The Etherkit CC1 1st Beta – A Trail-Friendly QRP CW Transceiver

About a year and a half ago, I posted that I had completed the first beta version of the Etherkit CC-Series QRP CW Transceiver.  It was a neat little rig, with low RX power consumption (of the order of 50mA – a bit less, I think), full DDS VFO coverage of any one HF band, a built-in keyer with memories, RIT and XIT, as well as firmware that could be updated at will with a simple AVR ISP programmer (you can get them for around $20). It also used a lot of SMT devices, and was my first serious project using these tiny parts (the KD1JV Digital Dial was the first).

My CC-20 beta worked, and I made quite a few QSO’s with it, including some DX. It wasn’t perfect though. The DDS VFO had some in-band spurs, the TX/RX switching produced a thumping sound, the input and output of the crystal filter weren’t as isolated as they should have been, you could hear some low-level processor noise on the receiver audio,  and the sidetone sounded a little rough too.  Although that sounds like a long list of woes, I think that anyone who designs circuits is used to tackling these kinds of issues one by one, until the dragon is slain. We (by which I mean Jason NT7S, the man behind Etherkit) did manage to improve the isolation of the crystal filter by a fairly good amount during this beta build.

Then he came out with the OpenBeacon kit and the EtherProg.  I knew he hadn’t forgotten about the CC-series, but I’m thinking he wanted to get a few other kits up and running before coming back to tackle it again, which he duly did.

The rig has been renamed the CC1 and although it retains the same basic architecture, there are a number of changes and upgrades to the design. It is still a monoband QRP CW HF transceiver (available in your choice of band) with an output of 2 – 3W (depending on the power supply), and it still has a DDS VFO (tuned with a real knob!) that covers the entire band, as well as RIT and XIT (useful for working split), freq readout in morse code and a built-in keyer with memories. The firmware is still also upgradeable via an AVR ISP programmer.  Although at this stage in the development it has not yet been implemented in the firmware, Jason thinks it should be possible to include APRS functionality and WSPR too. That’s quite a lot for a rig that is not much bigger than a pack of playing cards.

The beta kit arrived in a Priority Mail flat-rate box (what a neat sight on top of my mailbox!)  The enclosure is to the left, in the middle was the bag of parts for the EtherProg (a separate Etherkit product which can be used to update the CC1 firmware). The big bag on the right is the bag of parts for the CC1 –

The CC1 parts bag opened up to reveal the inner packaging.  The bag containing the bigger parts has been opened and those parts dumped into a mint tin.  The EtherProg, as I mentioned, is a separate Etherkit product and is available now, but I’ve included it in this photo. You can see the board slid partially into the enclosure –

A view of the underside of the board. Our beta kits had the microcontroller pre-installed. Currently, this was the only way Jason could supply it to us flashed with the firmware, but regular production kits will not have this IC pre-installed (it will have the firmware already flashed though) –

In true Etherkit spirit (the phrase “Open Source Amateur Radio” is on their home page), the beta testing forums are open for anyone to view here, and the forum for the CC1 beta is here. Only beta testers can post in these forums, but anyone can post in the product support forums which are here (you have to register first.)  The CC1 beta forums include schematics and an assembly guide which, although not final of course, will be of interest to anyone who might have an interest in the kit when it becomes available.

A couple of days of soldering, and the receiver section (which is about 85% of the circuit) was finished. Alignment consists of peaking 2 trimmer caps in the bandpass filter, and adjusting the BFO so that the wanted signal is in the center of the passband.  The passband for my filter is not flat – there is a definite peak in the response,  so I adjusted the BFO to place the wanted signal at the peak of the filter curve.  I already had a noise source that I had built to adjust the filters for my K2, and Spectrogram on my computer (for the same purpose) so I used these to adjust the BFO frequency.  Both the noise source and the use of Spectrogram are detailed here. With the receiver aligned, I have now spent every evening since just listening to it. I keep looking at it and thinking, “That little thing is a radio?”

Here’s the CC1 board with the receiver section completed –

You can see the GPS connector at the left-hand side of the board (the rear) immediately under the green key jack –

The onboard connectors are really great. They save a whole lot of hassle with wiring, and make it a lot easier to run the rig on the bench before putting it in an enclosure. In the following picture of the underside of the board, you can see U4, the 50Mhz master oscillator and to the right of it, U5, the DDS VFO chip. On the right-hand side of the board in the center, is U1, the NE5532 AF amplifier (I just saw a cat hair lying on top of U1 – those things get everywhere).  You can also see the space for U2, the transmit buffer –

At first, I thought the receiver wasn’t functioning correctly, because on attaching an antenna, I heard only a very faint increase in background noise. I tweeted to Jason and informed him as such, as well as posting to the other beta testers in the forum.  My theory was that the AF amp had low gain.  As it turned out, it was a combination of the bandpass filter being way off it’s peak, and the initial BFO freq placing the signal fairly well outside the passband of the crystal filter. Had I thought to peak the trimmers before jumping to conclusions, I would have realized that all was well.

The receiver was sounding good. The DDS spurs that were present in my CC-20 beta are no longer an issue.  The crystal filter has better isolation – there is still some room for improvement, and that will be improved further before it comes to market – in fact, Jason just suggested a circuit change in this direction that beta testers are implementing as we speak. The TX/RX switching is very smooth and the sidetone sounds nice. There is a sharp leading edge on the sidetone waveform which gives a clicking sound, but that will just require some simple shaping, which, once again will be taken care of in the production model. EDIT – another blog, and also a discussion in a Yahoo groupo, seem to have misread my last statement as meaning that there are key-clicks on the transmitted signal.  This is NOT the case. The transmitted signal sounds nice. I was referring to the sidetone only, which is a simple thing to take care of.  I emphasize also that this is a beta,  and we will most likely be taking this little rig through another beta before it goes into production. The other issue, the processor noise that was present in the audio, is vastly reduced and by the time you read this, will most likely be cured altogether, as Jason just re-wrote the firmware, which I am waiting to apply to my beta.  Things are looking very good for this little rig.

A couple more views of the board at this stage, before we move on –

Having confirmed that the receiver is working,  the final push was on to build the transmitter and complete the rig.  It didn’t take long – just the installation of 12 parts and 2 more toroids to wind.

Here’s the completed board, before installation in the enclosure –

The world of SMT seemed like a closeted world of intrigue and mystery before I built my first project using them.  I had read web sites detailing the use of solder paste and hair dryers, or toaster ovens for soldering these tiny little parts.  It was a while before I realized that you can actually solder them the good old-fashioned way – with a soldering iron and a roll of solder.  I pick up resistors and caps and place them close to their final resting place on the board with a fine pair of needle-nosed pliers. Then, with a small jeweler’s screwdriver, I gently nudge them into their exact position on the pads. While carefully holding the part down with the tip of the screwdriver, I tack-solder one end in place. Then I solder the other end, and go back to the first end to properly solder it.  I use a 1/32″ chisel tip and 63/37 .02″ solder with a mildly active rosin core.  0.015″ solder would be even better, as it’s easy to apply too much solder (which is where a good-quality de-soldering braid, such as Soder-Wick, proves invaluable.)

IC’s with fine lead pitch are a little trickier. The NE5532 AF amp was relatively easy, as the leads are far enough apart to solder them individually. Needless to say, a very clean and well-tinned tip is vital. I wipe my tip on a damp rag and tin it before every joint – unless I’m soldering a number of joints in quick succession one after the other, such as with IC’s.  The AD9834 DDS chip has leads that are too closely-spaced to solder them individually. The technique that I learned from Jason involves soldering all the leads on one side with a big wodge of solder, paying no attention to whether the leads are bridged together with solder.  Afterwards, you clean up the solder bridges with de-soldering braid and a larger iron tip. A larger tip is useful here because you can wick up the excess solder more swiftly in order to avoid destroying the chip. Jason posted a good description of how to do this in the assembly guide.  Search for U5 on that page and you’ll find the description, along with a picture.  Flux is said to be very helpful here.  I managed it with no extra flux (other than that in the solder) , but plan on getting some for future use.

The CC1 is billed as a trail-friendly rig, and the kit will come complete with a pre-drilled enclosure with silk-screened front and back panels.  The enclosure we received with our beta kits is the exact same enclosure that will go out with the kits, with the exception that ours weren’t drilled or printed.  So the following pictures represent roughly what the final CC1 will look like, without the silk-screened panels. There might be a slight adjustment in the spacing of the controls before the final production model too.

Firstly, this one’s for size comparison with my CC-20 beta –

The board slides into rails in the side of the extruded aluminum case and is held in place by the nut on the BNC connector at the back.  Here’s a couple of front views without the front panel –

Man, is this thing a beaut or what?

I’m very fond of this little rig. I’ve only made 1 QSO with it so far (with Jason NT7S) but have spent every evening listening to it. It’s great to have the earbuds in, listening to 40M on this diminutive little transceiver while working.

I’m hoping to get some audio up at some point, but it may take a while. If you’re wondering when you can get one of these, well, it’s still in development but at this point I think it’s safe to say that it will be coming out. I do know that Jason NT7S is a perfectionist and won’t release it until he feels it’s truly worthy, and all issues have been thoroughly worked out. The design is already very close to where it should be and there’s a great momentum behind it, but we still have a 2nd beta to go through  Stay tuned and we’ll keep you posted.

Advertisement