Author Topic: Dxing with Regenerative receiver  (Read 2294 times)

0 Members and 1 Guest are viewing this topic.

Offline Dom13c

  • Regular Contributor
  • *
  • Posts: 77
  • Country: 00
  • living in the stone age for lacking rf equipments
Dxing with Regenerative receiver
« on: December 04, 2018, 10:54:54 am »
Is it possible to do some dxing with a Regenerative receiver?As long as I know they are fairly sensitive to weak signals. I'm actually attracted by the simplicity in the design of them.So I would like to build one that covers the whole HF band to listen to distant HF signals with a tuned half wave end fed wire antenna. If anyone has first hand experience don't forget to share it.
 

Offline CJay

  • Super Contributor
  • ***
  • Posts: 3547
  • Country: gb
Re: Dxing with Regenerative receiver
« Reply #1 on: December 04, 2018, 01:10:37 pm »
It's possible but I'd suggest having a listen to Bill Meara's Soldersmoke blog and his thoughts on regens.

 
The following users thanked this post: Dom13c

Offline Dom13c

  • Regular Contributor
  • *
  • Posts: 77
  • Country: 00
  • living in the stone age for lacking rf equipments
Re: Dxing with Regenerative receiver
« Reply #2 on: December 04, 2018, 01:39:53 pm »
Could you plese tell me the podcast number. Having trouble to find the particular one
 

Offline CJay

  • Super Contributor
  • ***
  • Posts: 3547
  • Country: gb
Re: Dxing with Regenerative receiver
« Reply #3 on: December 05, 2018, 12:22:59 pm »
It's a recurring theme with them so there's no single podcast that outlines it all.

I have however just listened to 129, 130 and 131 and there were definitely comments in at least two of them about regens.

Boiling it down to the basic though, regens are fiddly, unstable and difficult to get right, it's the price you pay for the simplicity.

They are fun to play with but my experience has been 'frustrating' and it was easier to use slightly more complex schemes for greater ease of use and better results
 
The following users thanked this post: Dom13c

Offline bd139

  • Super Contributor
  • ***
  • Posts: 13810
  • Country: gb
Re: Dxing with Regenerative receiver
« Reply #4 on: December 05, 2018, 12:27:17 pm »
Indeed. Have to agree. I built the one described in experimental methods in RF design. It worked, eventually, but the experience was bad.

They had a place when it was an economical way to get on air but for interest’s sake now even the crappiest of simple direct conversion receivers (ne602+lm386) is a lot better.  With some sharp audio filtering, they even verge on reasonably good!
 
The following users thanked this post: Dom13c

Offline Dom13c

  • Regular Contributor
  • *
  • Posts: 77
  • Country: 00
  • living in the stone age for lacking rf equipments
Re: Dxing with Regenerative receiver
« Reply #5 on: December 08, 2018, 01:34:53 pm »
Had an out of topic question. Have any of y'all built a DDS VFO .Had a thought to build one with Si570,but can't source one fir cheap. Any idea about the exact price of them?
 

Offline retrolefty

  • Super Contributor
  • ***
  • Posts: 1619
  • Country: us
  • measurement changes behavior
Re: Dxing with Regenerative receiver
« Reply #6 on: December 08, 2018, 02:08:31 pm »
Is it possible to do some dxing with a Regenerative receiver?As long as I know they are fairly sensitive to weak signals. I'm actually attracted by the simplicity in the design of them.So I would like to build one that covers the whole HF band to listen to distant HF signals with a tuned half wave end fed wire antenna. If anyone has first hand experience don't forget to share it.

 HF propagation is such when the skip conditions are active then even a simple regen and low power transmitter can reach the world. The antenna is the real 'force multiplier' in such a station.

 
 
The following users thanked this post: Dom13c

Offline janoc

  • Super Contributor
  • ***
  • Posts: 3058
  • Country: fr
Re: Dxing with Regenerative receiver
« Reply #7 on: December 08, 2018, 03:26:46 pm »
Re regens - in addition to what has been said above, if you decide to play with one, do make you build one of the "decent" designs that have an amplifier stage in front of the regenerative one and good bandpass filters before connecting any sort of antenna to it. Otherwise you are going to produce a ton of nasty interference which is coincidentally why regen designs are banned/heavily restricted in many countries. Possible  problems with the authorities aside, causing QRM like this is just bad juju and it won't gain you any friends on the bands.


Had an out of topic question. Have any of y'all built a DDS VFO .Had a thought to build one with Si570,but can't source one fir cheap. Any idea about the exact price of them?

Sure, e.g. mine on my website (tooting my own horn):
https://janoc.rd-h.com/archives/502

But there are plenty of designs around the net, pick based on features you need and the way you are planning to operate it. E.g. my code doesn't have band switching but it allows continuous tuning in the entire range without twisting your hand off while operating the encoder - good for a receiver or generator. Other versions are more radio oriented, with explicit support for multiple bands and modes.

I wouldn't do it with Si570 unless you explicitly need the high frequencies it can reach - it has only a single output and is not particularly easy to program because it isn't really a DDS synthesizer but a PLL. AD9850 (DDS synthesis) modules can be had cheaply and have sine output, another popular choice for superhets is Si5351 (a PLL clock generator) which has 3 square wave outputs (depending on the version can have up to 8 ), so you can easily build both a VFO and a BFO with it. If you don't want to solder tiny parts, modules are also available (e.g. from Adafruit).

The Soldesmoke guys (Bill Meara & Pete Juliano N6QW) have a ton of resources on homebrewing this stuff - both in the podcast and in Pete's blog. He is a prolific radio builder:

http://n6qw.blogspot.com/ (usually work in progress)
http://www.n6qw.com/ (once done he moves it here)

Bill's SolderSmoke podcast site:
http://www.soldersmoke.com/
« Last Edit: December 08, 2018, 03:29:46 pm by janoc »
 
The following users thanked this post: Dom13c

Offline bd139

  • Super Contributor
  • ***
  • Posts: 13810
  • Country: gb
Re: Dxing with Regenerative receiver
« Reply #8 on: December 08, 2018, 03:48:29 pm »
Pete Juliano needs to slow down a bit  :-DD

Had an out of topic question. Have any of y'all built a DDS VFO .Had a thought to build one with Si570,but can't source one fir cheap. Any idea about the exact price of them?

Another vote for 5351a’s. Look for Si5351a clock generator boards on aliexpress. You can get three outputs for around $4. I’m using one of these as an LO. They work very nicely with balanced diode ring mixers. Plug it into a cheap Arduino clone and LCD and load up some off the shelf firmware and you’re done.
 
The following users thanked this post: Dom13c

Offline JohnPen

  • Regular Contributor
  • *
  • Posts: 228
  • Country: gb
Re: Dxing with Regenerative receiver
« Reply #9 on: December 08, 2018, 06:25:56 pm »
QRP Lab kits produce a VFO kit using the Si5351a if you don't want to design your own.   Their experiments show that the chip can reach ~290 Mhz a lot higher than the chip's specification would suggest.  Their particular kit can also produce 2 clock outputs in quadrature for it's frequency range providing you accept a lower frequency limit of ~3.5 Mhz which is useful for Tayloe type detectors.
 

Offline bd139

  • Super Contributor
  • ***
  • Posts: 13810
  • Country: gb
Re: Dxing with Regenerative receiver
« Reply #10 on: December 08, 2018, 06:38:37 pm »
I wouldn't buy the QRP Labs one. The voltage regulator implementation on it is faulty. They used an LM317L with a poorly chosen resistive divider meaning the Si5351A runs slightly higher than 3.3V plus on top of that they didn't consider the minimum dropout voltage of the LM317L for the load line which means some of them actually don't work or are unreliable under load. Hans is way too good at cutting corners. The regulator for the Si5351A on the QCX for example is two diodes in series from the 5V line.  Even Heathkit wouldn't dare pull shit like that.

Chinese no brand one is much better. And cheaper.
 

Offline JohnPen

  • Regular Contributor
  • *
  • Posts: 228
  • Country: gb
Re: Dxing with Regenerative receiver
« Reply #11 on: December 08, 2018, 06:40:49 pm »
A vintage thought.  My father built a 3 valve Short Wave receiver back in the 1930s.  It used an RF amplifier,  with the new screen grid valve :),  followed by a regenerative detector and an AF stage.  The tuning and regenerative detector both used slow motion drives for ease of tuning and regeneration gain.  It was a very sensitive receiver and certainly received < 20 watt signals from South America and Australia very clearly.  This was in UK and Solar cycle permitting.  As mentioned your popularity will be poor with locals if you do not have an RF stage in front of the detector. :(
 

Offline vk3yedotcom

  • Frequent Contributor
  • **
  • Posts: 539
  • Country: au
    • vk3ye dot com (radio articles and projects)
Re: Dxing with Regenerative receiver
« Reply #12 on: December 09, 2018, 02:36:37 am »
Sure!  Here's a demonstration.



SSB becomes hard at higher frequencies due to poor stability.  However you can overcome that either (i) by building your receiver like a battleship or (ii) using the receiver with the regen set just before oscillation and a stable external VFO (eg DDS) acting as a BFO. That pretty much eliminates frequency drift on a regen unless it's really bad.
If you're into amateur radio you might enjoy my books. Choice of 7. Electronic or paperback. Details here: http://home.alphalink.com.au/~parkerp/vk3yebooks.htm
 
The following users thanked this post: Dom13c

Offline kg4arn

  • Supporter
  • ****
  • Posts: 268
  • Country: us
Re: Dxing with Regenerative receiver
« Reply #13 on: December 09, 2018, 09:56:58 am »
Regarding the Si5351I am currently building a 40 meter single conversion superhet using the Si5351 for the VFO and the BFO.The IF frequency is 9MHz.  The VFO frequency is 16MHz. I am using the Adafruit Si5351 breakout board.

My experience so far: 
-It is very stable with very little drift.  It drifts a Hz or 2 over hours. 
-I detected faint spurious signals in the receiver pass band.  (A lot of them.) These birdies were definitely coming from the Si5351. The spurs were on the order of -70dBc (-60dBm), or less, as measured on an Agilent SA. I built a 16MHz band pass filter and this eliminated the spurs, birdies, and to my surprise the receiver baseline noise dropped noticeably. 
 The measured MDS improved from -129 to -131dBm, so the Si5351a must have also been contributing noise in the 7MHz range as well. 
I don't know if selecting a different IF frequency might improve the situation.As it stands now the Si5351 makes an excellent VFO but needing a BPF to remove broadband noise and spurs in the pass band which are then injected into the front end mixer.  The birdies were not tremendously loud but not acceptable for a CW op like me.

 

Offline JohnPen

  • Regular Contributor
  • *
  • Posts: 228
  • Country: gb
Re: Dxing with Regenerative receiver
« Reply #14 on: December 09, 2018, 09:59:01 am »
Another slightly more complicated arrangement that some tried in the past was a superhet front end with a regenerative detector operating at the IF frequency.   In my early days  I came across a valve based one that worked very well.  One thing I never understood properly was why this particular receiver suddenly became extremely sensitive when the  regeneration was wound far past the normal oscillation point.  Under this condition the valve lit up with a blue glow rather like it was gas filled.   It was not a voltage discharge and seemed quite stable and worked very reliably in that state.   At a guess I suspected it might have turned into a superregen detector as it had that sort of noise output.  In those days my test equipment consisted of a simple analogue meter and a'mains eliminator box'.  Also used ex WD 120 volt dc batteries when I could get them. :)
 

Offline bd139

  • Super Contributor
  • ***
  • Posts: 13810
  • Country: gb
Re: Dxing with Regenerative receiver
« Reply #15 on: December 09, 2018, 10:06:26 am »
Regarding the Si5351I am currently building a 40 meter single conversion superhet using the Si5351 for the VFO and the BFO.The IF frequency is 9MHz.  The VFO frequency is 16MHz. I am using the Adafruit Si5351 breakout board.

My experience so far: 
-It is very stable with very little drift.  It drifts a Hz or 2 over hours. 
-I detected faint spurious signals in the receiver pass band.  (A lot of them.) These birdies were definitely coming from the Si5351. The spurs were on the order of -70dBc (-60dBm), or less, as measured on an Agilent SA. I built a 16MHz band pass filter and this eliminated the spurs, birdies, and to my surprise the receiver baseline noise dropped noticeably. 
 The measured MDS improved from -129 to -131dBm, so the Si5351a must have also been contributing noise in the 7MHz range as well. 
I don't know if selecting a different IF frequency might improve the situation.As it stands now the Si5351 makes an excellent VFO but needing a BPF to remove broadband noise and spurs in the pass band which are then injected into the front end mixer.  The birdies were not tremendously loud but not acceptable for a CW op like me.

What IF? Is your mixer terminated properly?

I am using a 4.9152MHz IF and all is quiet. I did have problems with the front end mixer. I had to terminate it properly with a diplexer to get some mixing products to go away. Didn’t need an LPF on the output of the clock gen for that.
 

Offline kg4arn

  • Supporter
  • ****
  • Posts: 268
  • Country: us
Re: Dxing with Regenerative receiver
« Reply #16 on: December 09, 2018, 10:43:17 am »

What IF? Is your mixer terminated properly?

I am using a 4.9152MHz IF and all is quiet. I did have problems with the front end mixer. I had to terminate it properly with a diplexer to get some mixing products to go away. Didn’t need an LPF on the output of the clock gen for that.
 

Yes the mixer was properly terminated with a 6 dB resistive pad on the VFO port.  I am using a TUF-3 from Minicircuits.The IF port was terminated with a high intercept, broadband, feedback amplifier. 

The spurs are definitely on the Si5351 output (before they hit the mixer) as they are seen on the spectrum analyzer with the Si5351 terminated into 50 ohms.As you change the frequency output on the Si5351, the spurs move in frequency and amplitude as expected.  I didn't save a screen shot but If I have some time I will get one.

The birdies occur when the frequency output and the spurs just happen to be be precisely 9MHz apart, which is my IF frequency.  Lucky me.
 
I did some work terminating with higher impedance but it made no noticeable difference.I also did not see much difference with drive levels selected for the Si5351 output (I tested levels 1 1nd 3) 

I could not find any info on the Si5351 spurious output levels.I found a response from Silicon Labs to a question from a customer, on their website somewhere, that Silicon Labs would test the chip for spurious outputs for you if you specified your frequency scheme.  https://www.silabs.com/community/timing/forum.topic.html/si5351_spurious-I176


I am not criticizing the chip or condemning its use.  Just something to watch out for, as the chip might not be as "plug and play" as one might expect.I had ordered a VFO from QRP labs (I saw the criticism above) to see if it had the same issues.  It hasn't arrived yet.  I am thinking that programming of the chip might be different but I expect to see the same spurious responses.  We'll see.

 


 

Offline bd139

  • Super Contributor
  • ***
  • Posts: 13810
  • Country: gb
Re: Dxing with Regenerative receiver
« Reply #17 on: December 09, 2018, 10:49:47 am »
Interesting. Thanks for the write up. Drive level i optimised as low as possible so it didn’t burn up power. Also noted it made no difference.

I’m not sure SiLabs care about spurious output here as it’s a clock generator really.

If you receive the QRP labs kit you can improve it by substituting the LM317L for an MCP1700 LDO. Mine worked properly then.

I’m actually building analogue VFOs now as they are more interesting if I’m honest. I can get down to 20Hz/hour drift which I’m happy about  8)
 

Offline JohnPen

  • Regular Contributor
  • *
  • Posts: 228
  • Country: gb
Re: Dxing with Regenerative receiver
« Reply #18 on: December 09, 2018, 11:14:07 am »
With regard to the QRP Labs VFO using the Si5351a I did a quick check on my one.   I found that supplying 5.0v to the kit resulted in 3.3v to the Si5351a.  However increasing the 5v supply to 5.35v resulted in the 3.3v increasing to to 3.52v.  There was therefore a limited voltage regulation effect in that the 3.3v rail only increased by ~60% of the increase of the 5v rail. Adding an extra 10mA load onto the 3.3v rail however did not alter the LM317's output voltage.  Sadly when buying a kit one does not always check such things 'Oh its just a straight forward regulator'.  As BD139 stated this is not a stable regulator set up.  The LM317 spec says it needs at least a 3v drop to regulate properly.  That said with a stable regulated 5v supply on the complete VFO it is sort of stable in that variations on the current draw by the installed Si3351a shouldn't have an effect.  I suppose it is better than a resistor but with this arrangement one could accidentally exceed the voltage rating of the Si5351a when using it as a freestanding module not a good idea. :(
 

Offline bd139

  • Super Contributor
  • ***
  • Posts: 13810
  • Country: gb
Re: Dxing with Regenerative receiver
« Reply #19 on: December 09, 2018, 11:25:54 am »
That's exactly what I observed with mine. Shows how much narrow testing was done.

I was using mine in a sweeperino arrangement like this as a test case:



The Si5351 was being powered directly from the 5V 7805 here and it was locking on certain frequencies and then just dying. MCP1700 and all is good.

I don't use the sweeperino though now. I rebuilt the AD8307 board as a standalone proper one and use that, a scope and an HP AWG. That has a lot better performance! AVR ADC needs more bits for a start.

I bought a couple of these to use going forward - very nicely made: https://www.aliexpress.com/item/DC-3V-5V-CJMCU-5351-Si5351A-Si5351-I2C-Clock-Generator-Breakout-Board-Module-Signal-Generator-Clock/32881587285.html

Has proper regulator, same level shifting circuit and less crap board design. You can slide a TCXO on there as well if you want.
« Last Edit: December 09, 2018, 11:31:32 am by bd139 »
 

Offline kg4arn

  • Supporter
  • ****
  • Posts: 268
  • Country: us
Re: Dxing with Regenerative receiver
« Reply #20 on: December 09, 2018, 12:33:43 pm »
Here are some screen shots of the Si5351 spursI tested 2 units on the Adafruit breakout boards, one unit had significantly more spurs, I am not sure why.  but both units showed similar spurs in the 7MHz region.

CLK2 output is at 16MHz
CLK0 output is at 9MHz as it is used for the BFO
CLK0 output is terminated with a 51 ohm resistor soldered right on the breakout board.

There must be some interaction within the Si5351 or on the breakout board or both, since the spurious responses are reduced when CLK0 is disabled.

 

Offline bd139

  • Super Contributor
  • ***
  • Posts: 13810
  • Country: gb
Re: Dxing with Regenerative receiver
« Reply #21 on: December 09, 2018, 12:52:13 pm »
I reckon some of that is the absolutely terrible board layout. Lots of unintended signal coupling. Spot the turds:



1. Left block - those three vias go down to the board edge pins. They have to break the opposing ground plane to do that and are probably routed next to each other.
2. Right block - CLK2 output runs parallel to the DC input line.
3. Not much via stitching going on.

 :palm:

Adafruit: ship version 1 (this includes their buggy shit software too)

This is version 3 of something I'm working on. It took me 3 board revisions to get this clean:

« Last Edit: December 09, 2018, 12:54:42 pm by bd139 »
 
The following users thanked this post: cdev

Offline kg4arn

  • Supporter
  • ****
  • Posts: 268
  • Country: us
Re: Dxing with Regenerative receiver
« Reply #22 on: December 09, 2018, 02:00:46 pm »
@bd139Thanks for pointing that out on the board.
I am using the ke7er software to control the 5351.
Like you said above, it's a clock chip.  Not intended for this use.
In true HAM tradition, the chip was hijacked for VFO purposes.

I'd be interested in a, "spurious lite" PLL or synthesizer design that could be used for a multiband receiver.
I am just not keen enough to do something like that myself.

btw 20 Hz drift in an hour for an analog VFO is excellent, IMO, and certainly more than adequate for CW or sideband communication. 

 

Offline Dom13c

  • Regular Contributor
  • *
  • Posts: 77
  • Country: 00
  • living in the stone age for lacking rf equipments
Re: Dxing with Regenerative receiver
« Reply #23 on: December 09, 2018, 02:24:55 pm »
Sure!  Here's a demonstration.



SSB becomes hard at higher frequencies due to poor stability.  However you can overcome that either (i) by building your receiver like a battleship or (ii) using the receiver with the regen set just before oscillation and a stable external VFO (eg DDS) acting as a BFO. That pretty much eliminates frequency drift on a regen unless it's really bad.
Co incidentally though, after watchin this very video, I got the plan for making a Regen receiver  :-DD. Love your qrp projects vk3ye, keep designing them. :-+
 

Offline bd139

  • Super Contributor
  • ***
  • Posts: 13810
  • Country: gb
Re: Dxing with Regenerative receiver
« Reply #24 on: December 09, 2018, 02:56:57 pm »
I moved the sweeperino code over to si5351a from si570. Was basically cut and pasting stuff. Took about 30 minutes. I used etherkit si5351 library.

I have ported that to plain old AVR libc compatible C because I was getting pissed off with the crappy Arduino environment. Got into an interrupt handling nightmare trying to keep power usage down.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf