A Complete Transceiver
It works well!
As Project Yamhill nears actual physical implementation, it’s time to start building some more real-time, two-way communication. We’re starting to get some beta testing and discussions going! If you’d like to participate in Project Yamhill chat, please click on the link below. Thank you!
Well, it’s been a pretty consequential last few weeks here. As I mentioned in some previous posts, although I thought I had secured a new job that would tide me over for now after losing the previous one, that turned out to be fool’s gold. The need to secure some steady income became the most important factor here, and I could no longer count on dodgy contract work. I looked at all my employment options, and to make a long story short, I ended up applying for and getting hired for a full-time job at a local, reputable company doing what I went to school for in electronics! I’m grateful and relieved that I was able to get a good job so relatively quickly, especially at my age, which doesn’t always lend itself to the best prospects.
However, there’s no doubt that this is going to have an impact on what I’m doing with this publication, Project Yamhill, and Etherkit. I’m going to be working 40 hours a week, at a place of employment that is a bit of a drive from here, so unfortunately that’s going to eat into the time that I have for these other projects. I do intend to continue on with Project Yamhill and definitely want to see through the design of this latest transceiver. I need to wait until I start my new job and get settled into the new routine before I make any further declarations, but I feel very strongly about finishing what I started, even if it takes me more time than I initially projected.
In other words, stay tuned, as I will be producing more content here, even if it may take me a bit more time between posts. With that out of the way, onto the main part of this post.
I’m pleased to report that the last little bit of the MAX2681 transceiver that needed to be completed, the transmit power amplifier, has been built and is working well! The new PA can put out well north of the QRP full gallon on 20 meters without breaking a sweat. In fact, when I first brought it up, it was making just a bit shy of 8 watts, so I ended up dialing back the PA MOSFET gate bias to bring it down to more like 6 watts for testing.
The design I settled on is a standard Class D power amplifier, which fits well with the square-ish wave signal that is generated by the Si5351A. Instead of going with the old classic IRF510, I pulled a pack of MOSFETs out of my junkbox that I’ve been wanting to try for this purpose for a long time: the FDS86106. In many ways, its specifications are similar to the IRF510. However, it’s different in one big way: the package. Instead of a TO-220 power transistor package, it comes in SOIC-8. One pin is the gate, three are allocated to the source, and four to the drain.
I was wondering how well I could deal with the thermals of this package. The thermal resistance (RθJC) is listed at 2.5°C/W, which seems quite good to me, and beats that of the IRF510 by a significant percentage, as its RθJC is listed as 3.5°C/W. It was my hope that I could get away with a modest amount of heat sinking, so I ordered some 9 mm2 aluminum heat sinks with a layer of thermal adhesive on the interface that could just be plopped down on the IC.
In a result that was somewhat of a surprise to me, the device did very well in keeping cool under standard operating conditions. When connected to a 50 Ω dummy load, I could key down for a minute straight at full power and the heat sink would only get a little warm. You could easily touch it with a finger and not even be uncomfortable. Under a 25 Ω load, the device definitely got hot after a 1 minute key down, but I didn’t see any sign of thermal runaway to that point. Operating under an open for the load, the device also had no issues with warmth. A very promising device, I would say. More testing needs to be done later, but I don’t want to burn out transistors on this prototype if I don’t have to, as it’s kind of fragile as it is.
With a fully-built transceiver now on hand, the last major thing to do is to finish writing enough firmware to enable the rest of the transceiver functionality. The first order of business in that regard was to get the keyer logic up and running. As was always intended, I started with the firmware from Etherkeyer Mini and ported it over so that it would be compatible with the transceiver circuitry of this radio (meaning sequencing the activation of the mute line, key line, transmit LED, sidetone, and switching of the output of the Si5351A).
Without getting too much into the boring details, it took quite a bit of fiddling with the task scheduler library to ensure that the keying timing was accurate. After some test QSOs where I inflicted some buggy keying on unsuspecting victims (including AA7EE…thanks Dave!), I managed to iron out the bugs and was able to operate the radio without sounding like a total lid.
Which brings us to this weekend. It turns out to be the weekend of the CQ Worldwide DX CW contest, which is a perfect opportunity to test the receiver under the most grueling conditions it would face. Below is an audio recording I took this morning, tuning around the 20 meter CW sub-band. I haven’t yet made any quantitative measurements of the IMD behavior of the receiver, but simply going on a qualitative analysis, I think the receiver held up pretty well for what is intended to be a field radio and not a contest radio. I didn’t hear any of the “crunching” that you’ll get from bad intermod, and I only noticed a moderate amount of de-sense from nearby strong signals. Listen to this clip and see what you think:
I wanted to make a few QSOs today as well, but didn’t want to try to make contest QSOs, because my speed isn’t that great, and I was only running 6 watts. Instead I found that there were a few POTA activators hanging out in the refuge above 14.100 MHz, so I went hunting. KK6QZM in NV was easy to snag from here, as you’ll hear below. There’s a scratchy artifact in this audio when the receiver unmutes after transmit, but that’s a result of a cold solder joint on the AF amp, and not a flaw in the receiver design. This problem is fixed by the time I made the next QSO, recorded below this one.
A little later, I saw KF6SWU in CA on the spotting page and decided to give him a try as well. The QSB wiped him out briefly, but after that we were able to make a solid QSO, even if he was pretty weak on my end.
Overall, I’m incredibly happy with the state of this transceiver. I think the audio sounds quite nice and the transmitter so far is quite stable. I definitely want to make this into a real radio, one that I can take out on portable ops and that I can hopefully sell to others.
What’s Next
I started this rig with a microcontroller with 32 kB of flash memory, thinking that would be enough for all that I wanted to do, but that turned out to be incorrect. I’m already nearly at the limit, and I still have some more functionality that I want to implement. It also turns out that I need more GPIO pins than I currently have, and I could add a GPIO extender IC, but given that I need more program space anyway, I really just need to change microcontrollers. So I’ve ordered some AVR128DA32 with 128 kB of flash space, and quite a few more pins, which should meet any future needs on this radio. Once those come in, I’ll probably retrofit the current prototype board to adapt the new microcontroller, however…I also want to take this fragile beast out on a POTA activation, so I’ll probably wait until I’ve done that first. It’s been raining buckets here over the last few weeks, but the rain is supposed to let up soon, and once it does, I’ll be packing this radio out to my nearest POTA park, along with a new EFHW antenna. I’m going to be putting this radio through some serious testing!
