 |
TheRadioBoard
Forum for the homemade radio builder.
Newbie posts are most welcome here!
|
| View previous topic :: View next topic |
| Author |
Message |
qrp-gaijin
Joined: 28 Feb 2010
Posts: 68
|
 Posted: Sun Jul 25, 2010 2:13 am Post subject: Single signal regenerative receiver? |
 |
|
Hi again people. In my quest to understand and get the best performance from the regenerative receiver design, I was wondering: is it possible to get single-signal performance from a regen? I understand that DC receivers can achieve this (though not as well as a superhet can); is it theoretically possible with a regen also, or is it impossible? I didn't find any indications that this was possible (or impossible) in my searching on the web, which leads me to believe it's not possible; I hope you folks can prove me wrong!  |
|
| Back to top |
|
 |
exray
Joined: 30 Apr 2007
Posts: 1804
Location: Vieques, PR
|
| Posted: Sun Jul 25, 2010 5:46 am Post subject: |
|
|
Theoretically, regens have almost 'infinite' gain and selectivity thanks to the process of regeneration. They do very well for CW signals. If you need better then an audio peaking filter would be the most logical choice. This is done on many of the DC receivers as well.
AM is a different story because the regen isn't tuned quite into full regeneration when receiving AM. In this mode selectivity is going to be dependent on the quality of the tuned circuit. In my playing around with coils of varying Q on shortwave I've never been able to achieve any substantial improvement.
That said, in casual tuning of the SW-BC bands I rarely encounter bleedover from adjacent stations unless they are terribly strong.
-Bill |
|
| Back to top |
|
|
KR1S
Joined: 06 Jun 2009
Posts: 750
Location: South Florida
|
| Posted: Sun Jul 25, 2010 6:40 am Post subject: Re: Single signal regenerative receiver? |
|
|
| qrp-gaijin wrote: | | is it possible to get single-signal performance from a regen? |
Sort of. Single-signal reception refers to the receiver's ability to filter one sideband while passing the other. It's mostly applicable to Morse Code reception. On most simple receivers, you hear the same frequency of beat note at equal distances from the carrier frequency.
The methods of detecting CW signals with a beat-frequency oscillator are like those used to generate single-sideband suppressed carrier AM (aka SSB). If the carrier frequency and beat oscillator are at 455 kHz, one-kHz beat notes occur at 454 and 456 kHz. A filter with a center frequency slightly above or below 455 kHz, used with a 455-kHz BFO, will not pass one sideband of a signal at 455 kHz. This is the filter method.
Some elaborate direct-conversion receivers use the phasing method, where the signal is split into two, 90 electrical degrees out of phase. The phasing method was popular for transmitters in the early days of amateur SSB, because it did away with hard-to-build or expensive crystal filters. Amateur radio handbooks from the 1950s-1960s explain both.
So we have two methods: Filter and Phasing. Any DSP implementation uses one of these methods, but in the digital domain. Of the two methods, the one suited for regenerative receivers is "Filter."
After WW2 the market was flooded with lousy receivers, most of which had an IF at or near 455 kHz. Users found that making one IF stage regenerative greatly improved selectivity. Or you could tack on a parasitic regenerative stage, called a Q-multiplier. Assuming a fixed signal frequency (the IF), you could peak or notch, above or below the IF, and achieve reasonable single-signal reception. But the Q-multiplier usually didn't oscillate; that was still left to the BFO.
These were superhets, and the regenerative part was done at VLF. The regen stage only had to tune a few kHz. For the same Q, the bandwidth is less as you go lower in frequency, as it's a percentage of the frequency. You could do this at the signal frequency, with a non-oscillating detector and a beat oscillator.
The subject of separate beat oscillators came up here recently. In a classic regen article 1933, W3LW suggested them, as they allowed a CW receiver to get the most advantage from a regenerative stage. I believe exray mentioned trying it, with mixed results. I've tried it too, and maybe got better results, but it's a lot more work, and the secondary oscillator has to be pretty stable. I used a digital synthesizer! 
The article was "Regenerative Detectors, What We Get From Them -- How To Get More," in QST for February 1933. As I said, you might get single-signal reception by working at the operating frequency, but a superhet with regenerative, low-frequency-IF is a better way.
73,
_________________
Jim, KR1S
Just build it!
http://kr1s.kearman.com/
http://qrp.kearman.com/ |
|
| Back to top |
|
|
qrp-gaijin
Joined: 28 Feb 2010
Posts: 68
|
| Posted: Mon Jul 26, 2010 2:07 am Post subject: Re: Single signal regenerative receiver? |
|
|
Thanks for the detailed explanations. I'd like to pursue this idea further. Let's suppose I wanted to apply the filter idea to a regen, working at the signal frequency (and not at a superhet's IF). I know this isn't the ideal approach, but I'm not yet prepared to build a superhet (I don't yet even know what I don't know about superhets). Then...
| KR1S wrote: | Or you could tack on a parasitic regenerative stage, called a Q-multiplier.... You could do this at the signal frequency, with a non-oscillating detector and a beat oscillator.
|
If I'm understanding the idea correctly, it is to use the non-oscillating tuned regenerative stage as a filter at the signal frequency, adjusted such that the undesired sideband is attenuated while the desired sideband is peaked (or at least, attenuated less). The non-oscillating detector allows adjustment of the regenerative filter response (sharpness and center frequency) and simultaneously detects AM, while the separate BFO provides the necessary carrier for CW/SSB reception and allows setting the beat tone independently from the regenerative filter setting (which presumably must be tuned somewhat far off the signal frequency to allow the unwanted sideband to be attenuated sufficiently). Did I get the basic idea right?
How would such a configuration be tuned in practice? I imagine something like: first tuning the desired signal with regen oscillating and no BFO, then switching on (or in) the BFO and zero-beating the BFO with the desired signal, bringing the regen out of oscillation (which might change the regen frequency from the zero-beated BFO frequency, but let's ignore that problem), tweaking the BFO away from zero-beat for the desired beat note, and finally tweaking the regen's center frequency to attenuate the unwanted sideband. Something like that, or is there an easier way?
How about an alternate configuration: two regenerative stages? The first non-oscillating regenerative stage filters the RF to pass only one sideband; a second, separately-tuned, oscillating regenerative detector stage then operates as normal on the filtered RF from the first stage. Would this work? I imagine the first stage would need to amplify/filter but not detect the RF, allowing the second stage to do the detection.
I'm not sure if it's the same thing or not, but the "Q-dyne" seems to be based on this idea and is described as being similar to so-called "double-regeneratives" of the 1930's: http://www.qsl.net/wd4nka/TEXTS/qdynef.htm
Thanks again for the informative post. I am quite attracted by the ease of building wide-frequency-range "all-band" regenerative receivers, but the single-signal issue was still bugging me. Nice to see that there are some tricks available to get almost single-signal out of the regen without going all the way to a superhet design. |
|
| Back to top |
|
|
Rune
Joined: 15 Jul 2007
Posts: 510
Location: Akershus, Norway
|
| Posted: Mon Jul 26, 2010 3:35 am Post subject: Re: Single signal regenerative receiver? |
|
|
| qrp-gaijin wrote: | | Hi again people. In my quest to understand and get the best performance from the regenerative receiver design, I was wondering: is it possible to get single-signal performance from a regen? |
I have had dreams about a ''Variable Q, Wide Frequency Range, Tunable Band-pass Crystal Filter'' for a regen. Possibly placed in series with the tuning tank and the amplifying device, and ganged to the tuning capacitor. Of course it had to be non-digital too. Who would need ''pesky'' superhets anymore?
Well, I gladly admit that I'm too dumb to even approach the idea of the design of such a filter..
Interesting posts by both exray and KR1S.. |
|
| Back to top |
|
|
KR1S
Joined: 06 Jun 2009
Posts: 750
Location: South Florida
|
| Posted: Mon Jul 26, 2010 3:49 am Post subject: Re: Single signal regenerative receiver? |
|
|
| qrp-gaijin wrote: | | If I'm understanding the idea correctly, it is to use the non-oscillating tuned regenerative stage as a filter at the signal frequency, adjusted such that the undesired sideband is attenuated while the desired sideband is peaked (or at least, attenuated less). The non-oscillating detector allows adjustment of the regenerative filter response (sharpness and center frequency) and simultaneously detects AM, while the separate BFO provides the necessary carrier for CW/SSB reception and allows setting the beat tone independently from the regenerative filter setting (which presumably must be tuned somewhat far off the signal frequency to allow the unwanted sideband to be attenuated sufficiently). Did I get the basic idea right? |
That's correct.
| qrp-gaijin wrote: | | How would such a configuration be tuned in practice? |
You'd like a detector that won't shift frequency when going from oscillating to non-oscillating, of course. You make the detector oscillate, then back off and swoop in with the heterodyne oscillator. Takes some practice, and some time.
| qrp-gaijin wrote: | | How about an alternate configuration: two regenerative stages? The first non-oscillating regenerative stage filters the RF to pass only one sideband; a second, separately-tuned, oscillating regenerative detector stage then operates as normal on the filtered RF from the first stage. Would this work? I imagine the first stage would need to amplify/filter but not detect the RF, allowing the second stage to do the detection. |
There was a QEX article about a multi-stage regen, that was reprinted in ARRL's most-recent edition of "QRP Power." It reminded me of the old "TRF" receivers, which used several tuned (but not regenerative) rf amplifier stages ahead of the diode detector: Many knobs to turn! And Regeneration controls for each stage. The author was into SWBC, so he didn't run the stages in oscillation, though they would. You can't easily tune around the band with a receiver like that. Too much like work for me!
Frankly, this seemed like overkill on a global scale. It would be far, far easier to build a superhet with a regenerative IF, where you could obtain the desired selectivity in one stage. One way to think about circuit Q is the relationship between operating frequency and bandwidth. It's easier to get narrow bandwidths at lower frequencies. My MW AMBC regen will clip sidebands from broadcast stations, but it wouldn't manage single-signal CW reception.
| qrp-gaijin wrote: | | I'm not sure if it's the same thing or not, but the "Q-dyne" seems to be based on this idea and is described as being similar to so-called "double-regeneratives" of the 1930's: http://www.qsl.net/wd4nka/TEXTS/qdynef.htm |
I think Gary founded the "regenrx" Yahoo group. What you have there is a mixer stage with feedback, so you get two-for-one.
I believe Gary also coined the term "regenerodyne." (If I'm wrong, I apologize to the real inventor!) This is one way to get what you're after. You build one, tunable receiver, say 2-3 MHz, and optimize it.
If you can hunt down some 455-kHz IF transformers, you can try building a fixed-frequency regenerative IF stage with separate BFO and, ideally, a product detector. Once you have that section perfected, sticking tunable converters ahead of it is simple. You don't have to use 455-kHz, but you can get awe-inspiring selectivity down there. The problem with going lower is making inductors of reasonable physical size. I've never found a ferrite or iron-powder core material that worked well for tuned circuits down there. You can manage it down to ~100 kHz with -61 ferrites, but the coils are fairly large.
Superhets were less popular in the 1930s because they cost more to build, and there were technical challenges beyond many builders' abilities. Not so today, when parts are so cheap (assuming you're using solid-state devices).
73,
_________________
Jim, KR1S
Just build it!
http://kr1s.kearman.com/
http://qrp.kearman.com/ |
|
| Back to top |
|
|
Ham-er
Joined: 01 Dec 2007
Posts: 684
|
| Posted: Mon Jul 26, 2010 6:22 pm Post subject: superhet yes |
|
|
YEP
Everything jim said.
Build a superhet. Now wait a minute. A superhet does not Have to have 2 IF amps and 3 IF CANS. It does not have to be an elaborate Multi-stage(as in 5 or 6 active stages) receiver.
It does not in fact absolutely have to have an IF amplifier at all.
One can get most of the advantages of a superhet without building one as complicated as say an AA5.
A regenerodyne is just one example. Suppose instead we actually use a crystal filter in the IF. Then we could have a "Q" of 10000 or more.
Wouldnt we have a good bit of Insertion loss with no IF amps this way?
Yes but you could still get "single-signal" reception.
Still it is not all that hard or expensive to build a REAL superhet, Especially if one limits oneself to only 1 or 2 Bands, and say one MODE.
(80 and 40 meter) and say only CW.
73
kb0lxy - Eric |
|
| Back to top |
|
|
qrp-gaijin
Joined: 28 Feb 2010
Posts: 68
|
| Posted: Sat Jul 31, 2010 9:41 pm Post subject: Re: Single signal regenerative receiver? |
|
|
| KR1S wrote: |
The subject of separate beat oscillators came up here recently. |
Do you happen to have a link handy or could you suggest the keywords to search for? I couldn't seem to find the thread on this board.
I did find a few other articles on other sites which reveal an apparently complex issue. Gary WD4NKA says the regen will lock to the BFO frequency over about a 50 KHz range (ref. 1), which apparently improves stability of the detector and also raises interesting possibilities like a crystal oscillator used both to "lock" the receiver's detector frequency and to drive a transmitter's PA - a regenerative transceiver.
On the other hand John KB7NRN says running the BFO at half-frequency prevents the regen detector from locking to the BFO frequency, and that the BFO, as compared to running the detector in oscillation, seems to improve sensitivity and stability (Ref. 2).
This seems like an interesting variant on the regen.
Ref. 1 http://groups.yahoo.com/group/regenrx/message/7349
Ref. 2 http://www.antiqueradios.com/forums/viewtopic.php?t=68005 |
|
| Back to top |
|
|
qrp-gaijin
Joined: 28 Feb 2010
Posts: 68
|
| Posted: Sun Aug 01, 2010 4:15 am Post subject: Super-regen + BFO? |
|
|
| While we're on the topic of separate BFO's, how about a HF super-regenerative receiver (at e.g. 7 MHz) plus a separate 7 MHz BFO, for CW reception? The fantastic sensitivity of a super-regen plus a BFO seems, on the surface, to be an interesting combination. The selectivity would suffer terribly, but for CW, it seems like a narrow audio filter could take care of attenuating undesired CW signals. Would this work? |
|
| Back to top |
|
|
KR1S
Joined: 06 Jun 2009
Posts: 750
Location: South Florida
|
| Posted: Sun Aug 01, 2010 5:59 am Post subject: Re: Super-regen + BFO? |
|
|
| qrp-gaijin wrote: | | While we're on the topic of separate BFO's, how about a HF super-regenerative receiver (at e.g. 7 MHz) plus a separate 7 MHz BFO, for CW reception? |
Superregens are noisy!  I wouldn't bother with this approach.
I don't know how to search for the thread about the separate oscillator; the way threads go here, it could be anywhere. Not much was said; Bill, 'exray,' mentioned trying one. He's reading this thread, so maybe he'll recap his experiences.
In your previous post you touched on a few alternatives, such as a beat oscillator running at f/2. I've never tried that. My suggestion is to build a stable regen detector, and then try different methods of beating against the signal. Using the BFO to improve stability of an unstable detector is probably not a good approach. It's a Band-Aid solution to a poor detector. The benefit of using a BFO should be solely to allow the detector to operate just below oscillation, for best selectivity and sensitivity. I doubt you'll ever achieve single-signal selectivity at 7 MHz, though, and any improvement won't help much when 40 M is crowded. And there's the (to me) annoying need to adjust the BFO every time you change frequency.
73,
_________________
Jim, KR1S
Just build it!
http://kr1s.kearman.com/
http://qrp.kearman.com/ |
|
| Back to top |
|
|
exray
Joined: 30 Apr 2007
Posts: 1804
Location: Vieques, PR
|
| Posted: Sun Aug 01, 2010 7:54 am Post subject: |
|
|
It was the George Grammer superhet radio that I built that used a BFO.
http://www.sparkbench.com/homebrew/grammar/grammar.html
http://www.io.com/~nielw/3tube_xtal/3tube_xtal.htm
While a BFO seems superfluous its another simple tool in the box. With regeneration alone you're stuck with possibly tuning into the QRM to get your desired beat note. With the BFO you can tune away from the QRM and use it to establish your preferred tone.
Note that the B+ is switched to the BFO. You just turn it on when you want to use it. Its not running full time and doesn't get in the way of normal regen operation. |
|
| Back to top |
|
|
qrp-gaijin
Joined: 28 Feb 2010
Posts: 68
|
| Posted: Mon Aug 02, 2010 2:33 am Post subject: |
|
|
I tried an experiment using a non-oscillating Armstrong regen tuned to the CW portion of 40m. I held my GDO nearby (no physical connection) and tuned it to 7 MHz until beat tones were heard from the CW signals.
The unexpected thing was then that tuning the non-oscillating regen's main tuning capacitor changed the pitch of the beat note slightly. I don't see how this can happen because the beat note is determined solely by the frequency of the incoming signal and the BFO frequency, neither of which is changing when I tune only the regen's main tuning capacitor in non-oscillating mode.
In practice, when using a BFO and a non-oscillating regen, does the regen's tuning affect the beat note, or should the beat node be rock-steady when adjusting the main tuning?
P.S. As for single-single reception, I'm trying to work on a simple superhet design. Something like a 40m bandpass filter fed into a 2N3904 mixer, mixing with a variable Colpitts LC LO, followed by a single regenerative VXO IF amp/filter stage, again mixed with a 2N3904 mixer together with a VXO BFO slightly offset from the IF. I plan to post some schematics as soon as I figure out how the stages are supposed to be coupled together... |
|
| Back to top |
|
|
KR1S
Joined: 06 Jun 2009
Posts: 750
Location: South Florida
|
| Posted: Mon Aug 02, 2010 7:33 am Post subject: |
|
|
| qrp-gaijin wrote: | I tried an experiment using a non-oscillating Armstrong regen tuned to the CW portion of 40m. I held my GDO nearby (no physical connection) and tuned it to 7 MHz until beat tones were heard from the CW signals.
The unexpected thing was then that tuning the non-oscillating regen's main tuning capacitor changed the pitch of the beat note slightly. I don't see how this can happen because the beat note is determined solely by the frequency of the incoming signal and the BFO frequency, neither of which is changing when I tune only the regen's main tuning capacitor in non-oscillating mode. |
So you think! A GDO will pull like crazy when coupled into a resonant circuit, even from a considerable distance. It's living, breathing proof of why untuned rf amplifiers ahead of regen detectors are a Good Thing.
73,
_________________
Jim, KR1S
Just build it!
http://kr1s.kearman.com/
http://qrp.kearman.com/ |
|
| Back to top |
|
|
Ham-er
Joined: 01 Dec 2007
Posts: 684
|
| Posted: Mon Aug 02, 2010 2:23 pm Post subject: Re: Super-regen + BFO? |
|
|
| qrp-gaijin wrote: | | While we're on the topic of separate BFO's, how about a HF super-regenerative receiver (at e.g. 7 MHz) plus a separate 7 MHz BFO, for CW reception? |
Super-regens are not used at HF.
The reason is that the Quench frequency
would be in the AUDIO range.
73
kb0lxy |
|
| Back to top |
|
|
Georg
Joined: 13 Oct 2007
Posts: 161
Location: Germany
|
| Posted: Mon Aug 02, 2010 11:57 pm Post subject: |
|
|
| Quote: | | how about a HF super-regenerative receiver (at e.g. 7 MHz) |
At 7 MHz a Superregenerative receiver will work.
I`d try with a quench frequency at about 20 to 30 kHz.
There are problems at BCB frequencies, but not 7 MHz.
One could even work with a quench frequency around
10 kHz, which makes two samples per cycle for an AF
at 5 kHz. You need to filter the 10 kHz then, of course.
Georg |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
