This is a recovered file. The images in this post may be out of order, and there may be duplicates.
One Transistor Loudspeaker Superhet! - The RadioBoard Forums
Transistors, IC's and other new fangled devices forum
User avatar
transistor495
Posts: 808
Joined: Wed Jun 30, 2010 9:54 am
Location: Kochi, India
One Transistor Loudspeaker Superhet!
So far I don't have a clue about such a circuit exists or not, what about you? :D
Well, today I again had a look on this Machael's radio, and thought how well it can work if doing a superhet conversion. Machael's radio is said to be outputting roomfilling speaker audio by just utilizing the power of single MPSA18 and offers complete AM BCB band coverage. So the challenge would be to get it working on the same level by getting into oscillation and do some IF frequency reflex/regeneration to extract audio. Finally, audio transformer is there to help us hearing on speaker :D :D
Original article link below:
One Transistor Radio
So the first step would be make the original radio working as is. Hear some programs on it and get relaxed. Then convert it into a superhet .. :mrgreen:
one transistor radio.jpg
one transistor radio.jpg (32.28 KiB) Viewed 2889 times
I would like to hear the experience of anybody built this already. There were no video footage of a working version. How good it work?
Last edited by transistor495 on Sun Dec 03, 2017 5:45 pm, edited 1 time in total.
Selenium
Posts: 1997
Joined: Mon Jun 10, 2013 7:59 pm
Re: One Transistor Loudspeaker Superhet!
Post by Selenium » Sun Dec 03, 2017 4:32 pm
Here are 2 schematics for a 1 transistor superhet receiver.
The secret to getting it to drive a speaker would be to increase the supply voltage and add an impedance matching transformer to drive a speaker.
The second one appeared on these forums some years ago.
viewtopic.php?t=2797
transistor495 wrote: ↑
Sun Dec 03, 2017 4:16 pm
There were no video footage of a working version. How good it work?
Chuck Wenzel's version.
https://www.youtube.com/watch?v=NpCs7KW936w
Attachments
download (1).png
download (1).png (8.19 KiB) Viewed 2917 times
download (2).png
download (2).png (12.52 KiB) Viewed 2914 times
User avatar
transistor495
Posts: 808
Joined: Wed Jun 30, 2010 9:54 am
Location: Kochi, India
Re: One Transistor Loudspeaker Superhet!
Post by transistor495 » Sun Dec 03, 2017 5:01 pm
Did the Thiago's design work? Because the thread looks incomplete (oh yeah, this thread also soon would be... :D )
I saw this video already, wondering why there isn't any operational footage -which is what anybody looking for?
User avatar
transistor495
Posts: 808
Joined: Wed Jun 30, 2010 9:54 am
Location: Kochi, India
Re: One Transistor Loudspeaker Superhet!
Post by transistor495 » Sun Dec 03, 2017 5:42 pm
This is a different circuit and the guy is running it with audio amplifier and external antenna too( I did watch this video till end to realize this, and I was searching for where is the pulse transformer he used , throughout the video)
Coming back to original circuit, I think that the real speaker power comes from high voltage supply -18V! So at low voltages(around or below 9v) the audio might be quite low(average reception across the band)
I also think that if the local transmitter is not very far away, the audio output could be higher.
User avatar
transistor495
Posts: 808
Joined: Wed Jun 30, 2010 9:54 am
Location: Kochi, India
Re: One Transistor Loudspeaker Superhet!
Post by transistor495 » Sun Dec 03, 2017 6:21 pm
I quickly came up with a converted design :D
I retained the pulse transformer approach to handle the RF/AF frequencies just because it's already proved as a critical part to the original circuit. The frequency response of this transformer does seem affecting the overall audio level.
Does this has a potential to work or does this approach make sense?
1Tr_superhet.png
1Tr_superhet.png (73.42 KiB) Viewed 2881 times
Selenium
Posts: 1997
Joined: Mon Jun 10, 2013 7:59 pm
Re: One Transistor Loudspeaker Superhet!
Post by Selenium » Sun Dec 03, 2017 9:08 pm
Your superhet version will likely be successful, but I question the need for the pulse transformer.
In the 1960's there was a class of 2 transistor AM receivers that were called "Boy's Radios" that received favorable North American import tax breaks.
http://www.jogis-roehrenbude.de/Transis ... scot-2.htm
https://www.google.ca/search?q=boys+rad ... 79&bih=451
They used a reflex circuit with a choke or transformer winding in the Collector that was physically orientated to the ferrite rod antenna so that some regeneration was present which dramatically increased their performance.
I don't know what advantage a 1 transistor superhet receiver would offer over a 1 transistor regen-reflex receiver which could also drive a speaker by using an impedance matching output transformer and a higher supply voltage.
The original 1 transistor reflex receiver shown in the first post could likely be simplified by eliminating the pulse transformer and using a choke instead with the diode connected to the collector and regeneration could be added. The choke may not even be necessary as the reactance of the output transformer might be high enough to enable diode reflexing from the Collector.
The choke, if used, could be oriented to provide some regeneration or a small coupling capacitor could be used.
User avatar
transistor495
Posts: 808
Joined: Wed Jun 30, 2010 9:54 am
Location: Kochi, India
Re: One Transistor Loudspeaker Superhet!
Post by transistor495 » Mon Dec 04, 2017 3:16 am
Selenium wrote: ↑
Sun Dec 03, 2017 9:08 pm
Your superhet version will likely be successful, but I question the need for the pulse transformer.
The pulse transformer used I think being a ferrite core type and could be having a large choking bandwidth due to it's typical characteristics ie, across the BCB range + IF, and their typical response towards RF specturm and AF can be favorable to improve the performance. Even the performance gain over a cycle is small, it may matter in such a regen reflex configuration where a single tranny is doing multiple jobs. When I had few experiments with normal choke in the collector I did notice the response vs frequency/bandwidth criticalities. So in this circuit, the choke in the collector should have a response ranging from IF frequency through the upper bandwidth limit of AM BCB. I would think the pulse transformer offered a best response in the original circuit. One thing is sure that the choke response is very critical over tuning range and audio level.
So when we replace the pulse tranformer with the IF transformer, below points may need to take into account:
1. Response frequency range
2. The original circuit does seem to be getting a regeneration response from pulse transformer placed in the vicinity of ferrite antenna. A shielded IFT can block this effect. An unshielded IFT may push the circuit into instability.
The above are just conceptual observations only, that can likely happen. So basically I retained the pulse transformer there and used a separate IFT to extract IF.
This pulse transformer also said to be gaining the same regenerative advantage as per the article.
So basically I thought about retaining the pulse transformer approach due to their 1) typical frequency response which may add to the overall result 2) above mentioned proximity regeneration.
I don't know what advantage a 1 transistor superhet receiver would offer over a 1 transistor regen-reflex receiver which could also drive a speaker by using an impedance matching output transformer and a higher supply voltage.
These minimal design superhets may lack/kill many 'actual' superhet purpose/advantages like multistage selected frequency amplification and selectivity/sensitivity improvements, AGC etc. However, just to mention something -the superhet may not be as 'touchy' in performance as a regen. Also if it can get into working as expected, the sensitivity of the receiver would be worth comparing with a regen.
I welcome more thoughts/improvements, more ingenious designs, practical results on this subject.
Selenium
Posts: 1997
Joined: Mon Jun 10, 2013 7:59 pm
Re: One Transistor Loudspeaker Superhet!
Post by Selenium » Mon Dec 04, 2017 4:28 am
In a previous thread in which you were a participant I posted a schematic for a 2 transistor superhet which worked very well.
It can be used with higher supply voltages by changing the mixer Emitter resistor so that about 200uA flows and increasing the value of the 22K bias resistor.
I also wanted to try a regen IF using a single IF transformer and a Polyakov regen circuit which I have used successfully in my shortwave NE602 superhets.
I suppose the next step would be adapting it to directly drive a speaker by including a impedance matching output transformer and larger supply voltage.
viewtopic.php?f=4&t=5753&hilit=superhet
Attachments
1.5V 2 diode SIMPLE 2 TRANSISTOR superhet REFLEX (1).JPG
1.5V 2 diode SIMPLE 2 TRANSISTOR superhet REFLEX (1).JPG (26.05 KiB) Viewed 2776 times
User avatar
golfguru
Posts: 5235
Joined: Sat Aug 18, 2007 8:52 pm
Location: Australia
Re: One Transistor Loudspeaker Superhet!
Post by golfguru » Mon Dec 04, 2017 5:08 am
It seems one needs about 14mW (11-12dBm) of audio to achieve a "listenable" speaker level.
Weak signals would need to be amplified to at least that level.
Hope that helps with calculations.
........
User avatar
macrohenry
Posts: 318
Joined: Tue Aug 21, 2007 5:35 pm
Re: One Transistor Loudspeaker Superhet!
Post by macrohenry » Mon Dec 04, 2017 6:31 am
transistor495 wrote: ↑
Sun Dec 03, 2017 4:16 pm
I would like to hear the experience of anybody built this already. There were no video footage of a working version. How good it work?
I built that circuit and it works very well. In fact, Michael designed his after reading about it on my webpage, which addresses some of your questions and which you can find here: http://www.tompolk.com/radios/macrohenr ... ydyne.html
On my page you can hear audio clips that demonstrate the performance. Look for the mp3s under the waveform screenshots.
Best I can tell, here's the provenance:
1950s JF Towler designed a radio that was a lot likd the two transistor like the Regency XR-2 that used one transistor as an audio amp. (see my webpage to see Towler's 1 transistor schematic)
Robert Bazian had success with a similar circuit.
Bazian introduces Charles Wenzel to the circuit, and Charles builds his, publishes it on techlib.
I discover it on techlib and use it for my contest radio. Wenzel gives me the pulse transformer.
I develop what I called the "Macrohenrydyne," making three significant modifications to the Wenzel circuit.
Michael contacts me and creates his beautiful incarnation.
Very coincidentally, Bazian and I had previously corresponded about another matter and he subsequently sent me plans he was working on for a one transistor superhet. I never learned how it worked out.
Even more coincidentally, Charles Wenzel is my neighbor. One afternoon I was at his house. He got the one-transistor radio out and I took this impromptu video. So you do get to see how well his works (listen at very beginning of video): https://www.youtube.com/watch?v=OAzEOFHfsDE
User avatar
transistor495
Posts: 808
Joined: Wed Jun 30, 2010 9:54 am
Location: Kochi, India
Re: One Transistor Loudspeaker Superhet!
Post by transistor495 » Mon Dec 04, 2017 4:16 pm
macrohenry wrote: ↑
Mon Dec 04, 2017 6:31 am
I built that circuit and it works very well. In fact, Michael designed his after reading about it on my webpage, which addresses some of your questions and which you can find here: http://www.tompolk.com/radios/macrohenr ... ydyne.html
On my page you can hear audio clips that demonstrate the performance. Look for the mp3s under the waveform screenshots.
Thanks Tom, for your radio information -indeed a beautiful creation! I also see Bazian's radio used a custom wound ferrite rod RF transformer in place of the pulse tranformer and appears placed little bit away from the antenna rod 90 degree angled, also somewhat close to the loudspeaker. Wondering it did not create any instability there. But it seems like carefully oriented on an extended pcb -what it tells? A sweetspot? :D
When I was thinking little bit more on this, I think that the greatest challenge would be to maintain the quality of local oscillation and mixing while the transistor is attemting a high level of audio current flow. So instability issues will be anticipated.
Here's a question to all. How will be a normal superhet receiver respond to Scenario 1 and 2? What will be the audio output frequency detected out?
Mixer.png
Mixer.png (16.46 KiB) Viewed 2681 times
Bob Weaver
Posts: 2550
Joined: Sun Apr 08, 2007 8:02 am
Location: Saskatoon
Contact:
Re: One Transistor Loudspeaker Superhet!
Post by Bob Weaver » Tue Dec 05, 2017 12:54 am
transistor495 wrote: ↑
Mon Dec 04, 2017 4:16 pm
When I was thinking little bit more on this, I think that the greatest challenge would be to maintain the quality of local oscillation and mixing while the transistor is attemting a high level of audio current flow. So instability issues will be anticipated.
Yes, I think that is the primary issue. Reflexing low level signals can be done without too much grief, but as soon as the audio signal starts to cause large voltage swings, the transistor's operating characteristics are going to change, and everything will be modulated by the audio. So, very careful design will be required to minimize this characteristic shift.
As for the mixing components, I'm not sure why you would have modulation on the local oscillator signal, but this is what I get:
MixProducts01.png
MixProducts01.png (64.93 KiB) Viewed 2613 times
MixProducts02.png
MixProducts02.png (64.81 KiB) Viewed 2613 times
This assumes a balanced mixer. If it's unbalanced, then you'll also get all of the original components in there too.
Transmitting from my bio-containment unit, deep beneath the Earth's crust.
achu
Posts: 76
Joined: Sun Oct 01, 2017 9:35 am
Location: Trivandrum, india
Re: One Transistor Loudspeaker Superhet!
Post by achu » Tue Dec 05, 2017 9:28 am
transistor495 wrote: ↑
Sun Dec 03, 2017 4:16 pm
I would like to hear the experience of anybody built this already. There were no video footage of a working version. How good it work?
I've seen this circuit years ago.Really interesting though.
User avatar
transistor495
Posts: 808
Joined: Wed Jun 30, 2010 9:54 am
Location: Kochi, India
Re: One Transistor Loudspeaker Superhet!
Post by transistor495 » Tue Dec 05, 2017 4:04 pm
Bob Weaver wrote: ↑
Tue Dec 05, 2017 12:54 am
As for the mixing components, I'm not sure why you would have modulation on the local oscillator signal, but this is what I get:
This assumes a balanced mixer. If it's unbalanced, then you'll also get all of the original components in there too.
This is the very much information I was looking for. As you said, the question was asked because I would expect LO internal modulation which can affect the mixing products because of the audio voltage build-up -which obviously in case of a conventional superhet cannot happen, however I just wanted to see what will be the result products and audio output based on the above scenarios if we apply modulation to LO.
For scenario 1:
Detected audio will be 1KHz LO modulation[dominant] + disturbance and noises from RF modulation[1KHz]. Audio may appear slightly distorted.
For scenario 2:
Detected audio will be 3KHz LO modulation + 1KHz RF modulation[1KHz]. A mix of audio output.
So basically the 1 Transistor superhet will be likely addressing scneario 1 but the LO modulation component will be detected audio -which is actually same for received RF also. One can think about how this will be affecting the end results. My theory is that -in this superhet, this can appear favorable to the operation and end results. Opposite chances are also possible. The audio voltage swings[which has a potential to modulate LO internally] can be expected under 100mV -this is just a sensible guess.
My current observation is that a loudspeaker superhet is very much possible :) . However, although the original ingenious reflex regen and my converted superhet design will be a great starting point, anybody attempt to do this project may inititally find little bit of challenges to get the MPSA18 oscillating across BCB downconversion range relying on current biasing conditions. Altering the biasing may be changing the original performance of the 1 TR reflex/regen -which would be needed to sorted out. The point is that if MPSA18 can get oscillating reliably across the bandwidth without much tweaking the original biasing -the success will start to begin.. :D
achu wrote: ↑
Tue Dec 05, 2017 9:28 am
I've seen this circuit years ago.Really interesting though.
So why don't you give it a try then :), although not very much alike, a BC549 has good gain and somewhat similar characteristics to MPSA18. A custom wound ferrite core RF transformer can be tried although the winding pattern and exact core characteristics of 1:1 pulse X'former is not very known. These seems to be having a marginal performance gain in the circuit however.
