Superregen (a partial solution?)

[mstechca]

I was looking at a circuit on the web, and I wanted to adapt my superregen to it. It seems that the results are satisfactory, but at a low amplitude.

The circuit I looked at can be found at the following URL:

**broken link removed**

Basically, I took this circuit, and converted it into a common-base format, so that it can operate in VHF mode.

Here is how I modified the circuit.

The components I have omitted in my version are the PNP transistor, 0.1uF capacitor, and the output components (earphone and 10K).

The 100pF is grounded. (the tank is replaced by a short)

L1 and the 475pF is connected to the emitter. The components that were there are now omitted.

L2 and L1 are set to 0.1uF in my circuit.
RFC is 1uH in my circuit.

The variable capacitors in my circuit have no definite range because I'm looking for the best signal.

In my circuit:
I changed the 36K resistor to 18K.
I changed 560K resistor to 120K.
I omitted the grounded 120K so I can obtain maximum range.

When I added a 1.2pF capacitor between emitter and collector (feedback), I was able to pick up CHTV (air channel 11), but the volume was way too low.

Now here are the questions: (you can ignore my circuit.)

the 200pF trim capacitor in the original circuit must form part of the quench control. Is L2 the other component that adjusts this quench frequency or is there more to it?

The 475pF in the original circuit appears to select a station. Would I be better off to connect the capacitor between emitter and +ve instead of emitter and ground?

Is L1 the ONLY other component besides the 475pF that affects the incoming station frequency (carrier)?

Because I'm dealing with VHF/UHF, a PF makes a big difference. Is there any other component (including capacitors in transistors) that can directly affect the frequency? if so, how do they?

Sources tell me that the carrier frequency must be a multiple of the quench frequency. Also sources tell me that the quench frequency must be higher than the carrier frequency.

I want to be able to get the most gain out of this stage, and be able to use one trim capacitor to tune several dozen frequencies (or more).

 

[audioguru]

It is a regen receiver that doesn't have a quench oscillation. You must manually adjust its amount of regeneration with the 200pF adjustable cap.

A super-regen circuit has a quench oscillator in its design. The quench oscillation is the super part. When the quench oscillation is setup properly then the RF oscillator is at the threshold of oscillation when the quench cuts-off the transistor and the signal builds up in it and quenches it over and over.
The quench frequency is ultrasonic, about 100kHz. The RF oscillator's frequency is a few thousand times higher, isn't it?

There are plenty of VHF super-regen circuitrs on the web and you have posted them. So now why bother with modifying a low frequency regen circuit?

 

[mstechca]

This is the closest match on the web to my design. Also a transistorized ultra-audion oscillator I found in a book is the closest match as well. What I am ultimately seeking is a way to adjust several dozen Mhz using only one trimmer capacitor not two, AND be able to pick up CHTV (channel 11) at a decent amplitude.

My previous design worked, but the problem is that I had to adjust two capacitors, every time I want to change the frequency.
These capacitors are the positive feedback one, and the one in the emitter tank.

Is there a way I can do it?

 

[Nigel Goodwin]

This is the closest match on the web to my design. Also a transistorized ultra-audion oscillator I found in a book is the closest match as well. What I am ultimately seeking is a way to adjust several dozen Mhz using only one trimmer capacitor not two, AND be able to pick up CHTV (channel 11) at a decent amplitude.

My previous design worked, but the problem is that I had to adjust two capacitors, every time I want to change the frequency.
These capacitors are the positive feedback one, and the one in the emitter tank.

Is there a way I can do it?

Yes, build a proper super-regen, they only have a single tuning capacitor - if you had done that a year ago you would have had a working radio all this time!.

 

[audioguru]

MStechca,
Haven't you noticed that the tank is at the transistor's collector on all other super-regen designs? The tank needs a high impedance load which is at the collector of a transistor. The emitter is a low impedance. The other designs have the quench oscillator parts at the transistor's emitter but your circuit doesn't have quench oscillation.

 

[mstechca]

My latest circuit can be located at: **broken link removed**

I am unable to post attachments because the "manage attachments" button does not work in Opera 7.5.

anyways, what I forgot to mention in the circuit is that the junction between L1 and R2 is my output.

I got somewhat more lucky.

I managed to make L2 and my varactor arrangment (c3, transistor 2 and the tuning input) a tank circuit that can select the desired frequency.

I still have to figure out the best value for C2. I think some receivers have a "throttle" or quench control, and I think that is what it is.

What I really need is an equation that can help me determine C2.

Maybe I can adjust this circuit so that C2 can vary automatically and be always correct as the tank capacitance changes.

 

[audioguru]

MStechca,
A common-base oscillator works by C2 coupling a positive feedback signal from the transistor's collector to its emitter. Your circuit doesn't.

The varactor diode's capacitance connects to the high internal resistance of the battery and not to the tuning coil because your circuit doesn't have a ceramic disc supply bypass capacitor.

You are receiving the audio of a local very powerful TV station. I think your random circuit is operating only as a crystal radio, which is just a tuned circuit and an AM detector diode.

 

[mstechca]

Um, audioguru, I deliberately did that to the tank circuit because I want to eventually have a push button adjust the stations. I remember that you told me that a poor-mans varactor can be made with a reverse-bias diode and a high value capacitor in series. I did just that. The transistor is wired as a diode. for this purpose.

Is a capacitor still required even though I am using a varactor to generate capacitance?

If I add that coupling capacitor from emitter to collector again, then the frequency will change, and more math will come into play. If someone could explain the math (even if I decide to add the capacitor in the future), I would be happy. I think it has to do with neutralization, but I am not sure.

 

[Nigel Goodwin]

If I add that coupling capacitor from emitter to collector again, then the frequency will change, and more math will come into play. If someone could explain the math (even if I decide to add the capacitor in the future), I would be happy.

You've been bodging what used to be a super-regen for three years now and you're still no closer getting it working, simply because you never pay any attention to the answers you get. We've been telling you for those three years that you can't calculate the values, because you don't have any way to know what values you are using - and it your case, with two coils?, it's even more impossible to calculate values! (because it isn't really anything!).

The circuit you posted bears no resemblance to a super-regen, so it's not suprising it doesn't work - why not build a super-regen?, there are plenty of circuits on the net, and those will work - just build them as published, don't make random pointless changes!.

I think it has to do with neutralization, but I am not sure.

I see you've found another old word that you don't understand

 

[audioguru]

I think it is time you tried making a real super-regen receiver. Not an ordinary regen and not this random pile of parts.
It will probably get overloaded by your very strong local TV station.

 

[mstechca]

MStechca,
A common-base oscillator works by C2 coupling a positive feedback signal from the transistor's collector to its emitter. Your circuit doesn't.

I am taking that into consideration. I have C2 grounded, because the circuit I looked at also has it grounded. The problem is if I make C2 the feedback like you suggest, and I have no capacitor from collector to ground, the circuit won't work well for me.

In one diagram I looked at (probably Charles Kitchin's receiver), It claimed that connecting the capacitor from the drain of a FET to ground would make that capacitor a feedback capacitor.

The varactor diode's capacitance connects to the high internal resistance of the battery and not to the tuning coil because your circuit doesn't have a
ceramic disc supply bypass capacitor.

Let's ignore the varactor then. That will happen in the future.

The only things that are random in my detector are the capacitor and inductor values. HOWEVER, I made both inductors 0.1uH because I want high frequency, and because I think that I can achieve impedance matching through the same detector, using the inductors.

Let me tell you exactly what kind of mess I pick up.
I make the assumption that the tank at the emitter is the real "tank" that adjusts the carrier frequency, so I select a random capacitor value. (just because I want to pick a frequency, and I can't determine the capacitance of the trim capacitor). The inductor is at 0.1uH.

Now here is the interesting part. I changed the capacitor from base to ground to a ridiculously large value (22uF).

Now when I play with the capacitor value from collector to ground, I get interesting results. If I have it adjusted to a certain value (I can't measure capacitors ), I get white noise, and sometimes a faint voice behind the noise. If I adjust the capacitance one way too much, I get oscillatons. It seems that the 22uF capacitor helps determines the frequency of these oscillations. If I adjust the capacitance the other way too much, the white noise is reduced, I don't get a signal, and eventually, all I get is background hum.

It makes me think that the variable capacitor that I played with (see above paragraph) controls the quench frequency.

So if I instead connect the capacitor between emitter and collector instead, and I still want a decent signal from any VHF station ( > 108Mhz), what math is involved in calculating the two small capacitor values and the inductor values?

I need equations, so that I won't have to continue to play with trim capacitors.

Audioguru, I do understand why you suggest that I connect a capacitor between emitter and collector, but I do not understand why C2 must be that capacitor. (remove capacitor from collector to ground).

 

[audioguru]

Why not re-make a real super-regen that you tried a long time ago before you knew about stray capacitance that messes-up high frequency circuits and the importance of a low impedance supply bypass capacitor.

Andersen has a proper way of doing it:
1) The high impedance tank is at the transistor's high impedance collector.
2) A capacitor from collector to emitter provides positive feedback.
3) A few parts at the emitter provide a quench oscillator.
4) An RF amplifier isolates the antenna from the RF oscillator.

 

[bryan1]

TAP TAP and a hollow sounds echo's out of Mstensha's head as pointed out above just build a proven circuit that works. Ever since I've been visiting this forum you've been flogging that bit of junk you call a super regen and no matter how much people try to helps it goes in 1 ear and out the other. And you wonder why me and other people on this forum bag you and your posts. Put some grey matter back between your ears and just do something properly for a change instead of trying to be the kooky professor all the time as it's just not working!!!

 

[mstechca]

Listen.

I managed to get a signal using every combination including Audioguru's.

what I have been asking for for the longest time is an E-Q-U-A-T-I-O-N.

You know, some math!

Something that allows me to determine the correct frequency based on the components.

It isn't as simple as 1/(2 * pi * sqr(LC)) because I have to somehow mix in the feedback capacitor into the equation!

I guess no one here knows that answer. oh well

 

[audioguru]

An "equation" for a very small value capacitor that is used for positive feedback doesn't work. Stray capacitance and the capacitance of the transistor are up to double the capacitor's value and are unknown.
All VHF circuits like this use a capacitance of from 4.7pF to 10pF and they all operate fine.

If the value of the capacitor is a little too much, then the positive feedback is a little higher than necessary and the circuit is able to oscillate down to a lower frequency. If the value of the capacitor is a little too low, then the circuit will not oscillate at a frequency as low. So what.

The value of the capacitor that is used for positive feedback affects the tank's tuned frequency a little. So what, the tank must be tuned anyway and the value of the capacitor to tune it isn't a linear function of the frequency, and stray capacitance across the tank affects its frequency more anyway.

 

[Nigel Goodwin]

what I have been asking for for the longest time is an E-Q-U-A-T-I-O-N.

You know, some math!

Something that allows me to determine the correct frequency based on the components.

It isn't as simple as 1/(2 * pi * sqr(LC)) because I have to somehow mix in the feedback capacitor into the equation!

I guess no one here knows that answer. oh well

We've been telling you repeatedly for three years now!, the formula above is the one you need - BUT you can't use it effectively because you've no way of determining the values of L and C - there's also NO NEED TO CALCULATE THE TUNING FREQUENCY.

Note though that you would have a far better chance with a correctly designed circuit, with a random collection of components you can't calculate anything!.

Oh, other formulas you require are simple series and parallel capacitance and inductance, but it still won't help!.

 

[Hero999]

I changed the capacitor from base to ground to a ridiculously large value (22uF).

Which is probably behaving like an inductor at that frequency.

 

[mstechca]

the formula above is the one you need - BUT you can't use it effectively because you've no way of determining the values of L and C - there's also NO NEED TO CALCULATE THE TUNING FREQUENCY.

Note though that you would have a far better chance with a correctly designed circuit, with a random collection of components you can't calculate anything!.

I discovered something. When I tie a small value capacitor to the pull-up resistor, the frequency changes. if the value is large enough, audio is chopped up. This makes me think that this part forms the quench control.

I'll take the LC circuit in the emitter as is. Now the only two compnents that I still have to figure out is the positive feeback capacitor (capacitor between emitter and collector), and the inductor between the pull-up resistor and collector.

If there is an equation that can determine that, I'll be happy.

Now I see that my circuit is based on 3 equations:
The carrier frequency equation (determined by the LC tank) that Nigel claims he told me for 3 years.
The quench frequency could be determined using an RC equation on the pull-up resistor and the new capacitor in parallel with it.
AND
another equation that adjusts something (optimal feedback?) based on the emitter-collector capacitor and the inductor between the pull-up resistor and collector.

 

[audioguru]

I discovered something. When I tie a small value capacitor to the pull-up resistor, the frequency changes.

Of course. The inductor at the collector is the tank coil with some stray capacitance across it.

I'll take the LC circuit in the emitter as is.

Why? An LC tank is at a transistor oscillator's collector, not at its emitter.
A transistor is cutoff by a voltage change between its base and emitter. That is where the parts should be located for the quench oscillator.

You have the parts in your circuit mixed-up and backwards.

 

[mstechca]

Of course. The inductor at the collector is the tank coil with some stray capacitance across it.

Thats different. When I went to:

**broken link removed**

and looked at his circuit, he states:

Q1's emitter has a 6 turn RF choke connected to it; below that choke are a 10k resistor to ground and a .001uf cap across that resistor. These 3 components form a subcircuit that makes the superregen unique: the Quench Oscillator. (The 10k resistor is also the emitter load across which we take audio which is sent to the second transistor, Q2, for amplification).
...
The .001uF cap across the 10K resistor forms an RC time-constant circuit that sets the "quenching frequency".

He is agreeing with me.
I think it is just two components (resistor and capacitor), and you are saying it is the capacitor and inductor.

My resistor capacitor and inductor are connected together the same way. except the other end of the inductor is connected to collector and the resistor and capacitor are pull-ups instead.

and I have another question. When the emitter inductor is at 0.1uH I hear more "hum". When I replaced it with a 1uH inductor, I hear more stations. I want a low value instead, because I want to enter the aircraft and TV frequencies.

I wonder if the breadboard is causing difficulties, because I'm doing alot of my testing on a breadboard.