Sept 1, 2016:
Revised: v1.5

ZL2PD Builds the Lydford 40m QRP SSB Transceiver

Front view of Lydford transceiver
The Lydford transceiver was a HF QRP SSB transceiver kit produced by Walford Electronics in the UK during 2013-15. The kit instructions allowed it to be built for use on any single band from 80m to 20m. I built my version for use on 40m during 2014. On this webpage, I describe my construction of the kit and some of the minor modifications and accessories I made for it.



With Christmas approaching, I began to think about buying a transceiver kit. Of course, I could have started a new transceiver design and build project from scratch, but I thought I'd take the easier route of building a kit, and then use it as a 'test bed' for some other ideas. Looking around at the available options, I settled on the Lydford transceiver kit from Walford Electronics. It can be built for any of a number of HF bands, but I decided I would built it for use on the 40m band.

The parts in the kit will allow you to duplicate the kit as pictured on the Walford website. That's the photo you can see below. i.e. The parts supplied in the kit provide  the basics only, including a simple PCB for use as the front panel and for mounting the main tuning capacitor (on the left), the fine tuning control (centre) and the AF gain (volume) control (right) as well as sockets for the microphone and headphones or external speaker. Obviously, I've added some extra stuff for my version.

Watford's Lydford transceiver kit
Figure 1 : The Lydford transceiver as pictured on the Watford website  

I've mounted my Lydford transceiver board in an aluminium enclosure obtained from Maplin in the UK, added a DDS VFO which I designed specifically for the transceiver (and
described elsewhere on this website here), added my own CW active filter and AGC circuitry, and some front panel artwork to make it all look pretty.

Figure 2 : My Lydford transceiver in its enclosure with digital VFO and all ready to go

But before we get to that, here are...

Some Initial Comments about the Kit

I like this QRP transceiver, I really do. It has some interesting design features, and it really does perform up to expectations. As a kit, however, it is not the easiest to put together. It was clearly identified as a more complex kit on Tim's website - It's found in the 'Advanced' section. Mind you, I managed to built it so it wasn't  too difficult, and the transceiver works really well.

And it was not an expensive kit . It came well packaged along with a set of  detailed instructions running to 22 pages. These instructions describe the basic circuit operation and the kit assembly process, and some notes on possible ways to extend the basic transceiver.

The kit instructions wee complete, but they required much more effort to follow than many other kits where each step and the location of every component is carefully described. All of those details were provided but in a more concentrated manner, and it was in the "Advanced" section of their website where builders should have relatively good skills and equipment.

The schematic and layout diagrams were hand-drawn, but they contained almost all of the details required. They were not perhaps as clear as schematics and diagrams you might see in a typical electronics magazine or, say, like those on my website. (Sorry - I couldn't resist the self-promotion there)

Most importantly, Walford's printed circuit boards (PCB) are laid out by hand, just like we used to do them back in the 1970s. That means the PCB lacks features you might reasonably expect in a modern kit, such as perfectly aligned and well-spaced components, a printed PCB overlay on the topside of the PCB to help in component placement, and a solder mask on the copper side of the PCB. All of these make the kit-builder's job a little easier. But Walford kits are not like that. So builder skills are definitely required.

Building my Lydford Transceiver Kit

Here are a set of photos with some notes which describe the various stages of my transceiver’s construction.

Being more difficult to build than some other kits, I began by reading the instructions several times. I found I noticed some extra details each time I went through them. I also used a highlighting pen to mark where band-specific components had to be fitted to reduce the chance of making mistakes.

I then carefully inspecting the PCB. Because these boards are handmade, and not produced in a PCB factory somewhere in a distant foreign land, you do have to carefully check the PCB for any errors. It’s best to look for these right at the start. They are usually easiest to fix at this stage, too.

In my case, I noticed that a number of the drilled holes were not centred, and several tracks looked doubtful. You can see some examples in the photos below. In most cases, they presented no problem. Occasionally, and I’ll describe an example from my PCB, you may need to drill another hole or two. Where I noticed some tracks that looked doubtful (See Figure 4 below - Right-click on the photo with your mouse if you want to see the full detail), I dropped a blob of solder on each location on the tracks where they looked they might be open-circuit. Checking with an ohm-meter, they appeared to be OK, but I added the solder blobs just to be sure.

PCB holes offset from component pads
Figure 3 : Several PCB holes were offset from component pads

Possible open-circuit tracks
Figure 4 : I added a little blob of solder to permanently fix several potential open-circuit tracks

I then checked all the parts against the parts list, as suggested in the kit instructions. I had one of the first or second batches of Lydford kits sold and there were a few minor differences between the parts list, the schematic and what was supplied in the kit. Nothing serious, and Tim has since updated the parts list so new builders should not encounter any problems.

I usually also make a photocopy of the circuit diagram and the layout diagram and place them on my workbench right next to where I am building a PCB. This allows me to glance from one to the other quickly as I'm building the board. I also used my highlighting pen to mark the components on the layout diagram as I fitted them on the PCB. This helped me to more quickly identify the location for subsequent parts in the absence of the printed component PCB overlay.

In this case, I found the circuit diagram was more difficult to work from that I would have liked. It flows across two pages and, in some areas, especially the switching around IC4 and IC6, the circuit's operation was something of a mystery. Since I encountered a problem later in this area which I had to fix, I ended up spending a few hours early in the project redrawing the schematic onto a single page using one of my PC-based design programs. I printed out onto an A4 page and then photocopying it up to A3 size. No problems with reading it after that!

A small part of the one-page redrawn schematic

Figure 5 : Here is a section of the Lydford schematic which I redrew on my PC. When I photocopied up to A3 size, it made making the kit much easier to build,  track down faults, and (later) modify.

The copyright over the Lydford's circuit design is held by Walford Electronics. My redrawn schematic is not available for download. Walford has refused permission to publish it even though the radio is no longer in production.
And, no, sorry, I can't email it to you either.

Here’s how the board looked after the first steps. The power/antenna connector is in place, the front end coils installed, the trimmer capacitors, along with a few miscellaneous parts.

Photo 01

A couple of new parts (IC2, C20, C21) have now been mounted on the PCB, but things still look pretty lonely on this part of the estate. You can clearly see the countersunk method to isolate the top ground plane in this PCB. Unlike modern PCBs, there is a remote chance that approach can result in a short circuit should an angled component wire touch the ground plane at the edge of this hole. It pays to keep a close eye out for this potential problem as you build the kit, especially with the angled leads of the transistors.

photo 02

Continuing to follow the assembly instructions, I then completed and tested the receiver’s speaker amplifier.

photo 03

The audio preamp parts around op-amp IC7 were then added, and a speaker and (at the lower left corner in the photo below) the volume pot were temporarily wired in place for testing it, and subsequent stages. I used a 30mm diameter 2mm thin speaker for this. I replaced it later with a more sensitive
full-sized speaker but this tiny speaker is a useful size for testing during the build phase.

photo 04

Bt the way, I much prefer this "stage by stage" approach to kit building instructions compared to the alternative used in some kits from other suppliers. Each stage can be tested, and the kit gradually comes alive, section by section. In some other kits, the builder is led through adding sets of the same value component, for example a bunch of 1k resistors all over the board, then a set of 1k5 resistor, and so forth, until every component is fitted. Then the equipment is powered up and tested. For me, that approach is quite boring. It can be problematic too. For example, faults can be much more difficult to locate. The Walford approach, stage by stage assembly and testing, is much more interesting and satisfying.

I proceeded next to install the second mixer (IC5) and switching (IC6).
It was here that I then encountered some problems with the voltage measurements described in the instructions.

After a few minutes work, I found a fault in the PCB layout which required a little bit of careful work with a scalpel to fix. You won’t encounter this problem – Tim’s fixed the artwork to better ensure your transceiver will work for you first time. In this case, you may just be able to see in the photo below where I’ve removed some unwanted tracks which lay between several of the IC pads.

photo 05

Having sorted that problem out, it was time to install the crystals for the IF filter. This filter provides most of the selectivity in the receiver. Here I found another little problem – If you look carefully at the top corner of the IC in the photo below, you’ll see the crystal case which I placed here temporarily is grounding pin 1 and pin 2 of IC6. I re-drilled the mounting holes for this crystal, offsetting them by about 1mm, and all was well.

photo 06

The crystal then fitted perfectly, with plenty of space separating it from IC6.

In the photo below, you can see that I've fitted all of the crystals for the IF filter and the second switching chip (IC4) has also been added. The real estate on the PCB is starting to look much busier.

photo 07

The first mixer (IC3) and buffer transistor (TR5) were then added. Nothing more to test just at the moment. To do that, we first need a VFO.

But first, here’s a sideways view of the centre of the board across the first mixer to the crystals making up the IF filter. Very “New York skyline” appearance with those crystals in the background, perhaps. OK, so maybe I've been breathing too many solder flux fumes.

photo 08

As the next photo (below) shows, I then built the first oscillator section of the VFO. L1, a T50-2 toroid with a bunch of turns on it, is very much the feature of this photo. My VFO is, of course, for the 40m version of the Lydford transceiver, so ideally it should tune from about 4.2 to 4.8MHz. The coil winding instructions resulted in the VFO starting close to mid-range, around 4.4MHz. However, a few more minutes of work following the kit instructions had it tuning nicely across this range. Adding the (tiny!) SMD varicap diode on the underside of the PCB added fine tuning with the aid of another potentiometer.

photo 09

I quickly
became aware of the good stability of the VFO even though everything was still very much out in the open air of my workbench. You can just make out the tuning capacitor in the photo (above) peeping out from under the board. It was just tack-soldered in place. Certainly, the VFO was only operating on 4MHz, but it was still a very good performer compared to many oscillators around that frequency that I've had on my workbench.

typical reduction driveThe main tuning capacitor in an analog VFO like this would ideally be tuned using a reduction drive. An example of a reduction drive is shown on the right. This is a modest 6:1 unit, but for my transceiver, I needed a more elaborate drive. It needed to reduce the tuning rate from 150 – 200kHz per half-turn on the "polyvaricon" variable capacitor supplied in the kit to a more reasonable turning rate of, say, 10 to 20kHz per turn of the tuning knob. The Lydford kit does come supplied with an additional ‘fine tuning’ varicap control for the VFO to avoid the need for this sort of reduction drive. However, this requires you to use a two-step tuning process, first tuning to the general area of the band with the main control, and then fine tuning onto the exact frequency with the second control. This works, but I felt the usefulness of the transceiver would be improved if I used an alternate approach.
Typical vernier dial with integral reduction drive
Lacking a suitable reduction drive, or a vernier dial with its reduction mechanism (An example of a vernier dial is shown on the right), and not having access to a full-scale mechanical workshop and suitable parts to make something better (A friction drive or a string-dial are two traditional approaches that are possible), I had planned from the outset to replace this section of the transceiver with a DDS-based VFO once everything in the original kit was working.

A DDS VFO uses a rotary encoder for tuning (You can see an example of a rotary encoder here) with the tuning speed set in software. DDS software will often support several tuning speeds, and even variable rate tuning. The DDS VFO I used is one of my own design. It has four selectable tuning rates (I don't like the alternate 'variable rate tuning' approach) and it is described in more detail here on my website.

In the meantime, and as the next photo (below) shows, I continued using the analog VFO during initial construction. The VFO signal is divided by four in the 40m version of this transceiver. This uses a 74HC74 divider which I then fitted on the PCB. It's in the lower-centre of the photo below. The pink jumper to the right of the divider selects the divide-by-4 output required for this VFO, as well as for my DDS VFO that I added later.

photo 10

The complete receiver was operating at this stage, and I connected it to a 20m long-wire antenna. This immediately brought in a multitude of signals, demonstrating the Lydford’s sensitivity. (Very happy smiles on the face of the builder, and a gentle complaint from my ever-patient wife in the next room about the loud CW and SSB signals audible from the workbench. Headphones...Where are my headphones...?)

After some enjoyable time spent playing with the receiver, it was on to the transmitter stages, beginning with the microphone preamplifier. These parts, including the microphone gain control can be seen fitted in the photo below. The 8-pin NE612 second mixer can also be seen in the lower left of this photo.

photo 11

I didn’t have a suitable low impedance dynamic microphone which turned out to be required with this microphone preamp. It wasn't mentioned on the Walford website. With no easy way to obtain one, I had to look for an alternative.

Electret microphones are incredibly commonplace these days, with almost every transceiver, recorder, wireless device and telephone using them. I had three or four electrets microphone capsules in my parts bin, as well as some ready-to-use electret microphones. Several of these were tried after I made a small circuit modification to the Lydford's microphone preamp stage. This modification can be seen in the schematic diagram below.

Electret microphone modification

Figure 6 : Schematic showing the input section of the original dynamic microphone preamp and the modification I made to use a more widely available electret microphone

The four additional components required were added under the PCB. They are small enough to fit under the board without problem when the PCB is mounted in place on the chassis. I then tested a number of different electret microphones. I'll describe that shortly.

photo 12

Figure 7 : The added components for the electret microphone. The 10k resistor at left is wired to the +8V rail, and the yellow wire trails away to the electret microphone. A 100n disc ceramic capacitor provides RF bypassing to an adjacent ground rail.

Here’s a close-up view of the microphone preamp and divider with a line of four diodes heading in a line in the background. These drop the 8V rail to 5V for the divider – A simple and effective solution that avoids the use of another regulator and several bypass capacitors.

photo 13

Meanwhile, the relays and a pair of presets, one to set the transmitter drive level and the other to set the transmitter’s power amplifier bias voltage, were added to the PCB. The relays can be seen at top-left and top-right while these two presets can be seen to the right of the crystal filter.

photo 14

As you can see, the board is starting to look quite busy, yet at no stage did it feel overwhelming. If you saw this photo at the start, it might put you off, but actually building this, little by little, is quite straight-forward. The most important thing to remember is to check the location of each part VERY carefully against the layout diagram in the instructions. Install the part, then check its location again. Only then should you solder it in place.

photo 16About this stage of the kit, one of the parts which has to be added is D70. This 1N4148 diode is fitted between IC6, the one of the CD4066 switching chips, and two of the crystals in the IF filter, X5 and X6. I added some (blue) insulating sleeving to the leads on this diode as a precaution against possible grounding to one or other crystals.

The next step was to build the next transmitter amplifier stages. The photo below shows the first stage, a single FET, which you can see tightly tucked in beside the relay to the centre-left of the photo below. This grounded gate FET stage is DC-switched and,
in transmit mode, it buffers the low-impedance transmit output from the RF filters (L2, L3 etc) produced by the mixer (IC3) and switching (IC4) through to the high impedance input of the next stage (IC100 - AD8055) of the transmitter.

photo 15

The subsequent stage, IC100 (AD8055) and a pair of transistors (TR101, 102), was then installed. You might just be able to see a very thin -looking 100k resistor (R109A) fitted between IC100 and the Drive preset in the top-centre of the photo below. The 1/4W resistor supplied in the kit was a fairly tight fit so I replaced it with a smaller 1/8W type to maintain clearances here. This new resistor is
operating well within its ratings, so no problems are expected.

Incidentally, I labeled the RF Drive preset on the board at this stage because of the identical appearance of these two adjacent presets. The other preset adjusts the PA bias current. This was to prevent any mistakes later during alignment.

I've not used the AD8055 before. It's a variable gain low noise amplifier with gain adjustable in the Lydford from unity to more than 10dB. Its frequency response at that gain setting extends beyond 15m (21MHz).

photo 17

I tested my (40m) transmitter RF output from this stage into a 1k 1/4W load temporarily soldered on the underside of the PCB where the gate and source of TR100 (IRF510) would normally be installed. Everything worked perfectly, with speech from the electret microphone producing 1.2Vpp across this temporary load resistor. That matched the value suggested in the instructions.

As I mentioned earlier, I tried several electret microphone capsules as well as the electret microphone fitted in a cheap computer headset, and I even tried an external microphone fitted in an iPhone accessory microphone/earphone. The output varied considerably between all of these different microphones. The cheap computer headset was the least sensitive, while a small microphone capsule measuring about 4mm in diameter from an old cordless phone produced the most output. Ideally, I would have redesigned and rebuilt the entire microphone preamp stage to better suit these common-as-dirt electret microphones, but I was trying to leave as much as possible unmodified at this stage. (Some months later, I tried a "dynamic microphone" purchased in the UK. That's what the label said. Well, it turned out to be an electret type!)

The PA and its towering heatsink were fitted next. I had my doubts about the need for such a large heatsink, but these thoughts were quickly banished. The heatsink does get warm during use, but not particularly hot. It also provided a significant margin of safety, protecting the PA FET from any possible harm while I aligned the transmitter. That was especially true during my lengthy tests of a variety of microphones and accessories. (This photo below makes the heatsink look like the Tower of Babel, but it's actually only about 30mm high!)

photo 18

The last step in the PCB construction was winding and fitting the three output toroids, two of which are used in the output filter. The selection and fitting of the required capacitors took a little more time, especially the time taken to carefully work out which PCB holes to use. It's fairly crowded in that section of the board.

Actually, I temporarily removed C109 (100uF) from the PCB to give me a little more space in which to work. After fitting the low pass filter components, I then refitted C109 before testing the transmitter again.

photo 19

As I've noted, my transceiver was built for use on 40m, so if your transceiver is built for another band, your toroids will have a different number of turns on them and require different capacitors, so things may look a little different here compared to my PCB.

The last test of my newly completed transmitter (into a 50 ohm dummy load) worked exactly as expected. There were no signs of instability or strange spurious outputs, and I managed to achieve about 5W out with a power supply voltage of just under 13V.

Final alignment took place later once the PCB was mounted in the chassis and wiring was completed.

Finishing Construction

Here is the transceiver board fitted into its chassis, with the wiring nearing completion. The plain unfinished aluminium chassis was purchased from Maplin UK.

photo 20

The prototyping board which can be seen mounted on the front panel is my DDS VFO. Developed specifically for this radio, its display is fitted to the front panel. The main Lydford transceiver board can be seen mounted on the right of the chassis with wiring going to the antenna and DC
connectors on the rear panel and to the front panel controls.

Don't be misled by the current showing on my power supply in the background. The Lydford's receiver only requires a modest 50mA. The remainder of the current is mostly drawn by the very hungry AD9850 DDS chip used in the DDS VFO.

Some Accessories for my Lydford QRP Transceiver

The other (brown) prototype board seen hanging out from the left hand side of the transceiver in the photo above is an AGC circuit being tested. The Lydford receiver is quite sensitive, particularly with the low noise floor of the analog VFO. As I tested the receiver and tuned across the 40m band, I found listening to the vast range of different audio output levels, some signals being ear-numbingly loud while others were relatively quiet, rapidly became tiring. This is due to the lack of AGC in the Lydford transceiver. If I was using this transceiver back home in New Zealand, the number of broadcast stations and the audio level variations I encounter when tuning might not be quite so great. But here in the Middle East, I've got large and small signals right on my doorstep. So, adding AGC quickly became essential.

To be fair, I was aware of the potential need for AGC before I started the kit, but I wanted to see for myself just how necessary it was in this transceiver and in my location. With a few megawatts of MW and SW transmitter power less than 30 minutes drive away from my house, the need for AGC quickly became obvious!

I did not have the forethought to buy the matching Walford AGC kit, and the lack of a reliable mail service prevented it being sent to me here in the Middle East. So, as usual, I ended up experimenting with a number of alternative circuits and approaches. The one pictured in the photo above was based on a design used in the BitX and IFER SSB QRP transceivers. It was later replaced with another design which produced considerably less noise. The resulting combined AGC and CW filter board is described elsewhere on my website. The basic CW filter I added to this board also made listening to the CW signals located
between the broadcasters at the lower edge of the 40m band much more enjoyable.

I also built some other accessories for the transceiver, and most are described here on my website. These include a “silent” antenna tuner, so no RF output is required from the transmitter until the very last stage of tuning the ATU (shown on top of the finished transceiver in the photo above), and an RF power meter which is mounted inside the ATU. I may also add an S-meter to the transceiver at some future date.

For more details of these added items, j
ust click on the links below.

arrowDDS VFO for the Lydford Transceiver

arrowSilent-tuning QRP Antenna Tuner (ATU) and RF Power Meter

arrowAGC and CW Filter Board for the Lydford Transceiver


Front and rear panel artwork: FrontBackPanel.gif

Top cover speaker drilling pattern:  speaker_drill_guide.gif

Want to go back to the main page? Click here to return directly.