Dave Richards AA7EE

October 15, 2016

Comparing the Weak Signal Performance of a WBR Regen with a K2

The WBR seems to get a bit of a bad rap with some people for it’s sensitivity. A comment on the last post from a reader called Simon, reminded me that some WBR builders have experienced poor sensitivity. Based on my experience, this design does seem to be fairly deaf on AM, but the sensitivity on SSB/CW is fine. I think there are two reasons why some builders experience low sensitivity –

1) They follow the schematic from the original QST article, and do not include an audio pre-amp immediately before the LM386. In this case, the receiver is not necessarily insensitive – it’s just that the low audio is limiting what you can hear.

2) The value of Z1, the inductance between the coil tap and ground, is not high enough. In the original WBR design by N1BYT, this inductance was a 1-inch length of #20 solid copper wire. I followed this direction with my first WBR (for 40M) and it worked well. The WBR was tackled as a group build in the QRP-tech Yahoo Group, as I have mentioned in this blog before, and some builders experienced low sensitivity. The fix was to replace the 1-inch length of wire with an inductor wound on a toroid. Builders in the group found the optimum value of inductance to be somewhere between about 0.2uH and 1uH. I went lower with my 30M WBR, and found that a value of 0.03uH  (3 turns on a T37-6) worked well.

Of the above 2 reasons, my suspicion is that 1) is the main one for most builders.

We regen fans do get a bit braggy about the performance of our sets. I could never make the claim that my regens perform as well as a superhet, for several reasons. Obviously, the strong signal handling of regens is pretty poor, and the bandwidth is wide. When a regen is adjusted close to the point of oscillation, the nose of the response curve becomes quite narrow, but the skirts are still broad. Also, it’s a small difference, but the fact that a regen listening to SSB or CW hears on both sides of the oscillator, as opposed to a superhet, which only hears on one side of the LO, gives the regen an immediate 3dB disadvantage. Basically, for a given signal, a regen is listening to twice as much bandwidth as it needs to (a doubling of power is an increase of 3dB). It’s not a big difference, but it is there.

Having said all that, I am constantly surprised by how much my regens do hear. I remember one evening, a few years ago, when the Russian K beacon was coming through very, very weakly on 7039.3KHz on my K2. I was amazed to discover that I could also hear it on my WBR. Admittedly, I had to strain to copy it on the WBR, and the fact that it was sending the same letter over and over again – and I knew in advance which letter it was, all helped. However, the fact that it was marginal copy on the K2, combined with the fact that I could copy it at all on my WBR (albeit even more marginally) was an eye-opener.

With all that in mind, here’s a 3 minute video of my K2 and 30M WBR side by side, both tuned to the same weak signal, as I swap the same antenna between both receivers. Hope you enjoy it. PS – no cats in this one!

October 10, 2016

A WBR Regen On The 30M Amateur Band

Occasionally, I drag out old projects from their resting and display positions on my shelf, plug ’em in, and give ’em a whirl. It’s fun to watch as past home-brew rigs come back to life, and relive the feelings of wonder, as a handful of parts that I soldered together actually receive signals and in some cases, transmit them too. For me, the most wondrous times in building are those initial moments when a new receiver begins to pluck signals out of thin air. Those times of wonderment are often stretched out over a period of time, as a new receiver build progresses. I usually start with the AF stage of a receiver, and build backwards. The moment when I touch the input of the AF amp, and hear a mixture of hum and a general cacophony of broadcast stations isn’t so much a moment of wonder, as one of satisfaction that I can put that stage behind me and get on with building the real part of the receiver. Wherever the point is when RF is being converted to AF, and you’re hearing general atmospheric noise, it’s a magic time for me. It only gets better as subsequent stages are added, and the receiver begins to hone in on a very specific part of the RF spectrum. Mind you, there is something quite wonderful about hearing general atmospheric noise – it feels like an audio window into a wider world around us. I love that!

This is a preamble to the resurrection of the WBR that I built for the 31M broadcast band. Although I was initially happy with it, over time, I had to admit to myself that it seemed a bit deaf. Why was that? The original WBR that I built for the 40M amateur band was sensitive enough. Then I remembered that on the few occasions I had used to it to listen to 41M SW broadcast stations, it had also seemed a bit deaf. Perhaps it was just something about this design that doesn’t do well on AM? With that in mind, I decided to see how my 31M WBR performed on the 30M amateur band.

The existing receiver was already covering 9400 – 10000KHz, and a gentle adjustment of the trimcap in the tank circuit raised the frequency so that it was covering the 10100 – 10150KHz amateur band. The only other adjustment to be made was to limit the coverage to the 50KHz width of the amateur band, as it had previously been set up for the much wider 31M broadcast band. This can be accomplished by adjusting the range of voltages that are applied to the varactor diode, which usually involves nothing more complex than a judiciously placed resistor or two. I placed a 68K resistor between the bottom of the tuning pot and the trimpot, and changed the value of the trimpot from 5K to 10K –

The trimpot is used to set the lower edge of the band coverage, and the 10K value didn’t give me much adjustment range. I managed to get things set the way I wanted them, but suggest the values in parentheses, of 56K for the fixed resistor and 22K for the trimpot, as ones that would give more room for adjustment. If you’re building this from scratch, it might be worth looking into the use of 1N4001’s for the varactors. They’re cheaper and more widely available, and although they don’t give as wide a capacitance range as most varactor diodes, not much is needed when you just want to cover a 50KHz-wide band. You’ll probably need different values for the fixed resistor (if you even need a fixed resistor) and the trimpot. I’d start with no fixed resistor, a 5K trimpot, and go from there, if you do decide to experiment with a different part for the tuning diode.

The resulting receiver works well on the 30M band, with good sensitivity. Indeed, sensitivity is rarely an issue with regens – their main weaknesses are poor strong signal handling, and lack of selectivity. I have not yet heard a signal on my K2 that I couldn’t also copy on the WBR. This confirms my growing suspicion that this design just doesn’t cut it for AM, though it performs well on SSB/CW.

Another feature of this particular WBR version is the circuit of the LM386 AF amp, which provides enough gain to easily drive a speaker, and seems to have less noise than other high-gain configurations of this chip. I’ve heard from folk who built the WBR as described in the original QST article, and have been told that it has low audio. If you’re going to use that circuit, I strongly recommend that you include a preamp stage, as detailed in this post. Even better would be to use the circuit of the 31M WBR which, as well as including a preamp, also has the higher gain and lower noise LM386 amp stage.

If you’re into experimenting, Joel KB6QVI just bought some MD8002A audio chips from eBay. He reports that they have high gain (just like the LM386 in it’s souped-up circuit configurations) but, unlike that chip, is low noise. Like the LM386, it is intended for battery operation, so has low quiescent current. I’m thinking this chip could be a great substitute (not direct pin-for-pin though) for the 386 in many of our favorite well-known simple ham projects. Just a thought 🙂

I’m really happy with how the WBR performs on 30M. It would make a neat receiver for a simple QRP transmitter running from a 10.106MHz crystal. Here are 3 videos. The first one is probably more informative, though the third one includes 2 of my cats 🙂 I do tend to say some of the same things in all the videos, so apologies for the repetition, though I keep it more brief in the first one. If you’re only going to watch one video, watch this first one –

There was a lot of local noise during the recording of this next video. On top of that, I had not set the regen control properly. The set was well into oscillation, making it sound “hissier” than necessary. It also broadens out the response somewhat –

Once again, with this video, I had the regen control set too far into oscillation, widening the response and creating a bit more hiss than necessary. Really, if you’ve watched the other two, the only reason to watch this one is if you want to see some kitty action (2 of my gals feature in this one, beginning at around the 2 minute mark –

That’s the WBR on 30M, and I’m really happy with how it performs there.


October 5, 2016

An End Fed Halfwave Antenna for Portable Ops

I’m a very casual operator, and an even more casual portable operator. My main reason for not putting much effort into portable operation is that when I go out into nature, I want to enjoy my surroundings and not be distracted by radios. It sounds like an excuse, but it’s true. I spend quite a lot of time hunched over the bench and over my radios at home so when I go out, I don’t want to do the same. I’m more the kind of guy who builds small rigs, then operates them from the comfort of my own home. However, I had to take the SST out at least once simply to prove that I can!

The antenna needed to be compact and lightweight, as did the method of matching it. I just didn’t feel like carrying lots of boxes and interconnecting cables up the hill, and having to fiddle with them all once up there. An end-fed halfwave, often referred to by the acronym EFHW, seemed to be a good choice, as it only requires a support at the far end. I saw photos that Steve WG0AT had posted on Facebook of his little EFHW, with the matching unit built into a dental floss container, for a light and compact solution. I wanted an antenna that small and lightweight! Steve referenced a blog post by TJ W0EA, in which TJ detailed an EFHW matching unit he had made, based on the one in his Par End Fedz antenna. This little matching unit, that transforms the high impedance present at the end of a half-wave length of wire into the much lower impedance of 50 ohm coax, consists of a wideband transformer wound on a ferrite core, and a 150pF fixed capacitor. That’s it. Simple and compact!

What to put it in, was the big question. I spent several weeks looking in stores for suitable small containers, and finally decided on a Carmex lip-balm tube. Here it is with the lip-balm removed –

The remaining tube still has a corkscrew-like central element that needs removing –

It is a fairly simple matter to grasp  the corkscrew with a pair of long and slim needle-nose pliers, and push it until it pops out. You can discard the corkscrew, as it is not needed. The 2 parts on the left of the next picture, the snap-on lid and the main cylinder, are what you want –

The following pictures should show you how it all goes together. A plastic cable tie prevents the RG174 from pulling out of the bottom, and a generous squodge of hot glue keeps the toroid in check. If you have a dual temperature glue gun, use the hotter setting –

This matching unit is designed to work with a half-wavelength wire. Some folk build it so that they can change the wire length for different bands. I decided to make this a permanent 20M antenna, so started with about 36 feet, and continued to trim it down until the center frequency was close to 14060, at which point the SWR was 1.1:1. Not bad! I’ll state the obvious by reminding you that any antenna does need to be reasonably clear of nearby objects, particularly anything conductive, in order to make meaningful measurements. Laying it on the ground isn’t going to cut it – you need to suspend one end up in the air and have the antenna clear of obstructions. This is what my final EFHW looked like, all bundled up and ready for the trail, with a 10 foot length of RG174 –

Interestingly, a few days later, I checked the SWR again, only to find that although the center frequency was the same, the SWR at that point was higher, at about 1.4:1. The only thing that had changed was that the first time I measured the SWR, I was powering my MFJ SWR Analyzer from a “wall wart” transformer while the second time, it was powered from internal batteries. I’m thinking that the first time around, the AC wiring in the house was providing a bigger counterpoise and helping to lower the SWR at resonance. It might be interesting to try connecting a counterpoise wire at the rig to see if it reduces SWR any, but I did like the added simplicity of no counterpoise.

How does it work? I bundled the SST, antenna, small sealed lead acid battery, paddle from QRP Guys, and a few other things into my backpack, and cycled up to Vollmer Peak, a local high spot in the Berkeley Hills. I left rather late, had lunch on the way, and by the time I got up to the top, spent about 30 minutes eating trail mix and looking at the view, before realizing that I didn’t have much time. I didn’t get the antenna very high in the tree, and sat on the ground, listening, finishing off the trail mix, and putting out a few CQ’s before heading back down the hill. End result = no QSO’s, but I did get spots on the Reverse Beacon Network from Colorado, Arizona, and Alberta. The antenna works – it’s the operator who performs better in a cozy indoor shack 🙂

There is really only one more thing to try with my SST, and that is, as I mentioned in this post, to add extra filtering between the TX mixer and the buffer/driver. I think that a lot of harmonic energy is making it to the final and being amplified, before being filtered out by the LPF in the antenna lead. Better to nip all those naughty harmonics earlier in the process, I think. If I do any more work on it, that will be the focus.

Thanks to Ian MW0IAN (great callsign) for clueing me in to this PDF on the G0KYA EFHW.


Blog at WordPress.com.