Superregen revisited

[mstechca]

I have made the following circuit highlighted in yellow (see image) with a few modifications.

The tank circuit is connected from NPN's emitter to ground
C2 is connected from base to ground
The resistors are torn off. a 160K resistor is placed between base and collector. Another one between between base and ground. L2 is connected to colector (connected to a pull-up resistor) and C5. At C5, another capacitor forms a coupling capacitor.

Basically, it looks as if I turned the whole detector into a common emitter detector.

Here's the question. What equation(s) can I use with R1, R2 and R3 to get optimal results with as many frequencies as possible?

I fixed the pull-up resistor to be about 18K, and I played with the base-collector resistor. I ignored the 3rd resistor. If I make the base-collector resistor at around 360K, it seems that I only get one station in almost perfectly across several frequencies.
If I make the base-collector resistor at around 160K, it seems that I can get a few more stations in, but the majority of them do not come in clearly.

I still think C5 and another component make a filter. Is this true? if so, what filter?

I forgot to add that my L1 is 0.15uH and my L2 is 0.1uH and my transistor is 2N3904.

 

[Nigel Goodwin]

Try drawing your modified circuit and posting it. L2 is shown as an IF transformer, that will be MANY times greater than 0.1uH. I would expect it to form a low pass filter along with C5 - so 0.1uH will be pointless!.

 

[David Bridgen]

A regenerative receiver is for a.m., not f.m.

 

[Nigel Goodwin]

A regenerative receiver is for a.m., not f.m.

You can slope detect FM, although the quality won't be good, but it's not likely to be wonderful on AM either :lol:

 

[David Bridgen]

Yes, we did that in the lab listening to analogue cell-phone traffic on a rectum paralyser.

 

[mstechca]

Try drawing your modified circuit and posting it. L2 is shown as an IF transformer, that will be MANY times greater than 0.1uH. I would expect it to form a low pass filter along with C5 - so 0.1uH will be pointless!.

I'll post the modified circuit tomorrow.
I thought there was a filter in the circuit somewhere.

The reason why I'm using ridiculously small values for inductance is because I want to receive TV Channel 12. Channel 12's towers is about 4 km away from me.

 

[Nigel Goodwin]

The reason why I'm using ridiculously small values for inductance is because I want to receive TV Channel 12. Channel 12's towers is about 4 km away from me.

L2 isn't a tuning component, so doesn't need altering.

 

[audioguru]

Mstechca,
Please post the schematic for the original super-regen receiver that you are modifying. You seem to have removed its very important emitter resistor to ground, and instead you connected the lower end of L2 to the sensitivity pot.
Also please indicate on your modified schematic where you connect the antenna.

 

[mstechca]

I flipped the whole design around. Take a look.
and Nigel, Does L2 and the capacitor connected to it form the low pass receiving frequency filter or does it form an IF filter or what?

 

[Nigel Goodwin]

You appear to be randomly moving components, without any thought for what they might do?.

I don't see how it can possibly work as you've drawn it?, nor do I see how moving the base bias resistor can be expected to 'increase stability and gain'. The two are mutually exclusive, a super regen works by being unstable (just oscillating), with a signal being enough to stop the oscillation, thus giving the maximum possible gain. Connecting the resistor to the collector will give negative feedback, preventing it oscillating.

Depending where you are thinking of placing L2 (from the first diagram, not the second!) it should form a low impedance load for audio, and a high impedance load for RF, passing the audio through it, and rejecting the RF.

But I would suggest you go back to the original circuit (before you started altering things), I doubt random changes will improve performance?.

You should also be aware that a super regen, particularly if not designed and built correctly, can cause interference to other users, and could render you liable to prosecution?. You might have noticed, the circuit is almost identical to an FM bug?.

 

[audioguru]

You must keep the sensitivity control. A super-regen is easily overloaded by local stations so the sensitivity must be reduced. On weak stations the sensitivity must be increased.
Needless to say, a super-regen isn't a real radio.

 

[mstechca]

I forgot to mention that you should forget looking at the L#'s and C#'s and R#'s in my circuit because I whipped it up.

Basically what I appears that I am doing is taking the 1st circuit's common collector design, and turning it into a common emitter design.

You appear to be randomly moving components, without any thought for what they might do?.

the only thing random I am doing is changing component values because I don't know enough equations to pick a frequency and receive from it clearly. and 1/(2 * pi * sqr(L*c)) isn't enough either.

I don't see how it can possibly work as you've drawn it?, nor do I see how moving the base bias resistor can be expected to 'increase stability and gain'.

I'll forget the base to collector change then.

...with a signal being enough to stop the oscillation.

What frequency does this signal have to be, and does the R and C (in parallel) in the original circuit control the frequency of the signal?

Depending where you are thinking of placing L2 (from the first diagram, not the second!) it should form a low impedance load for audio, and a high impedance load for RF, passing the audio through it, and rejecting the RF.

do I use Xl = 2 * pi * F * L here where F = frequency wanted and L = L2?

What impedance value should I obtain? Maybe this is why my circuit isn't working.

 

[audioguru]

You should try this superregen. Its article explains well how it works.
**broken link removed**

 

[mstechca]

that was the very first circuit I came across while designing a superregen.

What a coincidence!

I tried making my own circuit with a capacitor leading from one end of the inductor to base of the transistor. I even posted that circuit here in the past, and I didn't get a solution for it.

I guess I am on my own now on this.

 

[audioguru]

Mstechca,
You aren't on your own. We have made suggestions about your modifications and discussed of the problems of a super-regen.

As the previous project explains, it is simply an RF tuned circuit which is part of an RF oscillator, and a supersonic oscillator that stops the RF oscillation at a very fast rate which keeps its RF gain and therefore its sensitivity very high.
The circuit performs like a single tuned-circuit sensitive but noisy AM detector that overloads easily. FM transmissions can be detected poorly with slope-detection, where it is tuned to one side of the broadcasted carrier.

 

[Nigel Goodwin]

Basically what I appears that I am doing is taking the 1st circuit's common collector design, and turning it into a common emitter design.

And what makes you think doing that would work?. Circuits use specific configurations for good reasons, that's one of the 'random' changes I was refering to!.

For that matter, what makes you think the original circuit was common collector?.

 

[mstechca]

from a D.C. standpoint, the emitter of my circuit is tied to ground, and in the original circuit, the collector is tied to VCC. hence the names "Common emitter" and "common collector".

and if you want to get real precise, there might be 1/10 of an ohm between emitter and ground.

and I got one more question.

What should I set the quenching frequency to when I want to listen to TV stations? I find that a 100Khz quench rate is perfect for FM radio.

 

[audioguru]

If you want to get really precise, both circuits are operating as common base, because their base is decoupled to ground, and their collectors and emitters are free to swing. They aren't DC amplifiers.

The spec's of FM sound for TV stations are exactly the same as for FM stations. Haven't you heard the sound from channel 6 on an FM radio? It is at slightly below the FM band. In Europe they are probably different.
Adjust the quenching frequency to whatever sounds best but when it is a low multiple of the stations' spacing on the FM band, it might cause tuning problems. :lol:

 

[Nigel Goodwin]

from a D.C. standpoint, the emitter of my circuit is tied to ground, and in the original circuit, the collector is tied to VCC. hence the names "Common emitter" and "common collector".

That's where you are making your mistake, as Audioguru has rightly pointed out, it's operating in common base - generally the DC conditions for all modes are the same (or they certainly can be), it's the AC conditions which usually set the mode.

In fact circuits can be in more than one mode at the same time, the classic FM 'bug' transmitter works in common base as an RF oscillator (with feedback from collector to emitter). Yet at the same time the modulating signal is applied to the base, with the transistor 'probably' working in common emitter - it's a fairly unusual mode of operation!.

 

[mstechca]

That's where you are making your mistake, as Audioguru has rightly pointed out, it's operating in common base - generally the DC conditions for all modes are the same (or they certainly can be), it's the AC conditions which usually set the mode.

thanks for letting me know that.
If I connected a capacitor between base of the transistor and the point where one end of an inductor, capacitor and the resistor from +ve meet (this capacitor is in parallel with the base resistor carrying the +ve voltage), does this capacitor go in parallel with the capacitor from base to ground when I calculate the capacitance for a low pass filter?

if I am losing you here, I am assuming that a low pass filter is present in the circuit, but the part that is interesting is that I get somewhat better results when I add more capacitors to the circuit.