my SX-100 doesn't really need a product detector. I think this is because the radio has separate audio and AGC detectors, and the point at which the BFO is injected.

https://bama.edebris.com/download/halli ... -Schem.pdf

Dave

So, are you saying that transistor matching is not . . . critical?