Note – If you have read this article before and are checking back in, it would be a good idea to clear your cache, to ensure that you are viewing the very latest version of this post. I do add material and make corrections from time to time.
It’s been about a year since I finished building The Sproutie, and it’s been a good year. Of all my scratch-built projects, it has been the most satisfying to own. It works well, looks pretty good and also, there is always the lure of possible of tweaks and improvements. This is partially because it’s a home-brew project, and also because it’s a regen :-) It was really enjoyable to build a receiver with the basic circuit architecture taken from the 1930’s, but with a combination of solid state devices and lovely old vintage parts.
I have continued to occasionally purchase vintage reduction drives and variable capacitors. After using a National N Dial for the main tuning control in The Sproutie, I became quite pre-occupied with what, to me, is close to the ultimate dial and drive for an analog receiver – the classic National HRO micrometer-type dial and gear drive. I wanted to find a good example of one of these, and use it in a regen. I also spent quite a bit of time performing Google searches using phrases such as “best regen receiver ever”, and “the ultimate regen”. These are the kinds of things I search for when at a loose end, in the vague hope that I’ll magically find the most amazing regenerative receiver ever designed and built! One very inspirational regen I did discover while searching for the “ultimate regen”, was Jim K4XAF’s build of Bruce NR5Q’s “Ultimate Regen”. What a beautiful receiver! It’s a tube set, built on 2 separate chassis. One chassis contains the main receiver, while the other houses the power supply, the speaker, and the “Selecto-O-Jet” audio filtering. It makes use of a National HRO dial and gear drive for the main tuning, along with National “Velvet Vernier” drives to control the regeneration and variable antenna coupling. Now this was the type of regen that inspires true longing, and convinced me that as enjoyable as The Sproutie was to build and own, I needed to build just one more regen :-)
Initially, I was hoping to use a different type of circuit for this receiver from the tried-and-tested front end used in The Sproutie. I did build VE7BPO’s regen #4 and had some trouble with it picking up a local FM broadcast station. In retrospect, I should have realized that I have had this issue with other simple receivers at this location, until they were cased up and grounded properly. The problem seemed to be a little worse than normal, but this could well have been due to the amplification factor of Professor Vasily Ivanenko’s hycas detector. I gave up far too soon and headed for the security of the front end I used in both the first version of my Sproutie, and the WBR. It is, of course, the circuit used in Nicky’s TRF, as featured in issue 70 of SPRAT (with a few corrections and suggested mods in issue 72). Incidentally, “Bear” NH7SR built a very functional version of Prof V’s Regen #4 which he described in this thread over on The Radio Board.
However, I didn’t just want to exactly duplicate the circuit of The Sproutie, even if the new receiver was going to have a different physical form and different hardware. This new receiver would have to have some alternate type of circuitry that would make it worth building. I was interested in trying a different type of filtering in the audio chain, and a tip from Prof V in his Solid State Regenerative Receivers group on Google+ clued me in to a great tool for designing active audio filters (more on that later). The pieces were beginning to come together. I had a bunch of NE5532’s in my parts stash that had sat unused for a couple of years and it struck me that a regen which utilized a series of active audio filters for different bandwidths, switched from the front panel, might be an interesting idea for a receiver. The LM380 output stage I had used in The Sproutie works well, so I saw no reason to change it. It is fairly low noise, a welcome factor that makes it possible to listen to a receiver comfortably for long periods of time.
Here’s the block diagram of The Sproutie MK II. As it contains 6 separate AF filters, I decided to also switch the +ve supply to the filter. A 5532 active filter draws about 7mA (14mA if using 2 x op-amp stages). Although it’s not a lot of current, it’s a fairly significant amount relative to the total consumption of this receiver if all 6 filters are continuously powered. One of the reasons I prefer solid state over tubes is the power efficiency, so no reason to keep all 6 filters powered if only one is being used at a time –
Fig 1 – Block diagram of The Sproutie MKII. Note that S1a, S1b and S1c are all part of the same rotary switch.
The front end, as I mentioned, is exactly the same one I used in the original Sproutie. It is the one used in Nicky’s TRF featured in issue 70 of SPRAT. I thoroughly recommend joining G-QRP. Your initial membership includes an archival DVD of past issues of the club magazine SPRAT, which is a very valuable resource for homebrewers. If you have access to this archive, you should also take a look at issue 105, in which a slightly different version of the same receiver is featured. It employs a simple passive LC audio filter, if you’re not keen on the extra complexity that my version here entails.
Here’s the schematic of the front end. The oscillator tank circuit has been simplified to just one variable capacitor, and all details of the plug-in coils removed, purely for the purposes of making the circuit a bit easier to understand. If I drew the octal coil socket without the coil (as I did with the schematic for The Sproutie) it would make the process of understanding the circuit diagram a bit less intuitive –
Fig 2 – The Sproutie MKII front end, with details of plug-in coils and fine tuning capacitor removed for simplicity.
Here are details of the coil base, using an octal tube socket. You can use any pin configuration you like – this is the one that worked for me. It is the same configuration as used in The Sproutie –
Fig 3 – Plug-in coil base wiring
The final AF amp is a simple LM380 circuit. It’s easy to build, is fairly low-noise, and it works. As well as a phone jack, I included a jack for an external speaker on the rear panel. It took me a while to figure out how to wire the internal speaker and the 2 jacks properly. I wanted the internal speaker to cut out if either headphones or an external speaker were plugged in. I also wanted the the external speaker to cut out if the headphones were plugged in. It’s a simple problem really, but simple things often elude me. I got there in the end –
Fig 4 – The AF output stage. The “bass” switch only gives a very gentle lift to the lower frequencies. The effect is so subtle that you won’t be missing much if you leave it out.
The thing that makes this receiver different from the original Sproutie, electrically speaking, is the bank of switched active audio filters. If you don’t want to be bothered with building multiple filters, and switching them all with a switch, you could permanently wire just one filter into the circuit. Another idea would be to replace this bank of switched filters with an adjustable filter made from op-amps, with the center frequency and bandwidth controlled by potentiometers on the front panel. Once you bring op-amps into the mix, all sorts of things are possible. Another idea suggested by Bear NH7SR, is a 5KHz audio notch filter, which could be quite useful for AM SWBC listening. The design tool that made all this happen for me was by Texas Instruments (thanks Prof V). There is an online version called Webench Filter Designer. It has a user-friendly interface that actually made the process of filter design harder for me than the offline software they also offer, called Filter Pro. Use which one works best for you – they are both accessible from this page (opens in a new browser window). Of the two, I recommend Filter Pro. You can use this software to design low-pass, high-pass, bandwidth, allpass (time delay) and notch filters. I stuck with low-pass filters. I was tempted to try a bandpass design for the CW filter, and may still do at some point. The CW filter I constructed was the very last filter out of 6. By that time, I didn’t have the patience for the slightly more complex design of the bandpass filter. I also rationalized that I might need to tune through a CW signal to hear the other side of it, if trying to escape QRM, so a lowpass would make this easier, as I’d be able to hear the signal all the way through to zero-beat and out the other side. This might simply have been my excuse for not wanting to build a bandpass filter :-)
I wanted a “straight-through” position to give me something to compare the other filters to. All the filters, with the exception of the narrow CW filter, were designed with a 6dB gain, so I designed my “straight-thru” filter with a 6dB gain also, so I could step through the bandwidths seamlessly. If doing this again, I would have given the filters a bit more gain. I’ll explain why later. Dan N7VE gave a talk to the Arizona QRP Scorpions a few years ago on (among other things) designing active audio filters. It’s definitely worth taking a look at his presentation, which is available here. In fact, I wish I’d paid attention to it before embarking on designing the filters for this receiver, as I would have tweaked some of the resistor and capacitor values a bit. Dan explains how it’s desirable to keep the resistors in the main signal chain fairly low in value, to avoid noise. He recommends trying to stay under 1K. I only read the presentation before designing the very final filter – the CW one with a cut-off of 700Hz – so while my resistor values in that filter are nice and low, they are not quite so low in the others (though in my defense, they are not atrociously high either).
Here’s the first, and widest filter. As far the ear is concerned, it’s not really a filter, as it has a cut-off set at 20KHz, with a gain of 6dB –
Fig 5 – The “straight-through” filter (an LPF with a cut-off of ~20KHz)
I wasn’t interested in the shape of the response as, for this stage, all I wanted was effectively an unfiltered stage with a gain of 6dB. For this reason, I used just one half of a dual op-amp 5532 package as a real-pole filter. Filter Pro doesn’t show the power supply and biasing arrangements, so I added the 2 x 47K resistors to keep the input biased at about half of the supply voltage. I also added the 10uF capacitor, which keeps the bottom end of the 1K resistor at ground potential for audio signals, while blocking the DC bias. I also added the lowpass filter formed by the 10 ohm resistor and the 100uF electrolytic on the supply line, as well as the 0.1uF ceramic RF bypass cap on pin 8 of the IC (mounted close to the pin). I don’t know how essential these 0.1uF caps are, but the datasheet suggests them, and they can’t do any harm.
The other filters were all 4th order low-pass filters (2 stages = 1 x 5532 dual op-amp package), with the exception of the 2.4KHz filter, which was an 8th order low-pass filter (4 stages = 2 x 5532 dual op-amp packages). The 8th order filter has a sharper cut-off, of course. Feel free to design your own filters, with the help of Filter Pro, for whatever cut-off frequency and rate of roll-off you wish. I’ll show you the R and C values I used for my filters but you might want to fiddle around with the software and come up with your own values that keep the R values in the main signal chain at or below 1K, if possible. The resistors in the first stage of the filter are particularly important, as the noise they produce is amplified more than noise produced in later stages. Just click on a component in Filter Pro, enter a different value, and hit return to see what new values of the other components the software has calculated. A bit of trial and error should get you close. Also note that you can specify the series of resistor and capacitor values you want to use (E96, E48, E12 etc), and watch how the filter response curve changes as you change the tolerances and values.
First of all, here’s the gentler roll-off 4th order filter that uses just one 5532 8-pin dual op-amp IC – or use the op-amp of your choice. I chose the 5532 because I had a bunch of them in my parts stash and because they are the 2N2222 of the op-amp world – plentiful, reasonably priced, and all over the place –
Fig 6 – Schematic for the gentler roll-off 2-stage LPF
Here are the component values I used for my 4th order filters –
Fig 7 – Component values for the 2-stage filters in my Sproutie MK II
For a sharper roll-off, an 8th order filter, which uses 2 x 5532 dual op-amp packages (or equivalent) –
Fig 7 – Schematic for the 4-stage sharper roll-off LPF
The 2.4KHz 8th order filter I used, although a bit on the narrow side for SSB, is good for listening when there are nearby stations higher in pitch that need cutting out. If you think about it, this 2.4KHz LPF is going to sound roughly like the 2.1KHz filter in a regular SSB rig. The reason for this is that your regular SSB filter is a bandpass filter, with the bottom edge being set to cut off at about 300Hz. This means that a 2.1KHz SSB bandpass filter will pass frequencies up to about 2.4KHz (2.1KHz + 300Hz). Here are the values I used in mine –
Fig 8 – Component values for the 4-stage filter
After I had built the receiver and all these filters, and done some listening, I concluded that for SSB and CW, a bit more filter gain would be helpful. The set has plenty of gain when listening to AM but on CW/SSB, the RF gain has to be wound right down to prevent the oscillator pulling. This creates a need for more AF gain in the CW/SSB modes. At the time of writing this, I have only just finished building this set and have no enthusiasm for building more filters. I actually had to build 8 filter boards to get the 6 that I used, and 3 front end boards to arrive at the final one. Together with the physical side of the construction, I am tapped out right now and have no desire to construct anything else at all for a while!
If you want to use this receiver mainly for SSB and/or CW, you may want to experiment with the value of the NPO capacitor in the front end that connects the hot end of the main tuning coil to the base of the 2N3904 oscillator transistor. It is listed on the schematic as being 39pF, and that is the value I used. However, it is possible that a lower value will cause the oscillator to pull less on strong signals. Of course, the lower value might also reduce the signal strength into the detector which will put you back to square one. It’s worth trying though. I’d be tempted to try a value as low as just a few pF. Remember that changing this capacitor will affect the frequency coverage – particularly at the top end of each range.
When building the filters, I originally built the 700Hz CW filter with a gain of 6dB, like the other filters. The idea is that if they all have the same gain within their passband, the operator can step through the different bandwidths without a change in the volume of the wanted signal in the speaker. This was the way it worked except with the 2 narrowest filters. The 2.4KHz 4 stage filter had a slight, but noticeable drop-off in volume. The effect was very pronounced with the 700Hz filter – so much so that I redesigned it with a gain of 20dB and still found that there was a slight drop-off in volume within the passband as compared to the other filters. I don’t know the reason for this. EDIT – Thomas LA3PNA Tweeted the following explanation – “The perceived loss when changing filters is because the power delivered to your ear is 10log(BW of filter) and less with less BW. So basically, the reduction in noise makes it sound like the volume goes down” He also gave a very useful tip for adjusting the gain of the filters so as to preserve the perception of constant volume – “I like to add gain in a filter circuit after the formula 20log(bw/orginal bw) for AF filters” That is very useful information Thomas. I’m a little tapped out after building The Sproutie, but if and when I decide to revamp the filter bank, I’ll be paying attention to this formula.
I may, at some point, rebuild the 3 narrowest filters with higher gains. If that ever happens, I’ll report the results here in this post. Incidentally, at this point, allow me to say one more thing about the filters. If building and wiring up all these filters sounds like it is making the construction of a regen overly-complicated, I can definitely sympathize. If you want to use this set for CW, SSB and AM and you want to permanently wire in just one filter, I’d go for a 4th order (2-stage) LPF with a cut-off of 3KHz. The one I have is perhaps a touch wide for SSB (it’s roughly equivalent to a 3.3KHz passband filter, as explained earlier) and a bit narrow for AM broadcast, but it’s a good compromise for both. If it were the only filter I had, I know I would get used to the sound of it. As for the gain, mine has a gain of just 6dB, but I’d like to up it in order to have a good volume when turning the RF right down, as is necessary to prevent oscillator pulling on SSB/CW. I can’t know until I’ve tried it, but I’m thinking something along the lines of 26dB gain. Just make sure to be careful when on AM, as you may find that you have way more gain than you need – so keep an eye (and a hand) on that RF gain control.
A big part of the inspiration for building this receiver, as I mentioned earlier, was the physical form of K4XAF’s version of NR5Q’s Ultimate Regen. In the search for a National HRO dial and gear drive in really nice condition, I bought several, and finally came up with a dial and drive combination that just cried out to be included in this receiver. This gear drive has a shaft rotation limiter, which was perfect, as the tuning capacitor I wanted to use didn’t have any kind of rotation limiting built in – it was the capacitor in the first photo in this post – a Hammarlund MCD-50-M. The final M stands for midline, referring to the fact that the off-center shaft and shape of the vanes help to make the tuning a lot more linear than with regular variable capacitors. With a standard capacitor, you’d find that the frequencies would become very compressed at the top of the tuning range i.e. the tuning would get a lot more fiddly. Try to get a midline unit. I believe they also go by other names, depending on the manufacturer.
Of course, a big dial and gear drive need a big chassis, and Terry from Seaside Chassis, who made the chassis for The Sproutie, came to the rescue again. I decided to use a chassis and front panel that would be compatible with 19″ rack cabinets, for a variety of good enclosure options. A chassis that big needs to be fairly thick in order to still be stout and solid. Terry does offer the use of 12 gauge aluminum for bigger enclosures, and I wanted this receiver to be big and solid (although compared to your average boat anchor, it’s still relatively light). As well as a large, stout chassis, I decided that I wanted to try designing a custom front panel with the services of Front Panel Express in Seattle using their free design software. Right at the beginning of this whole project, in the first month or two of 2015, I downloaded their software and casually laid out a very rough front panel, mainly for the practice, and the fun of learning something new. As the project progressed, I’d spend a few weeks working on circuit boards, then go back to the front panel, then do a bit of work on the plans for the chassis, to send to Terry. I had an idea that, with a bit of luck, I’d complete the whole thing in or around the fall, and that’s how it worked out. At no point did I rush though. Why rush? Besides, the longer a project takes, the less it costs per month. I could see that building this regen in the way I had chosen to build it was not going to be a cheap affair, so I took my sweet time.
Here is the chassis as it arrived from Seaside Chassis, along with 2 side braces for supporting the front panel, 2 mounting brackets for the main tuning capacitor, and 2 mounting brackets for the regeneration pot. I only needed 1 of each of these brackets, but like to have extras on hand. As it turned out, an extra bracket was needed to help secure the main tuning capacitor which I forgot to ask Terry for, so I put in an extra order. The shipping from Canada dwarfed the cost of the bracket but at this point, it was easier to ask him for it than to find someone local and besides – I just wanted him to fabricate all the chassis components. Terry’s work is first-rate. It’s good to give him as much relevant information as you can. Simple drawings with penciled-in dimensions work well. If it’s important to you, remember to take into account the thickness of the aluminum if there are any dimensions that are particularly critical. Also remember that he is bending and fabricating these components by hand, so allow for a certain amount of tolerance in the final dimensions. Having said that, the chassis he supplied was remarkably close to the exact dimensions I requested, and within the tolerances I had allowed for. If you have any dimensions that are particularly critical or non-critical, I think this is all good information to pass onto him when making your request –
Figuring out exactly where to drill holes for controls in front panels and enclosures usually takes quite a lot of time. It’s a bit like a game of chess in that every decision you make affects everything else down the road. To make matters tougher, I have trouble thinking about more than one thing at a time, so juggling all the variables in my head takes a lot of thinking, measuring, and drawing. For front panels, I always draw the shape of the panel on a full-size sheet of paper, and place all the knobs and controls on it to see how they look in various configurations. Just when I think I have it right, I leave it and walk away, often overnight. On returning, I inevitably come up with an improvement or two. Building something like this is all baby steps for me. I am impressed and amazed by builders who claim to be able to throw something like this together in a few afternoons – this one took me over 6 months. Heaven knows how long a more complex receiver, such as a multiband superhet, would take me.
I took a great deal of time and care in designing the front panel. They are worth every penny, but they are not cheap. I didn’t want to make a mistake that would result in having to re-order the whole thing. So after checking, rechecking, going to sleep, then waking up and rechecking again, I went through this whole process several more times before finally clicking “order”. A week or so later, this beautiful 4mm thick aluminum panel arrived via UPS, packed with a little bag of gummie bears –
The front panel as it arrived from Front Panel Express, vacuum packed to a stiff baseboard. The metal ruler is 18″ long (the panel is 19″ wide).
Look at this beautiful, black anodized front panel!
I just couldn’t get enough of this thing when I first saw it –
There were some scratches in the black finish on the rear, but this is normal. I later found out that it is possible to add a note when ordering, to ask the people working with the milling machinery to take extra care with the back side of the panel. The front surface is guaranteed, but not the back. I decided I was OK with the rear of my panel as, well, it was the rear, and the bottom half of it would be in direct contact with the front of the chassis anyway –
You’ll notice a number of “blind” holes milled on both the front and rear. The panel is so thick (4mm) that controls sticking through both the chassis and this panel wouldn’t protrude far enough for the nuts to thread onto the bushings. For the RF gain, AF gain and filter rotary switch, the blind hole was milled on the front side, as the knob would cover it. For the phone jack and bass switch, the blind holes were milled on the rear. Here’s a close-up of the blind hole on the rear side for the bass toggle switch. You’ve probably figured (if you didn’t already know) that a “blind” hole is one that doesn’t go all the way through the panel –
After the initial euphoria of receiving this fantastic front panel had subsided a little, it was time to put some time and labor into making all the remaining cut-outs in the chassis. I had asked Terry to make the holes for the octal tube socket and the main controls, but there were others that still needed to be done. My usual method of making non-standard cut-outs and holes is very time and labor-intensive, but it works quite well. I mark the edges of the cutout with a pen or pencil, then with a hand-drill, drill lots of small holes around the perimeter. Then, with an old screwdriver, I knock out the piece of metal in the center, and clean up the edges with files, usually using a bastard file first, and finishing off with something finer. These photos should help illustrate the process. The speaker cut-out was inspired by a WW2-era British military R107 receiver that I owned as a teenager. It is simple – just 4 large holes arranged in a square. This is the “during” photo, showing the series of small holes drilled around the edges of the holes. The rectangular cut-outs to the right were made using the same technique, incidentally –
– and after –
Here’s another photo, taken a bit later during the assembly, showing the placement of some of the main components. This particular National HRO NPW gear drive, unlike most that I have seen, has a shaft rotation limiter. The tuning shaft is a little on the short side. I needed to mount the gear box as close to the front panel as possible in order to be able to mount the dial properly. If you scroll back and look at the photos of the front panel, you’ll see there are 3 smaller holes located around the main hole for the gear drive. These holes helped in locating the gear box as close to the front panel as possible (the 3 screw heads fit into the 3 smaller holes on the front panel). Most of these gear boxes don’t have this rotation limiter, so the extra holes won’t be necessary. Also, do you see the aluminum shaft couplers on the regen pot and fine tuning capacitor? Those are quality parts personally machined by John Farnsworth KW2N. The one on the right is a standard 1/4″ to 1/4″ coupler, while the one on the left was made to order. It couples the 1/4″ shaft of the 10-turn regeneration pot to a short 3/16″ shaft that the National knob fits onto. I wanted to use the same type of National Velvet Vernier knob and escutcheon plate for the regeneration that I used for the fine tuning, but I didn’t want to use the 5:1 reduction drive. I wanted to use a 10-turn wirewound pot instead, as I like the feel of those pots. From the front (as you will see in later photos) the 2 National knobs and escutcheon plates look the same. However, the knob on the left is connected directly to the 10-turn pot and not to a National Velvet Vernier reduction drive. The black escutcheon plate for the regen control is spaced away from the front panel by one washer thickness, and bolted to the front panel with 4-40 hardware. It is not used for anything, other than looks.
Now, let’s look at some of the boards. They were built, as always, with W1REX’s very useful MeSQUARES and MePADS. This is the AF output stage and the 4KHz filter mounted on one board, and installed in the chassis. The idea was that this board, together with the main RF board, would form a working receiver, after which I could build and install the other filters, one by one –
Mounted above the AF output stage, on the stand-offs, is this next filter board, carrying 2 LPF’s. The first filter to be built was the 6KHz one –
Next came the 3KHz filter (in the foreground of the next shot). The grey rectangular poly capacitors were from Tayda Electronics. Thier prices are low, and the caps seem good. The resistors are 2 types – either 5% carbon film from the parts stash I had as a kid in England in the early 80’s. They lasted a long time, but I am beginning to run out of them. The others are newly-acquired Xicon 1% metal film parts, purchased in lots of 200 from Mouser –
The same board, taken from above (3KHz filter on the left, 6KHz filter on the right) –
The same board, with the 3KHz and 6KHz filters, mounted in the chassis above the 4KHz filter and AF output stage –
Here are 2 shots of the 4-stage (8th order) 2.4KHz LPF, with temporary leads in place for testing. It’s quite the QRM-buster –
At this point, allow me to introduce the main RF board. Electrically, it is exactly the same as the one in the original Sproutie. I tried a couple of small mods but went back to the original. So – nothing new here, except for a small physical detail that I learned from my experience with The Sproutie. There are 2 pads on the board that connect to the octal tube socket with (ideally) short, stiff wires. In the original Sproutie, I used very short lengths of solid 16 gauge wire. They were so stiff that, over time, with repeated insertions of coils, the wires placed enough stress on the pads to detach them from the board. It took me a while to figure out why the dial calibration was suddenly off by about 10KHz. The pads had only separated from the board by 1mm at most, so it was hard to see, but it was enough to throw off the dial calibration, and cause it to change slightly in an unpredictable fashion. I did 2 things to remedy this. The first was to replace these 2 wires with thick stranded wire, which I tinned thoroughly. The tinning stiffened the wire, but it still had more flex than the original solid 16 gauge wires. Secondly, I removed the 2 pads, and re-attached them with epoxy instead of superglue. Problem solved! When I built the Sproutie MK II board, I attached these 2 pads with epoxy (I used JB Weld). Superglue gel was used with all the other pads, as before. The 2 pads in question are at the very front edge of the board in the next shot. They are the second and third pads from the right. Missing from this shot is the 0.1uF capacitor that couples the audio to the next stage. There is also one extra capacitor that was part of a mod I later uninstalled. I’m showing you these shots to give you the overall idea of layout, what it looks like, and because it’s fun looking at circuit boards. For absolute accuracy of the circuit, follow the schematic –
The next 2 shots are the same board, but at an earlier point when I was using 1uF caps for interstage coupling. They are the 2 blue box-like caps. They didn’t make it to the final version of the board –
Here’s a wider view of the underside of the chassis at this stage of the construction, showing the main RF board wired to the octal tube socket, as well as the AF output stage with 4KHz filter, the 3KHz and 6KHz filter board on top of it, and the 2.4KHz filter board sitting on it’s own for the time being –
A closer view from a slightly different angle. At this point, it was beginning to dawn on me that keeping all this wiring tidy would take a bit more work than I had anticipated. I never did get the wiring as tidy as I wanted, but it’ll do –
This one’s a bit boring. It’s the “straight-through” real-pole LPF with a cut-off frequency of 20KHz, shown installed in the chassis on top of the 2.4KHz LPF –
and the 700Hz CW low-pass filter –
Here are all those filter boards stacked on top of each other. Looking at the left-hand stack first, from the top down is the 700Hz filter, the 20KHz “straight-through” filter and, at the bottom, the 2.4KHz sharp roll-off filter. On the right-hand side is the 3KHz and 6KHz filter board, with the 4KHz filter and AF output stage board on the bottom. You can also see the 6-position rotary switch that selects between the different filters. I had no trouble finding a 6-position switch with 2 poles but when I decided to also switch the +ve supply line to the filters, finding one with more than 2 poles proved tricky. I finally found it from a supplier of parts for musical instruments. It is distributed by AllParts, and is part number EP-0920-000. It is a 6-position 4-pole switch (one pole goes unused). Prices vary a bit, so search around for the best deal if you want this particular switch. If you want fewer filters, then you’ll probably find it easier to locate a switch that has 3 or 4 poles and 4 positions or less. This is the finished receiver, by the way. Well, finished for the time being – until I decide I just have to modify something –
Some more views of the underside of the finished receiver. You’ll notice that I designed a rear panel too. That also came from Front Panel Express – a more detailed view of it is coming later. The speaker is an 8 ohm, 4 inch, 6 watt unit made by CUI, model # GF1004. I got it from Digi-Key, part # GF1004-ND. Before finding this speaker, I purchased one from a company well-known for supplying vintage radio parts. It turned out to be very lightweight, with a small magnet, and generally rather disappointing. I liked the speaker I used in The Sproutie, so got the 4″ version of that one instead (the one in my Sproutie is a 3″ version). The aluminum speaker grille was custom cut by speakerworks.com –
Coils were constructed in the same fashion as the coils for The Sproutie. In fact, the pinouts used on the octal socket are the same, so my Sproutie coils work in the Sproutie MK II, though they cover a wider range, due to the greater maximum capacitance of the tuning capacitor I used – a 2 x 50pF instead of the 2 x 35pF used in the original Sproutie. I decided to wind a complete set of new coils for this receiver. As of writing this post, I have 6 coils wound with a few more to go, as needs and desires dictate.
With The Sproutie, I used nylon hardware to secure the larger T68-6 toroids for the lower frequency bands. On the higher frequency bands, I used T50-7 toroids, and secured them with hot glue. This time around, I found that hot glue worked perfectly well for securing the T68-6 toroid cores too, so I used that technique exclusively. It’s faster and easier than using nylon nuts, washers and bolts. I didn’t think it would be the case, but if you need to re-make a coil, you can peel/break the glue off and re-use a tube base. In fact both the coils in the photos below were made with bases that were used at least once before –
I like these ceramic bases, because they are just a little higher than the phenolic ones, offering a bit more protection to the toroid. Wherever you get yours from, if they’re ceramic, they may well be the same ones as these, as most of these bases and sockets seem to be made in China these days –
“Take us to your leader”
You’ll need to figure out the exact details of your coils with the help of online calculators (I like the ones on W8DIZ’ site) and good old trial and error, but here are my details – they should give you a start. Remember to take into account the values of main and fine tuning capacitors, if they are different from mine – and the value of that 39pF capacitor between the tank and the base of the oscillator transistor, if you try a different value. I’ll update this table as I wind more coils. The plan is to wind general coverage coils up to about 21MHz or so, and a few more coils for specific bands. It is much easier, with the aid of the dial calibration graphs, to pinpoint exactly which 5KHz channel you are on, when the band coverage is limited to 1MHz or less. With the 20:1 reduction ratio of this National HRO drive, and the large, relatively massive dial, I found it quite easy to tune in stations even on the 13500-18300KHz coil, which spans almost 5MHz. For pinpointing which 5KHz “channel” I am on though, a general coverage receiver to listen to the oscillator of The Sproutie is more reliable (and faster) than reading dial calibration graphs.
For dial calibration, I use a piece of freeware called Graph. You won’t be able to read the following graph, as it’s a bit small. The original is a bit larger, and the software has an option for zooming in on a particular area of the graph. This is the dial calibration graph (so far) for my 5475-8450KHz coil –
I guess it’s time to do a reveal and show you what this little feller looks like from the front. You wouldn’t think it, but I spent a great deal of time on the front panel, figuring out the exact placement of all controls, placement of the lettering, and fonts. I was looking for a specific type of vintage knob for the RF gain, filter, and AF gain controls, but didn’t find any in good condition while building the receiver. Then I found some knobs on clearance at my local Radio Shack. Those are brand new RS knobs, but I think they look good and fit in well with everything else on the front panel. My main concern was to not “overdo it”. When sitting at the computer with the Front Panel Designer software running, it’s quite tempting to go overboard on the lettering, or try a colored panel, and colored letters in a fancy font. Just because you can do something though, doesn’t mean you should, and I wanted a front panel that was understated, functional, and that would still look good, regardless of how my personal aesthetic might change. Minimalism is the key, though the one extravagance I did allow myself was the larger “Sproutie MK II” declaration, and my callsign. I did try my callsign in red but decided that it looked gaudy. Best to play it safe, I think. I was also concerned that the finish might be a bit too shiny or glossy, but it turned out to be matt with a slight sheen. I’m very happy with how this looks –
Although the regen and fine tuning knobs, and escutcheon plates look the same, the ones on the right are attached to a National “Velvet Vernier” 5:1 reduction drive, via a fairly long coupling shaft. The knob on the left is connected to the 10-turn regen pot (via a 3/16″ to 1/4″ shaft coupler), while the black escutcheon plate is spaced away from the front panel by washers, and attached to it in a fixed position with 4-40 hardware –
Here’s The Sproutie MK II with her little sister, for size comparison –
A quick word about that regen control. On first installing the main RF board, one filter, and the AF amp into the chassis, I noticed that occasionally, when receiving a strong carrier, I’d hear a ringing in the speaker. My first thought was that it was microphonics caused by physical feedback between the internal speaker, which was bolted to the chassis, and some part of the circuit. Plugging in headphones didn’t cause it to go away, however. Undeterred, I continued building, and it was only after finishing the whole receiver, that I realized what was going on. I discovered that if I hear a ringing, all I have to do it back off the regeneration control a bit, and it disappears. I think this ringing is due to the high Q of the circuit when set right at the threshold of oscillation. If you recall Dan N7VE’s presentation on filters that I referenced earlier, he talks about how ringing in filters is caused by abrupt phase changes at the edges of the passband. The cure, when designing them, is to limit the Q in any one stage. Similarly, if you experience ringing in your regen, backing away from the critical threshold of oscillation will lower the Q of the circuit, and should solve the problem. Fascinating. I’ll be interested to hear if any other regen operators have experienced this. My guess is this would be less likely to happen in a regen that utilizes a bipolar transistor for the detector (or combined oscillator/detector if it’s just one device, unlike this design).
I had to try a shot from a lower angle, for that authoritative look. When I’m spinning that big old dial and listening to CW, I can almost kid myself that I’m intercepting enemy broadcasts for the valiant code-breakers at Bletchley Park. In reality, I’m usually just listening to some ham tell some other ham what the weather is like at his QTH! –
The rear panel (thanks again, Front Panel Express) –
I like this receiver as it is, with the partially open chassis. From using The Sproutie, I have become used to seeing the vanes of the variable capacitor rotate as I tune the band, and I like that. I like seeing these vintage radio parts in action. However, I did learn from The Sproutie that whatever isn’t covered picks up dust – and when you’re living with 3 cats, 2 of whom are long-haired, a lot of cat hair too. I designed this receiver so that it would fit any standard 19″ enclosure that is also 6RU (rack units) or more high. The first plan was to make use of a hack (as the kids call it) of an IKEA product to make a low-cost rack cabinet. The IKEA Rast nightstand is the right size, and only costs US$14.99. For that, and the cost of a pair of rack strips, you can have a rack cabinet that is either 6RU or 8RU high, depending on how you construct it. A Google search on “IKEA Rast rack cabinet” or similar will yield a lot of sites and info on how to do this useful mod.
That was my plan for this receiver until I came across nice-racks.com. David Tatelbaum makes beautiful studio racks out of his workshop in Massachusetts. He uses furniture-grade pine, though he will use other woods if you request them. The mahogany racks look gorgeous (but they do cost a bit more, of course). From his website –
“Nice-Racks are constructed of solid Pine furniture-grade panels…not just pressed wood or particle board covered in a laminate like some studio racks, but 100% real wood. The panels are cut to size and the components joined together securely using pocket-hole construction and self-tapping pocket-hole screws. The front sides and top edges are rounded, then the racks are sanded and stained. The finish is a clear matte enamel, scuff sanded between coats, to preserve the look and feel of real wood. Finally, hardware is installed and the fully assembled rack is boxed up and shipped out to its new home.”
After finding his site, I was hooked. Yes, it was going to cost more than the cheap IKEA hack but at some point during the design and construction of this regen, I decided that I wanted it to look really nice, and expense was going to be a secondary concern. I was going full-hog on this. Besides, it took me over 6 months to plan and put together and spreading the cost over that time, the cost per month for my hobby was actually quite reasonable – especially if I take into account all those movies and dinners I didn’t go to because I was at home building! Incidentally, David’s racks are most definitely not expensive. When you consider the cost of your standard rack cabinet – the ones you find in music stores that are made of heavy particle board and covered in black veneer, his racks compare in price very favorably – and they are vastly nicer. They’re not ideal for the rough life of touring but for a home studio, they are perfect – and very good-looking.
OK, so time for the big reveal. This rack cabinet makes The Sproutie MK II look so great –
David also fitted rack rails to the top half of the cabinet at the rear, and supplied a 3RU-high steel panel to help enclose the receiver, and hopefully keep the cat hair away. On the left interior side you can see the recesses for the self-tapping pocket-hole screws that hold the whole cabinet together –
And in case anyone ever wonders who made this fantastic cabinet for my regen, David left his mark. What a quality job! –
The National HRO gear box does have a small amount of backlash – even when I apply as much tension to the anti-backlash gear as my poor little fingers can manage. At first, the backlash was something like 1 – 1 1/2 dial divisions. After increasing the tension on the anti-backlash gear as much as I could reasonably easily manage without the use of tools, the backlash, though still there, decreased to about 1/2 a dial division. It is a small amount, and also predictable, so not really a problem. For the purposes of dial calibration, I always turn the dial in the direction of increasing frequency before taking a dial reading. This ensures consistency in the readings.
Some videos of The Sproutie MK II in action. In the first one, the towels are on top because Sproutie, my 3 1/2 year-old kitty (aka Sprat The QRP Cat) likes to sit on top of it, and I don’t want her claws to do to the wood what they have already done to my leather sofa –
The Sproutie MK II on 49M, 41M and 40M (though mainly 49M) –
There are some more videos of the Sproutie MK II, showing how a regen can be used in exalted carrier mode to enhance reception of weak AM stations, and on the 25M band, in this slightly more recent post.
There are a couple of things about my Sproutie MK II build that I’d like to change. The first is that, especially at the higher frequencies, the set is slightly sensitive to physical shock. A knock on the cabinet will shift the frequency slightly. I don’t recall noticing this effect with the original Sproutie, although to be fair, I didn’t do as much listening to SSB and CW with it as I do with the new receiver. I think one reason for this slight frequency shift may be the fact that the thick wires connecting the 2 stators of the main tuning capacitor to the coil socket are longer than in the original Sproutie. Although the effect is only slight, it is there, and that bugs me. It may also have to do with the much larger chassis, meaning more metal in the vicinity of a tuned circuit, that flexes when a physical shock is applied. The more I think about it, the more I think this second factor is the main reason. In practice, it is not a problem, but it is there, and I’d like to reduce it, if not eliminate it completely. Fixing a bottom plate to the chassis may help in this regard. Incidentally, banging the desk on which the Sproutie MK II is sitting has no effect. The rubber feet probably help a lot.
The above phenomenon is responsible for an interesting ringing effect that happens occasionally when the regen is set close to the critical point for receiving AM. It only happens with the internal speaker, so is being caused by sound from the speaker vibrating the main chassis. I did mount the internal speaker on small grommets, but this didn’t cure the issue. I have been looking for a reason to purchase a Palstar SP-30B external speaker, and this may be it! This is what the ringing effect sounds like –
The second thing is that, because the cabinet is wood, the receiver is not completely shielded. This would be useful were I to wind a coil for the 2-3MHz region and use The Sproutie MK II in conjunction with crystal-controlled converters to cover specific bands. This is a Regenorodyne approach, like Gary WD4NKA’s inspiring Regenerodyne receiver here. It would also be nice to reduce the possibility of picking up very local QRM in the shack. I could achieve better shielding with my Sproutie MK II by either simply housing it in an all-metal rack cabinet, or by cladding the interior of the existing wooden rack cabinet with thin metal plate or mesh. There is absolutely no hand-capacitance effect when using the set, due to the metal front panel, but when my cat Sprout jumps up on top of it (as she often does) the frequency shifts by about 20Hz. This is also due to the lack of shielding on top of the set. Again, it is not much, but it is there.
Sproutie (aka Sprat The QRP Cat) and The Sproutie MK II. Her contribution to the dial calibration of this receiver was carefully knocking the plug-in coils off the top of the receiver and watching them hit the floor.