The Minima Is An All-Band HF Transceiver For Under $100

If you have ever browsed an amateur radio magazine you could be forgiven for receiving the impression that it is a pursuit exclusively for the wealthy. Wall-to-wall adverts for very large and shiny transceivers with hefty price tags abound, and every photograph of someone’s shack seems to sport a stack of them.

Of course, this is only part of the story. Amateur radio is and always has been an astonishingly diverse interest, and away from the world of shiny adverts you’ll find a lot of much more interesting devices. A lot of radio amateurs still design and build their own equipment, and the world of homebrew radio is forever producing new ideas.

One such project came to our attention recently, the Minima, an all-band HF SSB transceiver. It’s an interesting device for several reasons, it uses readily available components, it’s an impressively simple design, and it should cost under $100 to build. This might sound a little far-fetched, were it not from the bench of [Ashhar Farhan, VU2ESE], whose similarly minimalist BITX single band SSB transceiver set a new standard for accessible SSB construction a few years ago.

The circuit shares some similarities with the tried-and-tested BITX, using bi-directional amplifier building blocks. The mixers are now FETs rather than diodes, the intermediate frequency has moved from 9MHz to 20MHz, and the local oscillator is now an Arduino-controlled clock generator. The whole thing is designed to be built dead-bug-style if necessary, and two prototypes have been built. We’d expect this design to follow a similar evolution to the BITX, with the global community of radio amateurs contributing performance modifications, and no doubt with some kit suppliers producing PCBs and kits. We think this can only be a good thing, and look forward to covering some of the results.

We’ve featured [Ashhar]’s work here at Hackaday before, when we covered a BITX build. if you’re left wondering what this amateur radio business is all about, we suggest you have a read of [Bill Meara]’s guest post on the subject.

Thanks [Seebach] for the tip.

38 thoughts on “The Minima Is An All-Band HF Transceiver For Under $100

  1. Very nice job. I’ve been working on a DC-through-HF single-band transceiver design for a while, with the major sticking point being that I don’t want to use a PLL local oscillator, and can’t come up with anything else that would be stable enough to be usable. I’ll probably end up breaking down and doing it the way Ashhar did.

    The one thing I don’t understand is the claim about “automatic switching” of sidebands. How the heck does that work?

      1. Part of the goal here is an all-analog ultra-low-power radio, or at least all analog in the signal path (I will probably use an Atmel-based frequency counter for accurate tuning indication, but will turn it off once tuned up). So direct-synthesis has the same down-side as PLL.

    1. Build it but make the “VFO” external. That way you can try different things, including a crystal oscillator or single band vfo so you can use it while figuring out the best scheme.

      One reason double conversion like in Collins equipment was that it kept the tuneable oscillator frequency low, and no switching. Synthesizers later replaced the bank of crystals, though the PLL reference frequency was only 500KHz or so (hence easy to get a clean signal.

      Swan went with an HF range IF, and switched the VFO tuned circuits to change bands. Not so stable.

      Some, I think including Drake, used an HF IF, but a fixed VFO. So the VFO was mixed with a crystal oscillator before getting to the signal mixer. So you had to carefully filter the output of the “premixer”.

      A PLL can do that filtering. The VCO on the needed frequency, but inside the loop a mixer and crystal oscillator to bring the VCO frequency down to the reference frequency. The reference frequency was the fixed range VFO. One crystal per band was still needed.

      A PLL with steps every fraction of a KHz came later to amateur radio.


      1. Thanks, Michael, that’s probably a good idea. I say ‘probably’ because decisions like how wide the VFO tunes affect the frequency plan for the radio. I’ve studied the architecture of a number of Collins, Drake, and other transceivers. One common trait of the ones known for rock-solid stability is that they use low-frequency VFOs, but the tradeoff for that is that you get a very narrow tuning range. The Collins R390A, for example, can only continuously tune over a 1 MHz range. Which is fine for amateur HF work, since the 10 meter band is the only one wider than 1 MHz, but it’s a royal pain for just scanning shortwave bands, since you have to spin the VFO dial all the way back to go from 9.99 to 10.01 MHz. But I’m toying with the idea of using a narrow VFO with a control that switches direction at each band change, so when for example you’re tuning from the top of the 9-10 MHz band to the bottom of the 10-11 MHz band, you switch bands and start scanning the other direction. Yeah, I don’t know about that, either.

          1. Mostly because of a touch of OCD. But then, that would probably screw me up because half the time it’s tuning the “wrong way”. :) Anyway, that was why I wanted to tune using bands as wide as possible, like six five MHz bands from DC to 30 MHz. (I’m an old spectrum analyzer guy, so I also have an aversion to radios that won’t tune down to DC.) But once again, wider VFOs means less stability.

    1. I have never built a radio before, but I have built Manhattan-style xtal oscillators, wound toroid inductors, a rat’s nest mint tin filter, and programmed microcontrollers before. Would this be a good first radio to build?

      1. There are a couple of reasons this might not be a first full radio build:
        1) Getting a radio to tune over this wide a range poses a number of challenges, particularly with things like getting a clean enough transmit signal (since as I_failed_C++ says below, the first few harmonics are within the pass band for the low-pass filter). A single-band transceiver is much easier to get right.
        2) This radio, along with the BITX radios, uses a bidirectional IF amplifier that allows the same mixers and crystal filter to be used with the signal flow from microphone to antenna on transmit, or antenna to speaker on receive. Note that Q6-Q8 are the transmit amp and Q3-5 are the receive amp. It’s a very clever design, but I’ve seen a number of people’s implementations of BITX radios where they’ve had to add other circuitry to isolate the input from the output (or is it the other way around?) in order to fix oscillation problems. So even though this cuts way down on the parts count, it does present additional challenges.

  2. You don’t have to be wealthy to be in ham radio. One of the nice aspect of ham radio is most equipment doesn’t become outdated. The market for used equipment is quite robust. For a couple hundred bucks you can get a commercial rig with lots of bells and whistles that is perfectly adequate. Or you can get simpler new rigs for $50 that are limited in capability but still get you nicely on the air. The other nice thing is market for used equipment is quite robust. That rig you bought for a couple hundred will sell or trade-in for not much less than you paid. When you upgrade you can get most of your money back toward that newer rig.

    Don’t be fooled by the really expensive rigs. You’re not going to know what to do with all the bells and whistles anyway. As always it is the basics you need to learn before getting really fancy. You can get that fine with an nice, inexpensive shack. Biggest tip: keep it clean, don’t contaminate it with cigarette smoke (in the US at least), and keep the manual and box if you got them with the equipment.

    1. Okay, that’s probably the oddest qualifier I’ve seen all week. Does cigarette smoke do less damage in other parts of the world? Or is it just hopeless to try to be smoke-free anywhere else?

      1. I suspect the cigarette smoke caution referred to buyers of used equipment. Buyers in the US are much less likely to pay full price for a cigarette smoke contaminated radio then perhaps ham radio operators in other parts of the world because without a doubt you can find a similar not contaminated rig just down the block .

      2. If I buy something and it smells like cigarettes when I receive it, yet the ad made no mention of it, you bet it’s going back, and appropriate feedback will be left. Selling stinky equipment to unsuspecting buyers is not acceptable.

    2. “For a couple hundred bucks you can get a commercial rig with lots of bells and whistles that is perfectly adequate.”

      I read this sentiment several times, but I haven’t seen many $200 all-band rigs that weren’t QRP, broken, missing something major. Even if I look at QRP, I’m not seeing much.

        1. I’ve seen Kenwood TS-140s for 2-300 and they’re good radios, I have one myself that I inherited from my grandfather when he passed away. Downside is it’s all transistor so it’s not as forgiving as a radio with a tube final such as a Kenwood TS-830 (Those can often be found for 3-400 as well), tubes are a lot more forgiving of SWR mismatches and I’ve had a few hams recommend them to n00bz, especially if you’re building your own antenna and don’t have the $ for an automatic antenna tuner.

        1. Indeed – why, you can buy a chintzy PC for that price. But again, you can buy a low power code transceiver kit for under US$100. I picked up a really clean working Icom IC-745 HF all-mode transceiver for $215 on eBay. Most ham clubs will have a number of members who have spare “rigs” (transceivers) that they’ll lend to beginners to get them on the air. If you don’t have a license and are just curious, the weekend of June 24th – 26th is Field Day when many hams and ham clubs get on the air in remote or public places. That includes “Get On The Air” activities where anyone can try ham radio out and make contacts on the air without a license as a control operator will show you how, making it entirely legal and free. How can you beat that?

          Yes, there are multi-thousand dollar radios, antennas, towers and so forth. Advertisers emphasize where the money is. But as a hobby, it’s still less expensive than motorcycling, power boating, RVing, high performance computing, many 3D printers, a full woodworking or metalworking shop, or any number of other fields HaD often covers. Hooray for the freedom of choice – and the ability of HaD folks to overcome obstacles!

        2. When you start out to get the ticket, you already probably have a few old timers around that want you to get on the air as fast as you can get the ticket. Many of these good old boys and xyl’s help a whole lot, as well. There is always the local show and sale…get with a club to get your ticket and you make new friends as well, the old timers are the ones teaching the classes…the $$$ really won’t keep you off the air as long as you might think. Wait a few more years and you will have a old Halicrafter sitting on the shelf looking for a new home, because the latest and greatest new radio you want to buy has to have a place to sit….ooops …there goes that old halicrafter out the door to a n00b that did not have $300. What goes around, comes around…… Enjoy the trip!

    3. Any properly working transmitter produces the same signal as another. SSB is SSB, FM is FM, etc… afterall. Assuming you are getting something that has or would pass FCC (or your country’s equivalent) regulations anyway. The one functional difference is power level but that really isn’t as important as people think it is.

      Remember, as your signal gets farther away from you it is also spreading out in 3 dimensions. Signal strength vs distance drops as an inverse square. What really makes the difference is the antenna. That is what allows you to focus your signal so that it doesn’t drop as quickly. Fortunately one can easily build a pretty good antenna with just a few dollars worth of wire!

      There may be something to be gained with a fancier receiver. It will introduce less noise of it’s own and have better features for filtering out interference. That means you can more easily hear a weak signal in a crowd of other hams or when natural static and/or man made electrical interference is high. That’s still just getting you a little bit further in to the weakest signals. Increased money spent there will be subject to diminishing returns. Again it’s the antenna that is key.

      1. And this is what makes for good radio, even if you never see a color ad for hanging a wire dipole in a tree it will do more than a tuned $80 ‘whateverstick’. Even in a hotel the time spent bluetacking an aimed loop is paid off by not investing in some mini wonder antenna.

          1. In the US, radios made by licensed amateur operators do not need a TYPE certification, but they still need to meet FCC regulations. Specifically, in the HF band, spurious emissions cannot exceed 43 dB below the carrier. I think that’s what I_Failed_C++ is referring to.

        1. Not only that, but the local oscillator is pushing a SQUARE WAVE into the mixer, which makes for a very low loss mixer, but it also means very high 3rd and 5th harmonics. Specifically, the 3rd harmonic would only be about 6 dB below the fundamental. And the Si570 isn’t specified as highly symmetrical, so yep, plenty of even harmonics as well. And in addition to that, the same would apply to the received signal, so if you tune to the 80 m band, you’ll be hearing 40 and 20 meters as well. Good catch.

          1. Yup, the draft of the new design is not going to cut it from the RF filtering point of view. The mixer needs a tap on the input transformer secondary for balancing. The Si570 asymmetry is cured by using a high speed flip-flop to divide by two. Overall RX gain may be OK on 40 & 80 meters, but I doubt it is going to be high enough on the higher frequency bands. I would suggest adding a a switchable low-noise preamp to the RX front end. I’m not a fan of the bi-directional amplifiers at all, I call them the product of a misguided obsession with minimalism. I always split them up. That IF amp before the crystal filter would be a lot better if it was termination insensitive. In-fact since the filter crystals are a 24 MHz (high for this type of filter), I would book-end the filter with step-up/down transformers. That would not only allow good interstage matching, it would improve the Q factor.

            It’s like I said: a misguided obsession with minimalism.

          2. First thing I would do is replace the Si570 with the Analog Devices AD9834 which outputs either sine or triangular wave output. This would cut down on most of the harmonics created by SI570 square wave out. Alternatively you can run the SI570 at 2f divide by two with a flop to get 50% duty cycle and drive it into a class E circuit to make a fairly clean sinewave (but this is starting to sound like a rube goldberg contraption).

  3. I am a bit puzzled by the KISS Mixer as there seems to be no path to Ground for the current through the FETs. They are biassed via the 4k7 R and the local oscillator is being fed in to the gates, but there is no path to ground through T1.

    1. Yeah, it looks like he left something out, there. In the article BobMounger linked above (, Farhan says he had trouble with the KISS mixer, and switched back to a diode ring mixer. This could be why he had trouble. The problem here is that he’s using N-channel FETs, which means the sources need to go negative with respect to the gates and the drains. This can be fixed by feeding the 78L05 and pot into the output of the mixer through an RFC, then a resistor to ground on the sources. The RFC prevents the bias network from affecting the impedance match for the crystal filter. But I haven’t done a full DC analysis of the circuit, so don’t just take my word for it.

      1. Farhan’s basic problem with the mixer was that the oscillator was getting through the mixer and showing up in the output. IIRC there was a mistake in the diagram (several actually) and I think that was one of them. The receiver was very good but stuff was coming out of the transmitter that would make it difficult to sell in a kit form. Seems in the US the FCC doesn’t really like someone selling “dirty” rigs, or so Farman thinks. Recently he said that putting a band pass filter in the output would eliminate the problem

  4. If you like the Minima then another one to look at is the LBS (Let’s Build Something). This one was published as an article in a magazine, ‘QRP Quarterly’ so you will probably have to get ahold of the back issues. Unfortunately I wasn’t a subscriber :-( when I do get time to start building my radio I will probably try to get those issues.

    Anyway… the LBS I think it probably similar to the Minima. It’s another radio using a DDS chip controlled by an Arduino. What I think is interesting about the LBS is that it is intended to be built in stages. It’s always best to build these things in modules as it makes them easier to test. The LBS is planned to be built in such a way that those modules start out as a direct conversion receiver and then parts are added, adding functionality until you end up with a fully functuional transciever. This way even in a only partially completed state you have something to play with and something tangible to show for your effort!

    Especially as a first homebrew rig I would think that that is a pretty cool thing.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.