Tell me if the outputs are mislabeled

Sorry,

It seems so. The 56 and the other resistor make about 50 ohms on the AUX ou tput. It is of course somewhat attenuated.

The main output you got 133 ohms in parallel with 616 ohms. That comes out to about 109 ohms, unless th ose transistors are conducting heavily, don't forget that.

The BC547 is obviously biased on but not into saturation. So it is not down to 133 ohms anyhow.

The output seems to be biased at about a third of Vcc, and being a source f ollower it can be considered pretty solid. The BC547 under it has no input to linearize the voltage output and the whole output stage looks like it is running off a 15 pF cap, so there is no upsetting the DC.

In my considered opinion you are right. The "aux" is 50 ohms, and what they not labelled "50 ohms" is another output at a higher voltage and impedance .

I will await those who love to call me wrong, but I am right and so are you . They labelled it wrong. :-)

That assumes the BC547 is in saturation. Otherwise the collector/100R is going to be hi-z.

Except there is 1.7V (Vled - Vbe) across 33R, implying there is approx

Yet the gates of 3N211 are at 4V. A Vgs of 0V implies the BC547 is in saturation?

I can't work out from the datasheet the FET conductance, which should provide the source follower output resistance. That might be lower than the resistive elements.

In short I don't know the answer. Certainly the output labelled aux is

I can't say I was all that right, I started thinking the label was wrong, and then just decided the output was just messed up. Now I see the labels are reversed and the actual 50 ohm output is 51.032 ohms + or - a couple hundredths of an ohm. Thanks, Mikek

The impedance of the "50 ohm" (upper) output is dominated by the transconductance of the 3N211, which is poorly specified on the data sheet. But it is probably 50 ohms or a bit less.

The AUX output impedance depends on whether the 50 ohm output is loaded, but will be about 51 if the upper is loaded with an external

The 3N211 dual gate mosfet in that circuit has a transconductance of ~20mA/V hence the 50 0hm output impedance.

going to be hi-z. "

It was kinda best case scenario. The actual Z is probably much higher than that, but that would be the minimum. Certainly not 50 ohms.

That was assumed simply to derive the best case scenario. And incidentally, if it is such a good current source how come the 100 ohm resistor ?

It can't be with that 100 ohm resistor. IF the BC547 is in saturation that is... If not the impedance is even higher.

With a 100 ohm load resistor ?

The BC547 is a constant-current sink, so the 100 ohms is mostly invisible. I'm not sure why it's there.

This is a pretty weird circuit.

invisible. I'm not sure why it's there. "

It still means it ain't 50 ohms.

If the BC transistor is in constant saturation, and the FET runs totally in class A, it could be 50 ohms but then there would be no reason for the transistor. A piece of wire would work.

If the NPN is a current sink (and it's not clear from the schematic) then Zout is 1/Gm of the mosfet paralleled by the resistive divider down to AUX. It might be around 50 ohms.

The dual-gate fet is very poorly specified, and "red LED" is not too specific either.

Looks like someone fiddled this until it worked. That's OK for a one-off.

many "50 Ohm outputs" mean they are designed to be loaded by 50 Ohms.

If the load is 50 Ohms, it won't really matter what the source Z actually is.

Mark

Not going into a transmission line.

The LED-BC547 combo with 33R emitter resistor makes a constant current sink of approx 30mA.

My guess is that 100R is to absorb some voltage headroom and reduce power dissipated in the mosfet. The mosfet is already heatsunk so the designer might have been concerned about its dissipation.

The 3N211 datasheet I saw rated Idss at 6-40mA so this circuit may have struggled with lower Idss parts!

The use of 3N211 makes me think this circuit is from a bygone age. The source follower output impedance is not well defined - as others have mentioned already.

The rest of the circuit seems not at all weird - looks like a reasonably competent ham radio type wide range RF signal generator with ALC levelling.

piglet

If the NPN is operating constant-current, the 100r resistor doesn't affect mosfet dissipation. It does reduce the dissipation of the NPN, which probably doesn't matter.

One of my pet peeves about RF signal generators is that they tend to have horrific harmonic distortion. The 100r probably makes this one worse.

You are right, the 100R is not helping the mosfet. Only other explanation I can think of is the designer intended it to be parasitic stopper like a ferrite bead?

piglet

If the far end of the transmission line is properly terminated, it still doesn't matter because there are no reflections.

It matters only if the Tline is not properly terminated on the far end.

Mark