TheRadioBoard

A number of people have found that the Simple Kitchin Regen circuit will work better if the collector were tapped down on the coil, like this example:

ftp://ftp.sonic.net/pub/users/amckenna/ ... _Rcvr_.pdf

Norman
KA1GUK

Norm...

I am not trying to give anyone a hard time, but a "regen" detector uses feedback. In the case of the last Kitchin you posted, there is L3, the tickler coil. It provides a path for drain current to influence gate current. This increases the gain of the stage and causes oscillation. I recently built this circuit and it works fine.

The original Japanese circuit uses the first FET as a simple RF amplifier. The diode is the detector.

If someone can show me a feedback path (naming which components are involved) in the original Japanese circuit, I will be glad to change my mind.

The originally referenced Japanese circuit seems to be a "crystal set" with an RF amplifier ahead of it and an audio amplifier following.

Rich

_________________
Homebrew Radio ex-Silicon Valley

Hi Rich.

I've assumed that the feedback path is:

- the pot in the source lead,pt 8, has RF & AF;
- this is fed through the 47pF capacitor to the tickler coil, pts 1-2;
- to ground (-) at pt 3.
- The tickler coil is coupled to the detector coil, pts 3-4.
- and the JFET is acting as a regenerative amplifier.

Since I'm new to this - I could be full of beans.

73, Dan

Congratulations!

I was assuming 1 and 2 to be an antenna coupling coil. The schematic notation seems pretty strange. Vr center rotor actually does get connected to the first coil and the antenna is directly coupled to the the Vc.

OK, now we have a feedback path and the circuit can be a regen!

Strange to see that 47 pF cap in series with the tickler, but if it works... what the heck!

Sorry for adding to or creating confusion....

So, performance may be improved by making sure that the coils are phased correcly, or by changing that 47 pF cap to some other value. If it oscillates only at the hi end of the band, increasing the 47 pF to a higher value might give you more feedback at the low end?

Rich

Rich

_________________
Homebrew Radio ex-Silicon Valley

Hi Rich.

The commercial coil(s) he substituted probably have different coupling than the hand wound ones too.

We'll see what he finds. Hope he gets it to work.

73, Dan

Well, I may as well jump in here too.

First of all, welcome to the forum Qrp-gaijin.

After reading the article, it does claim to be a regenerative receiver. If so, then it appears to me that the feedback is due to the drain-gate capacitance. The old triode tube RF amps would oscillate for this reason if they weren't neutralized. Varying the source resistance will adjust the gain until it goes into oscillation. It doesn't look like a particularly good circuit by any means, though.

The way the variable source resistor is drawn shows the wiper unconnected, but I think the intention is that it's connected as a two terminal variable resistor. So, the wiper would be connected to one end.


Edit:
Oops, I missed Dan's first post. Looking at the circuit again, it looks like Dan's got it figured out.

Hey, I built that! The original Electronics Now version, that is.

At the time, I wasn't on the internet and knew nothing about Q and coils, or even transistor capacitance. I had to use different capacitors for different parts of the band. I thought it was because of how I wound the coil. Thanks to this thread, now I understand it may be the design. It worked, though, and now I understand why I had to switch in different caps. Here are pictures of it.





Macrohenry

Nice looking radio as usual, Macrohenry.

Norman
KA1GUK

Hi folks,

Thanks for all the highly interesting discussions and advice. First let me confirm the circuit: point "2" on the tickler coil should indeed go to ground, for instance by connecting it to point 3, right?

OK, to try to improve the oscillation I think I'm first going to try increasing the FET source resistance:



If I add an additional fixed resistor, should it be between the ground and the pot, or between the pot and the source, or does it not matter?

What's the idea behind this? Is it to force more current to flow back through the wiper to the input, thus increasing regeneration?

It's going to be the weekend before I get to this - and I also need to improve my desoldering skills before I can change the circuit around. Next time I'll use a solderless breadboard for my experiments... but for this first time, I wanted to build it exactly as the original designer had laid it out.

I'll post here if I get it working better.

Thanks again for the great advice and information!

It doesn't matter which side of the pot the resistor is connected to. The idea is to increase the gain. This article may help. http://www.w7zoi.net/jfet101.pdf

Tune the radio to the low end of the band and adjust the resistance so the circuit oscillates with the pot almost fully advanced. You'll be able to back off on the pot as you tune up the band.

BTW, connecting a digital display to a regen is considered heresy by some!


73,

_________________

http://kr1s.kearman.com/
http://qrp.kearman.com/

Good luck qrp-gaijin, keep us informed.

Just to be nit-picky ----

- increasing the source resistance - decreases the gain (gm) & drain current;
- decreasing the source resistance - increases the gain (gm) & drain current.

73, Dan

it was pointed out to me, and stated in the article as well, that the feedback is taken from the center tab on the Rs pot and send it back to the primary side of the coil -> which looks like an antenna on the schematic.

so it is indeed a regen.

Exactly, and no source resistor at all will give you the most gain without regeneration, though regeneration gives you the best gain of all! If you add the extra resistor, let us know if it helps the regeneration.

Another idea: Using an RF choke between the battery and the FET drain, rather than a resistor will help increase the gain of the RF amp stage. The resistor helps to keep the RF where it belongs, but it also cuts down the voltage available to the FET. An RF choke is much more efficient. See the diagram:



Norman
KA1GUK

Just a quick thought. When regeneration started becomming popular in the late teens and early 1920's, some of the circuits did not use tickler coils and such. They put a variometer in the plate circuit and this tuned the plate to the grid circuit, and being due to the feedback capacitance of the triode tubes, the sets would oscillate just fine. In fact, sometimes the plate variometer was a couple feet away from the grid circuit.
Curt

_________________
Curt, N7AH
CW forever
Connoisseur of the cold 807

an RF choke is essentially a variable resistor: it has a huge ac impedance (thus improves ac gain) but low dc resistance. This way, you can run the rf stage at a low rail voltage without lowering the gain.