What is also interesting about the noisy regen is it is a negative resistance circuit. An increase in the voltage at where the tuned circuit is connected to the collector of transistor A and base of transistor B causes an increase in current flow through transistor B and cuts off the current flow through transistor A. Therefore the current flow at collector A falls with increasing voltage which is the opposite of positive resistance.
I noticed something interesting in LTspice when simulating this circuit. I'm not sure if it's related to the above comment, but what I observed was that in some cases, at higher frequencies, the noisy regen can oscillate even if the common-base transistor in the pair is separately biased to cut-off, or even if the common-base transistor is removed completely. I think that only way that oscillation can happen in this case is for the remaining common-collector (emitter follower) stage to have parasitic in-phase feedback via parasitic base-emitter capacitance, causing it to oscillate.
In other words, it may be that this parasitic base-emitter capacitance of the common-collector (emitter follower) transistor in the cross-coupled oscillator is always contributing to the tendency for oscillation, independently of the "main" feedback path through the common-base transistor.
It might be interesting to build a regen with a transistor operating in that region. It's possible you might directly get smooth regeneration control down there. FETs I seem to remember have a linear response mode at very low source dain voltages but I never got that to work properly, if you could find that zone and there was enough gain you would get very smooth regeneration control.
I think another method, that I haven't tried in a regen, is to swap the emitter and collector of the BJT for lower gain. That might also give smooth regen control. LTspice can correctly simulate a BJT in this mode.
I used a BJT with collector and emitter swapped as part of a 1.2-volt AGC circuit. The "reversed" BJT acted as a variable resistance, forming the bottom half of a voltage divider. When I used the BJT in "normal" mode, the gain was too high and the transistor reacted too suddenly to base bias changes. In contrast, in "reversed" mode, the BJT reacted more slowly. This was confirmed both in LTspice and in hardware. Some notes here: https://qrp-gaijin.blogspot.com/2015/09/12-volt-audio-based-agc-for.html .
Â
For your amusement, here is a differential pair/cross-coupled oscillator that is oscillating with one of the BJTs connected in reversed fashion. It might have smoother regen control due to the lower gain.
The above circuit wouldn't oscillate at supply voltages of 1.2 volts or lower, indicating that the gain is indeed lower in this configuration.
The circuit will also oscillate if the Q2 C/E connections are normal, but the Q1 C/E connections are swapped.
The circuit has difficulty oscillating, however, if the C/E connections on both Q1 and Q2 are swapped, probably because the gain is too low overall. But it can be made to oscillate under some circumstances, if the L/C ratio is high enough.
In this configuration, you can see that the waveforms quickly become distorted, I think due to the fact that the transistors are quickly becoming saturated at low voltages, which is a characteristic of using BJTs in this reversed fashion.
So there we have it -- an "upside-down" cross-coupled oscillator! Is it useful -- who knows? 😀
Here is an improved version of the noisy regen which consists of an AC-linked differential pair Q-Multiplier in which the regen control is done by changing the base bias current of both Q1 and Q2. R5 is the course regen control, R6 is the fine regen control. The current mirror detector circuit Q3-Q5 can also be connected at the top of the tank circuit L1/C4. Circuit has been drawn in DigiKey's Scheme-it at https://www.digikey.nl/en/schemeit/project.
Advantages of this circuit:
Q1 and Q2 don't have to be equally matched, because each BJT has its own independent bias. This will give a smooth and backlash-free regen control;
The base-collector junctions of both Q1 and Q2 will always be reverse biased, which reduces loading of the tank circuit L1/C4 and causes less phase noise when the Q-Multiplier is oscillating while detecting SSB and CW;
The tank circuit L1/C4 has been lossely conected to the base of Q2 by using a small coupling capacitor C5, which also reduces loading of the tank circuit and decreases phase noise.
The increased collector emitter reverse bias would reduce the collector emitter capacitance as you can see in many transistor datasheet. I understand people only want to use an untapped turing coil to allow easy bandswitching. Still I feel if you had a tap on L1 maybe 1/4 to 1/2 the way up and connected the collector of Q1 to the tap and C5 (which could be increased in value) to the tap you would get better impedance matching and also reduce the negative effects of semiconductor junction capacitances.
There are choices to be made. If someone wants to use the AC-linked differential pair Q-Multiplier that is a good choice too.
At the moment though I want to experiment with the noisy regen Q multiplier and even one simple reason is I'm not set-up to prototype with more than a few components.
The price you pay for using the noisy regen circuit is:
Increased collector emitter capacitance (as you can see from most datasheets).
Somewhat reduced gain from operating somewhat near the saturation voltage.
Somewhat reduced collector impedance for the same reason.
Reduced voltage headroom like you said.
Noise injection from the high value emitter resistor.
I don't know how serious the last point is, it could be important or unimportant.
If it is important then you could use a FET as a current source, in place of the emitter resistor.
A 2sk596 with its base to ground will provide a few hundred microamps which should be enough. You could use 2 if it is not. And then control regeneration by having a moveable coil couple the negative resistance into the tuned circuit. Of course you cannot use so much current that the circuit oscillates with the coupling coil. I never found any problem with that though. The phasing of the coupling coil is unimportant which is funny compared to a normal tickler coil where phasing is of red green importance.
Thank for the FM radio circuit. I would not use a superregen for VHF because I understand it is possible to get some hearing loss from high intensity ultrasonics. The circuit you indicated looks like a great alternative.
While the Noisy Regen may be a poor AM detector, but this circuit is very suitable for using as a synchronous FM detector to detect wideband FM broadcast stations. See https://www.electroschematics.com/small-fm-receiver/ "Simple FM Receiver Circuit".
If you measure the collector emitter voltage of the transistors in the noisy regen circuit it is about 0.6 or 0.65 volts which is above the collector emitter saturation voltage of small signal transistors. That is how the circuit can still work, even with what you could say is a very odd biasing technique.
As you get nearer the staturation voltage the gain of the transistor falls. And actually often the saturation voltage is defined as the point where the gain of the transistor (Hfe) has fallen to 10% of its normal value.
Even there this is some gain.
It might be interesting to build a regen with a transistor operating in that region. It's possible you might directly get smooth regeneration control down there. FETs I seem to remember have a linear response mode at very low source dain voltages but I never got that to work properly, if you could find that zone and there was enough gain you would get very smooth regeneration control.
I suppose the follwing noisy regen circuits would work and of course you get to keep the very easy band switching capability of the basic noisy regen circuit:
You can clearly see Q3 and Q4 form a current mirror at DC if you view L1 as a short. You can possibly omit C2 especially above 2 or 3 MHz. The disadvantage of the circuit is the LC resonant tank L1 C1 is directly connected to the base input impedance of Q1 which might only be 5 or 10k. And the tank is directly connected to a number of semiconductor junctions which is not good for frequency stability and SSB reception. You could put a bypass capacitor (100n) between the collector and emitter of Q3 but I left it out to avoid any RC constants that might cause any squealing effect.
Does the input impedance of the detector Q3-Q5 change when adjusting the regeneration level of the Q-Multiplier Q1-Q2? If that's true, regeneration backlash will occur.
In the world of regenerative receivers it does seem to be the only circuit that is repeatable and reliable. At the very least you get a good Q multiplier. The tricky question is how to link a proper AM detector to the circuit. What ususally happens is AM detector input impedance shifts with signal evelope disturbe the delicate balance of forces that make the circuit a good Q multiplier.
Taking a signal from the emitters is generally safe as that is at low impedance.
I agree with the above, except for the last point. In my regenerative superhet, I found that even taking the signal off of the emitter yielded poor performance, which was most easily observed by trying to listen to CW above threshold, which was impossible due to very noisy oscillation above threshold, instead of the pure sine wave audio tones that you should hear above threshold. I also noticed that just below threshold, there was an unacceptable increase of noise instead of the expected seashell-like sound caused by narrowing the bandwidth. And there was also other highly annoying and troublesome behavior like a tendency to motorboat on strong signals (although I later found out that poor voltage regulation probably also contributed greatly to the tendency for motorboating).
For a long time this puzzled me, until I finally decided to take the RF signal from a low-impedance capacitive tap on the tank. At that point, behavior was vastly improved and I could finally narrow the bandwidth below threshold and listen to clean CW above threshold.
My guess is that in my circuit, my detector stage was probably causing highly variable loading, which when connected to the emitter of the differential-pair oscillator, was probably dynamically affecting the regeneration level and hence causing unwanted excessive noise at threshold.
The really annoying thing about debugging this is that it was quite easy to misinterpret the excessive threshold noise as being a natural property of the regenerative amplifier at critical threshold, which was wrong -- the excessive threshold noise was due to bad interaction with the following detector stage.
Of course, taking the signal off of a low-impedance capacitive tap on the tank gives a much lower signal level for the detector to work with. Because my circuit was a superhet, I could simply add more IF gain in order to provide enough signal for the detector to work with. For a straight regen, you might need a large antenna (or even double regeneration) to get enough signal for the detector.
To avoid any bad interaction (fringe howl/squegging, excessive threshold noice, regeneration backlash) between detector and Q-Multiplier stage, a cascode or a differential pair RF stage between tank circuit and detector must be used. Shifting imput impedances of the AM (voltage doubling diode) detector circuit with signal envelope/regeneration level are not reflected back to the RF tank circuit in such RF amplifier stages. In any cascode or differential pair amplifier stage the output is isolated from the input by using a common base stage.
In the world of regenerative receivers it does seem to be the only circuit that is repeatable and reliable. At the very least you get a good Q multiplier. The tricky question is how to link a proper AM detector to the circuit. What ususally happens is AM detector input impedance shifts with signal evelope disturbe the delicate balance of forces that make the circuit a good Q multiplier.
Taking a signal from the emitters is generally safe as that is at low impedance. However low impedance means low signal voltage.
Taking a signal from the tuned circuit to feed into an AM detector requires great care about how the AM detector is designed.
What is also interesting about the noisy regen is it is a negative resistance circuit. An increase in the voltage at where the tuned circuit is connected to the collector of transistor A and base of transistor B causes an increase in current flow through transistor B and cuts off the current flow through transistor A. Therefore the current flow at collector A falls with increasing voltage which is the opposite of positive resistance.
Also if you couple the noisy regen circuit to a tuned circuit instead of connecting it directly to the tuned circuit it doesn't matter which way round it is coupled, which is different to how a normal tickler coil cares about turn direction.
The tricky question is how to link a proper AM detector to the circuit. What ususally happens is AM detector input impedance shifts with signal evelope disturbe the delicate balance of forces that make the circuit a good Q multiplier.Taking a signal from the emitters is generally safe as that is at low impedance. However low impedance means low signal voltage.Taking a signal from the tuned circuit to feed into an AM detector requires great care about how the AM detector is designed.Taking a signal from the emitters is generally safe as that is at low impedance. However low impedance means low signal voltage.
You can also connect an RF interim amplifier between the emitters of the MC1648-based differential pair regenerative stage and a voltage doubling diode detector. See circuit diagram below. Best pratice is using a cascode or a Long Tailed Pair RF amplifier to isolate the diode detector from the tank circuit in any regenerative receiver.
I am in component count minimization mode, for a number of reasons, including just time limitation factors (the amount of time I have available to build things). I fully agree you could buffer the output of the Q multiplier and then use say a diode detector whose input impedance varies all over the place with signal envelope.
@Sean O'Connor wrote:
I noticed something interesting in LTspice when simulating this circuit. I'm not sure if it's related to the above comment, but what I observed was that in some cases, at higher frequencies, the noisy regen can oscillate even if the common-base transistor in the pair is separately biased to cut-off, or even if the common-base transistor is removed completely. I think that only way that oscillation can happen in this case is for the remaining common-collector (emitter follower) stage to have parasitic in-phase feedback via parasitic base-emitter capacitance, causing it to oscillate.
In other words, it may be that this parasitic base-emitter capacitance of the common-collector (emitter follower) transistor in the cross-coupled oscillator is always contributing to the tendency for oscillation, independently of the "main" feedback path through the common-base transistor.
@Sean O'Connor wrote:
I think another method, that I haven't tried in a regen, is to swap the emitter and collector of the BJT for lower gain. That might also give smooth regen control. LTspice can correctly simulate a BJT in this mode.
I used a BJT with collector and emitter swapped as part of a 1.2-volt AGC circuit. The "reversed" BJT acted as a variable resistance, forming the bottom half of a voltage divider. When I used the BJT in "normal" mode, the gain was too high and the transistor reacted too suddenly to base bias changes. In contrast, in "reversed" mode, the BJT reacted more slowly. This was confirmed both in LTspice and in hardware. Some notes here: https://qrp-gaijin.blogspot.com/2015/09/12-volt-audio-based-agc-for.html .
For your amusement, here is a differential pair/cross-coupled oscillator that is oscillating with one of the BJTs connected in reversed fashion. It might have smoother regen control due to the lower gain.
The above circuit wouldn't oscillate at supply voltages of 1.2 volts or lower, indicating that the gain is indeed lower in this configuration.
The circuit will also oscillate if the Q2 C/E connections are normal, but the Q1 C/E connections are swapped.
The circuit has difficulty oscillating, however, if the C/E connections on both Q1 and Q2 are swapped, probably because the gain is too low overall. But it can be made to oscillate under some circumstances, if the L/C ratio is high enough.
In this configuration, you can see that the waveforms quickly become distorted, I think due to the fact that the transistors are quickly becoming saturated at low voltages, which is a characteristic of using BJTs in this reversed fashion.
So there we have it -- an "upside-down" cross-coupled oscillator! Is it useful -- who knows? 😀
Here is an improved version of the noisy regen which consists of an AC-linked differential pair Q-Multiplier in which the regen control is done by changing the base bias current of both Q1 and Q2. R5 is the course regen control, R6 is the fine regen control. The current mirror detector circuit Q3-Q5 can also be connected at the top of the tank circuit L1/C4. Circuit has been drawn in DigiKey's Scheme-it at https://www.digikey.nl/en/schemeit/project.
Advantages of this circuit:
Q1 and Q2 don't have to be equally matched, because each BJT has its own independent bias. This will give a smooth and backlash-free regen control;
The base-collector junctions of both Q1 and Q2 will always be reverse biased, which reduces loading of the tank circuit L1/C4 and causes less phase noise when the Q-Multiplier is oscillating while detecting SSB and CW;
The tank circuit L1/C4 has been lossely conected to the base of Q2 by using a small coupling capacitor C5, which also reduces loading of the tank circuit and decreases phase noise.
While the Noisy Regen may be a poor AM detector, but this circuit is very suitable for using as a synchronous FM detector to detect wideband FM broadcast stations. See https://www.electroschematics.com/small-fm-receiver/ "Simple FM Receiver Circuit".
Circuit diagram (Source: https://www.electroschematics.com/small-fm-receiver)
If you measure the collector emitter voltage of the transistors in the noisy regen circuit it is about 0.6 or 0.65 volts which is above the collector emitter saturation voltage of small signal transistors. That is how the circuit can still work, even with what you could say is a very odd biasing technique.
As you get nearer the staturation voltage the gain of the transistor falls. And actually often the saturation voltage is defined as the point where the gain of the transistor (Hfe) has fallen to 10% of its normal value.
Even there this is some gain.
It might be interesting to build a regen with a transistor operating in that region. It's possible you might directly get smooth regeneration control down there. FETs I seem to remember have a linear response mode at very low source dain voltages but I never got that to work properly, if you could find that zone and there was enough gain you would get very smooth regeneration control.
I suppose the follwing noisy regen circuits would work and of course you get to keep the very easy band switching capability of the basic noisy regen circuit:
You can clearly see Q3 and Q4 form a current mirror at DC if you view L1 as a short. You can possibly omit C2 especially above 2 or 3 MHz. The disadvantage of the circuit is the LC resonant tank L1 C1 is directly connected to the base input impedance of Q1 which might only be 5 or 10k. And the tank is directly connected to a number of semiconductor junctions which is not good for frequency stability and SSB reception. You could put a bypass capacitor (100n) between the collector and emitter of Q3 but I left it out to avoid any RC constants that might cause any squealing effect.
Sean O'Connor wrote:
I agree with the above, except for the last point. In my regenerative superhet, I found that even taking the signal off of the emitter yielded poor performance, which was most easily observed by trying to listen to CW above threshold, which was impossible due to very noisy oscillation above threshold, instead of the pure sine wave audio tones that you should hear above threshold. I also noticed that just below threshold, there was an unacceptable increase of noise instead of the expected seashell-like sound caused by narrowing the bandwidth. And there was also other highly annoying and troublesome behavior like a tendency to motorboat on strong signals (although I later found out that poor voltage regulation probably also contributed greatly to the tendency for motorboating).
For a long time this puzzled me, until I finally decided to take the RF signal from a low-impedance capacitive tap on the tank. At that point, behavior was vastly improved and I could finally narrow the bandwidth below threshold and listen to clean CW above threshold.
My guess is that in my circuit, my detector stage was probably causing highly variable loading, which when connected to the emitter of the differential-pair oscillator, was probably dynamically affecting the regeneration level and hence causing unwanted excessive noise at threshold.
The really annoying thing about debugging this is that it was quite easy to misinterpret the excessive threshold noise as being a natural property of the regenerative amplifier at critical threshold, which was wrong -- the excessive threshold noise was due to bad interaction with the following detector stage.
Of course, taking the signal off of a low-impedance capacitive tap on the tank gives a much lower signal level for the detector to work with. Because my circuit was a superhet, I could simply add more IF gain in order to provide enough signal for the detector to work with. For a straight regen, you might need a large antenna (or even double regeneration) to get enough signal for the detector.
Anyway, you can read about my experiences here: https://www.theradioboard.org/forum/main/comment/7e06906d-2163-430d-94bc-5161e33f2699?postId=62cd4c7ddec9c100122bee3b .
Interesting circuit. Has anyone tried it?