SP-8: a homebrew 28MHz SSB transceiver for a UR3LMZ 144MHz transverter

Ten meters or the 28MHz band is showing glimmers of life from sunspot cycle 25. Even so, its not an obvious choice if you want to build a portable SSB monobander and have lots of contacts. But 28MHz is the IF of choice for VHF and UHF transverters. After noticing the 6, 2 and 70cm transverters available on eBay from the workshop of UR3LMZ and the good reports from buyers, I cooked up the idea to build a 28MHz ‘transverter IF’ transceiver, to be paired with one or more VHF or UHF transverters.

The result is a homebrew single conversion superhet SSB monobander on 10 meters with a UR3LMZ two meter transverter in the case. Bringing the 28MHz receiver and transmit drive signal out to connectors on the rear panel makes this a tunable IF for an external 6m or 70cm transverter down the track.

Homebrew 28MHz SSB transceiver for a UR3LMZ 144MHz transverter (Summit Prowler 8) by VK3HN

Concept

The objective of this project was to get a portable, compact 2m SSB transceiver for SOTA and portable use, something with sufficiently good performance to match some of the older 2m SSB superhet rigs. I toyed with the idea of homebrewing a dedicated 2m SSB transceiver, probably achievable with an si5351 as the VFO and a conventional IF. The 144MHz stages would need special care.

However I decided that buying the VHF module and installing it in front of a monoband HF transceiver would be much more achievable and more versatile. I’ve never built a transceiver for any higher frequency than 17m so getting good gain and overall performance on 28MHz was a challenge in itself.

Design and schematic

The design of the transceiver is conventional — a BF961 dual gate MOSFET RF amp with bandpass filtering front and back, a Minicircuits L7 diode ring mixer (JMS-1), post mixer Class A amp (2N3866), three stage IF amplifier using pairs of BF246 JFETs, SA612 Gilbert Cell product detector, audio preamp and a ULM2002 power amp.

On the transmit side, the mic amp, balanced modulator and transmit mixer use 2N5245, TL072, LM1496, LM1496 respectively. At 28MHz a 2N3904 predriver and BFU590GX driver excite a Mitsubishi RD16HHF1 for around 5 watts. More than the usual number of bandpass elements were added (easier to achieve in a monobander) to give the 28MHz signal a higher degree of spectral purity.

CW was omitted in this rig on the hunch that in VK what limited 10m/VHF/UHF activity to be found would be telephony. That eliminated some of the sockets and circuitry normally required by this mode and simplified the build.

The full kicad schematics are here.

Case

The case is an aluminum angle and sheet creation. This one uses 32 x 32 x 1.5mm angle for the front, rear and sides, with a snug fitting 2.5mm sheet base. The top is 1.25mm sheet, cut at the supplier.

Small brackets are added around the top of the front, rear and sides, dropped by 1mm to allow a sheet aluminum top to drop in.

After the aluminum angle piece is rubbed down with 0000 grade steel wool and cleaned with bathroom cream cleanser, DecaDry lettering is applied; the piece is then sprayed with three light coats of a satin clear enamel, a very light rub with 800 grit between coats to take away any rough points.
Completed rear panel.

VFO, BFO, Controller

VFO/BFO/Controller double sided board supports front panel controls on the front.
Arduino Nano and si5351 breakout board in sockets (and 5v regulator) on the component side. Note the breakout board was trimmed to make it fit vertically inside the narrow chassis!

I built yet another Arduino Nano and si5351 VFO, BFO and Controller onto a narrow 28mm wide PCB. It had to be this narrow to fit vertically into the 32mm high case. After discovering that the si5351 breakout board was 2mm too high I took to its right side with a miniature grinder, turning 2mm of it’s edge into dust, while watching for any evidence of a copper trace. Fortunately none were found.

IF module

This module consists of a Yaesu XF-8.9HS crystal filter, found on a Hamfest table a few years back. It is from the FT-901, is on the unusual frequency of 8987.5kHz, and is 2.4kHz wide. The three stages of tuned amplification after the crystal filter are taken from EI9GQs book and use a pair of JFETs in a cascode arrangement. This module showed signs of instability, so brass plate shield was installed which made some difference. However to really get it to settle down it was necessary to add a 5k6 damping resistor across the tuned circuits of the first two stages.

Receiver front end

Nothing special here, a BF961 RF stage, JMS-1 mixer, balanced tee diplexer (see below) and 2N3866 post mixer amp. A MMBT2222 gets the VFO level well above that required by the L7 mixer (a pluggable 3dB pad sets the level on mixer VFO input).

The receiver uses a 4-pole bandpass filter on the front end (EI9GQ) and a two pole one on the output of the RF amp. This is probably excessive filtering but I chose it to get a clean signal to the mixer. The diplexor is built on a small board off to the side. The receiver front end module is quite stable, despite the fact that the signal path snakes around from one side to the other, navigating shielded compartments, slightly violating the golden rule of keeping outputs well away from inputs.

8.98MHz balanced tee diplexor sweep, around -3.8dB at the intermediate frequency, the -6dB bandwidth is fairly wide at 3.2MHz; at the 37MHz VFO around -30dB.

SSB exciter

The exciter is a familiar sequence of microphone amplifier, balanced modulator, then off to the crystal filter and IF amplifier. Upon return the 9MHz SSB signal is mixed with the VFO to 28MHz using another LM1496 Gilbert Cell mixer to get to 28MHz.

28MHz driver and PA

The output device is a Mitsubishi RD16HHF1. This FET has flat gain to 50MHz and the convenience of a source (not drain) tab making it easier to mount and heatsink. From the LM1496 transmit mixer there are two gain stages, an untuned 2N3904 common emitter stage followed by 28MHz bandpass filter to a BFU590GX driver, to a broadband collector transformer and followed by a low pass filter. Output power is 3.5 watts with a bias of 150mA. Being a monoband exciter, it is easy to build in bandpass and low pass filters in place on the boards.

144MHz transverter

The UR3LMZ 2m transverter was ordered mid lockdown via eBay and took around 10 weeks to arrive due to backed up freight around the world — no fault of the seller. Once it arrived in a small but exotic world-weary looking parcel with Ukrainian stamps and travel scuff marks, I was impressed by the quality and construction of the unit.

The transverter is conservatively designed and uses a Minicircuits ADE mixer, BF series MOSFETs and solid state T/R switching. I fired it up on the bench with care and heard the local beacons solidly. So far so good.

On transmit, the required drive is just 50mW, so I was fastidious about working out how to drive the thing. If you are using an FT817 you have no choice but to dissipate your 5 watts of RF power and sniff a bit off as drive. But when you build the exciter from scratch you can take the drive from wherever you want. So I set up a relay switching arrangement to route drive from the BFU590GX stage, via a pi attenuator, to the transverter.

On the bench, with the ability to vary drive, and with judicious access to the transverter’s onboard RF drive and PA bias trimpots, I had full control, and quickly saw 10 watts of 2m SSB output. A YouTube video of this transverter on a test bench with some serious test gear indicated that spectral purity is considerably better if it is throttled back to around 3 to 5 watts RF output.

10m/2m switching

Band switching turned out to be a challenge in routing the low level exciter drive and the outputs from the two separate PAs around. This was achieved using miniature telecom relays.

Performance

On 28MHz the single conversion superhet worked very well, delivering a sensitive receiver and a clean 4 to 5 watts SSB. A 28MHz quarter-wave vertical was erected on the house roof as a resonant 10m antenna. This allowed reliable reception of a local VK3 beacon for receiver testing. Although the 10m band is usually empty here in Melbourne, digital modes can be heard down the bottom end, and more recently, a mid afternoon sporadic E opening brought in VK4s peaking at 5 by 7. Ten meters is an amazing band when it opens.

The UR3LMZ two meter transverter also worked well on the bench. Once mounted snugly in the SP-8 case, some RF feedback became apparent. Three gain controls had to be judiciously adjusted, the microphone amp gain, and on the transverter board, the 28MHz drive level and the RD16HHF1 FET bias. After extensive monitoring using a digital power meter, oscilloscope and a local 2m receiver, an acceptable settings were found. The RF power on 2m needed to be wound back to around 5 watts to keep things under control.

On air

Before going portrable with SP-8 a 2m antenna was required. I started with a homemade J-pole on a 6m squid pole. This got me on air from a nearby hill top, about a dozen contacts were made with stations all over Melbourne (most with high gain antennas), but the cross polarization penalty had to be addressed.

In QSO with Rob VK3MQ and Chris VK3KIH one Sunday morning, Rob mentioned the Hentenna, a wire loop with horizontal polarization. I knocked one up and immediately got reports from Melbourne stations around two S-points better.

After a few months, David VK3KR shared plans for his 4-element DK7ZB collapsible yagi. It is constructed from 20mm PVC drilled for 6mm aluminium elements. The driven element is configured as a dipole and is quite cunning. The inside dimension of 6mm aluminum tube can be tapped for an M5 bolt. The PVC pipe is drilled for a pair of M5 bolts, retained by one nut. Each driven element then twists easily onto the protuding M5 bolt’s shaft. A picture illustrates this much better than words. A coax matching transformer is added, wound conveniently around the PVC pipe. I’ll post on this sometime soon. Now with 6dBi gain antenna, I was a 2m QRP portable Big Gun!

Wrap-up

SP8 was a most enjoyable lock-down project. Use of the UR3LMZ transverter was a good choice. It’s performance is good on both receive and transmit. Special care must be taken to limit RF drive to a low level. Also, others results and my experience confirmed that the transmitted SSB is cleaner at 5 watts than 10. The lost 3dB is more then compensated for by a lightweight 3 or 4 element yagi.

One interesting thing I wasn’t used to was a few hundred hertz of receiver drift in the first few minutes after switch-on. This is the 116MHz overtone oscillator in the transverter. I’m not used to drift after adopting si5351 clocks as VFO and BFO. In some images you may notice the display sometimes shows 100Hz low (For example, 144,099.900).

Most of my portable operation has been on two meters with this rig. Being able to carry a 4 element beam, rig, small pole and battery in a light bike pack is fantastic and brings local hill-top operations into range. Once up on a hill, elevation and line of sight is everything on 144MHz. Two meters SSB has turned out to be an excellent band for postable QRP operation. Signals can travel hundreds of kilometers and are usually strong at both ends. And the portable QRP operator benefits from the high gain DX-catching antennas of the serious 2m base stations. just make sure you are horizontally polarised. Two meter portable operation is highly recommended!

Tagged , ,

19 thoughts on “SP-8: a homebrew 28MHz SSB transceiver for a UR3LMZ 144MHz transverter

  1. jackhaefner says:

    Excellent work, Paul!!

    I had only been participating in the Sept and June VHF/UHF contest using my FM rig in my Jeep. This was my only solution after frying my IC-7000 a few years ago. You now gave me an idea of adapting a transverter for 144 SSB/USB using my Elecraft KX3…..

    Cheers,

    Jack
    NG2E

    Like

    • Paul Taylor says:

      One of these 2m transverters will get you back on 2m easily enough. The instruction sheet (minimal) shows a 5 watt dummy load and T/R switching for a rig like a KX3 or similar.

      I just remembered that Glenn VK3YY did just this:

      https://vk3yy.wordpress.com/2019/02/07/144-mhz-for-the-kx2-transceiver/

      The KX2 even adapts it’s frequency display to show 144MHz.

      Good luck!

      Like

      • jackhaefner says:

        FB, Paul! I’ll let you know how it goes.

        BTW, making progress since we last talked. I’m in contact with Bill (N2CQR) and Pete (N6QW). Building a Michigan Mighty Mite and testing…. Adding filters and testing…. Doing my best to understand what is occurring in the circuit and why. Lots to learn!

        73,
        Jack
        NG2E

        Like

      • Paul Taylor says:

        Excellent! Those guys are world’s best homebrew mentors! I love Soldersmoke. And all of their projects. Good luck with the MMM.

        Like

  2. Bill Meara says:

    Paul: This video was especially cool because of the many different sources of circuitry: W7ZOI, SSDRA, BITX, W3NQN, all coming together in VK3… This is truly an “International Brotherhood of Electronic Wizards” rig.

    Jack NG2E got his Michigan Mighty Mite transmitter working yesterday. Great fun!

    73 Bill

    Like

    • Paul Taylor says:

      Hi Bill

      Great to hear from you. Thanks for commenting.

      I continue to pull stages and modules for my piecemeal projects from SSDRA and EMRFD, I also often check other W7ZOI designs — they are the benchmark. For example, I recently made up a 4MHz crystal BFO on its own little board with a FET source follower and a bipolar gain stage to bring it up to +7dBm from one of Wes’ designs. Powered it up, and first time, it delivered +7dBm into 50 ohms. Brilliant!
      Just like Wes said. The design just worked.

      I have started annotating my kicad schematics with a small comment field containing the page and figure number of the circuit origin, for traceability back to the source. This is good practice fo maintenance and also serves to give acknowledgment where it is due.

      Jack contacted me a month back, asking for a few suggestions to get started with homebrewing. When I realised he was in Virginia I suggested he contact you. Great that he has had MMM success.

      73 Paul VK3HN.

      Like

  3. max373849 says:

    That looks really nice. Would it be possible to build a receiver with that si5351 module für SAQ 17.2kHz or other VLF?

    Like

    • Paul Taylor says:

      Yes, relatively easily in fact. A VLF receiver for Grimeton and 137kHz,amateur bans would be an interesting design challenge. The si5351 would be a fine oscillator for an up-converter to, maybe, 455kHz. Then a narrow filter, maybe a ceramic resonator, or just a few IF transformers, and a product detector for audio. There are probably circuits on the Interwebs for something like this. An alternative simpler option is to make just the front end part and up-convert to a ham band, say 160m.

      Like

      • max373849 says:

        That sounds interesting. I asked an other OM earlier and he wasn’t sure if it’s possible. I would prefere a independent receiver. I know the up-converters to a ham band but it would be much nicer to have just a simple compact device for showing around for friends and familiy when SAQ is transmitting without the need of an additional TRX

        Like

      • Paul Taylor says:

        Here’s a good design and discussion from an experimenter who knows their stuff. This is an up-converter to HF. For a self contained receiver you would need to build a back end, that is, product detector, BFO, and audio stages.

        https://www.giangrandi.org/electronics/lwupconv/lwupconv.shtml

        I am re-thinking the wisdom of using a si5351 PLL. Read the article and you will see why.

        73 Paul VK3HN.

        Like

  4. max373849 says:

    Thanks I will have a look. Yes I know that you need additional stages like the one you mentioned, it would be a nice try for myself as a new OM and an opportunity to learn. I only know some of the theory and never built a receiver or TRX (besides of a kit).

    Like

    • Paul Taylor says:

      Best of luck wirh your project, a receiver like this is a good option!

      Like

      • Max says:

        Thanks, well I’m really at the beginning with that kind of circuits. Your link looks great and I ordered a Sbl-1 already.
        What would be your concerns replacing the osc with the si5351?
        73, max

        Liked by 1 person

      • Paul Taylor says:

        Phase noise. As 17khz is so close to the whole megahertz oscillator, you don’t want noise for even 20kHz around the oscillator frequency. The article explains it. You could try a PLL oscillator like the si5351. Then if you are not happy, replace it with a crystal oscillator. All good fun!

        Like

  5. Tony, G4CIZ says:

    Very interesting. I noticed those pairs of JFETS in the IF stages. I wonder if they are a substitute for a dual gate MOSFET? If so, I wonder if you’ve spotted the enhancement mode MOSFETS from Toshiba, eg 3SK293 and 394. They are surface mount types, but might help save space if they will do the same job as a pair of JFETS.

    Thank you for some inspiring projects. Perhaps I should re-make my old 4metre transceiver in smaller, low current consumption form. Then I might be tempted to take it to a nearby summit.

    Like

    • Paul Taylor says:

      Hi Tony, Thanks for commenting. I read your QRZ page, I like that you built an SSB transceiver core and have used it and reused it over so many years. I have a few old rigs like that, the analogue transceiver blocks dont change much, although the RF power and digital controller and VFO PLL or DDS options have improved dramatically.

      Yes the J310 pairs are a substitute for a dual gate MOSFET, which are difficult to source now. The circuit cones from EI9GQs book. Three stages have too much gain so you have to damp down the drain tuned circuits.

      I have not seen 3SK293/4, Thanks, that’s a hot tip! I see they are available thru RS, who I use for semiconductors, and 68c each! Brilliant!

      If you can, its a good time to start a new transceiver build, particularly if you are happy to use the larger surface mount 1210, 1208 parts. When I adopted surface mount, the size and weight of my projects halved. Perfect for portable rigs. Building for a single band greatly simplifies the project. With a digital PLL or DDS you can make your filters plug in, to allow shifting your selected band.

      4m is an interesting band, we don’t have it here, I listen sometimes on Hack Green VHF, I have heard SOTA activations on 2m SSB and FM, but not on 4m yet.

      73 Paul VK3HN.

      Like

    • Paul Taylor says:

      Here’s a good summary of dual gate MOSFET options and it links to a Hackaday post, also worth reading.
      https://groups.io/g/qrptech/message/1913?p=,,,20,0,0,0::created,0,3sk293,20,2,0,76455105

      Like

      • Tony, G4CIZ says:

        Thanks for the dual gate MOSFET link. Very interesting. I had read the “component that won’t die” before, and I agree. I came across the Toshiba dual gate MOSFETS while building a 6 metre preamp. I do wonder how long they will last as current parts though. The broadcast receiver market was what these devices were aimed at, huge volumes. They have moved on to more integrated front ends, so all these interesting parts from our point of view will slowly die out. I guess that’s life.

        Thanks also for your nice comments on my old SSB transceiver. It still works fine, but a nice DSP noise reduction/ variable filter add on would help it deal with all the noise on the bands these days, plus the usual QRM.

        Like

      • Paul Taylor says:

        All good Tony. You can always ‘simulate” a dual gate MOSFET with a bipolar and JFET pair as per the W7ZOI Hybrid Cascode IF. I’ve used that arrangement and it seemed indistinguishable from a BF960 or 998. I will try to remember to get some of those 3SK devices in my next RS order. Back when you built your SSB transceiver the difference between what you could make at home and the commercial products was not always large. Thinking about the Elecraft K2, or Yaesu FT7. Not so any more. But you can still build more than acceptable performance with older mostly analogue designs. If you make them compact and portable, the gap is narrower. 73 Paul VK3HN.

        Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Shack of VK2KMI

Ham Radio Blog of VK2KMI

Ham Historian

A passionate amateur looks at Melbourne history

R.IT

My journey of repairing and recycling anything I put my hands on that I believe is still useful. Not just hardware, but including software with relevant content and issues in the field of Cyber Security, Vulnerability Scanning and Penetration Testing.

G4YDM Ham-Radio and SWL news

Radio craft, homebrew, QRP/SOTA, AM

VK3HN

Radio craft, homebrew, QRP/SOTA, AM

VK1DA.blog

Amateur radio experiences with VK1DA

Tj Lab

信号処理とラジオ

WA2MZE

Amateur Radio and unrelated technologies

Paul Gacek

W6PNG/M0SNA

%d bloggers like this: