Not a Chinese QRP Transceiver #1 – Rockmite ][ 40m

You could say that my QRP transceiver kit addiction is getting slightly out of control but where’s the fun in life if you can’t build things that put a smile on your face?

This whole flirtation with the Chinese kit market has caused me to do some delving into the history of the original kits that they look to emulate or some would say “rip off”.

As I’ve alluded to before the Ham fraternity in America is huge and a lot of the QRP kits or ideas originate from that side of the globe. My plan post Pixie/Frog was to have a look at a thing called a Super Rockmite, which again hails from the Asian markets and can be readily found on eBay and the likes. It is an evolution of the other kits and adds a computer interface to allow the transceiver to be hooked to a PC and to generate CW directly from software. Sound familiar?


Well one thing that this has taught me is that despite liking the idea of a PC doing the hard work, it’s very much a case of missing out if you’re not pressing the key or the paddle yourself and with that in mind I’m having a serious go at learning code using the Koch method using IZ2UUF’s Morse Koch CW app for Android. I’m not saying it’s easy and to become proficient will take a lot of time and effort but then reward isn’t without endeavour!

Now with that very much at the front of my mind when I started reading the forums about the shoddy standard of both Super Rockmite kits and fully assembled purchases, poor audio quality and inability to find English version software it was very much a case of “Do I really want to waste my time on this one?” Short answer, no!

Then a bit more delving into the term Rockmite revealed the history of Dave Benson’s, K1SWL, nifty RockMite QRP transceiver kits, Small Wonder Labs and the NorCal QRP club. What a shame Small Wonder Labs are no more! Until I found out that QRPme have taken on the mantle of supplying an updated version of the kit which is a true testament to the original device. Even better Kanga Products in the UK are the authorised distributors avoiding the pain of being mugged by Border Force and Royal Mail for import duty charges!

It was an absolute no brainer and after a brief email exchange with Rex Harper at QRPme and Dennis Anderson at Kanga I secured a bespoke RockMite ][ enclosure (unfortunately you can’t get the Heathkit green one anymore)


along with a RockMite][ in the 40m flavour


The kit landed today (12/02/16) and I genuinely can’t wait to build this thing!


The enclosure landed on a month later and is a work of art and nicer than anything I could make myself


Unlike the Asian kit’s this thing is a work of perfection on all levels, with everything included and loads of support and documentation available online to assist in the build and getting it up and running. I won’t be rushing this that’s for sure as it’d be great to have this working perfectly from the outset and running as my outdoor rig for the summer in conjunction with something like an Emtech-ZM-2 ATU and a Palm Radio Mini Paddle.

That’s more money and several months away so lets finish the projects sat on the desk first!


Chinese QRP Transceivers #3 – Frog v3

The Frog Sounds Forty-9er clone is the big brother of the Pixie I built last month.


The Forty-9er originated in the Norcal QRP club, the original documentation for that creature is here and there are a lot of hardcore dead bug and Manhattan iterations out there. Unfortunately I just can’t get on with dead bug builds, my only attempt being my Intermediate Licence VFO project.


I’m not doubting it as a build technique and applaud those who can do it as it works, but I’m very much from the school of regimented PCB and breadboard construction!

Like the other Asian QRP kits there are various YouTube videos around showing how various builds have gone for those brave enough to launch in with the soldering iron. If you detect a hint of scepticism emanating, it’s not without reason. eHam didn’t rave about it only rating it a 3.3/5 on the “ham radio enthusiast’s satisfaction-ometer” and I can safely say on receiving my kit I can see why. At least I received a component list and a schematic unlike others, but on sorting the components things didn’t add up or look good.

Running a multimeter over the resistors I’m wondering which factory floor they were swept up from as the markings were dubious in several cases along with the supposed 5% tolerance. Several resistors were way off value, as in Merchant Banker LIBOR rate meddling off! At least this time all the components are in the pack, along with several spares and some completely random bonuses. The crystals are again the Asian market 7.023Mhz variety which I may substitute with the 7.030Mhz flavour which believe it or not I had to get from Greece. At least we’re supporting their failing economy!

There are several notes in Chinese on the documents which got me wondering what I was missing and a very helpful friend did the translation for me, which unfortunately revealed the square root of absolutely naff all! I was hoping that there would be a clue regarding the substitution of various components which were mentioned in my kit’s listing on eBay, which increase the Frog’s output to closer to 2W.

After seeking advice from the assembled Alumni of FPARC I was closer to knowing what I was doing and decided to start building, having dutifully sorted all my components (in the warm) into bags to make things easier.


I took my time on this one mostly because it was bloody freezing and within an hour I’d lost the feeling in my fingers, toes, nose and ears. That said I managed to install all the resistors and a chunk of the ceramic capacitors soldered in place in that time.


Another failing of the kit came to light during this initial stage. There were components listed on the bill of materials which just don’t exist in the circuit schematic or on the PCB and a lot of time was spent trying to work out where these elusive creatures were supposed to go, when in reality they don’t feature and aren’t required.

One failing on my part was my build technique in the face of the unknown! Normally, with a kit you attack things from the lowest level and work upwards through the height of the components so you can physically get at things. That’s all well and good with something which is a banker or so simple you can readily diagnose any construction faults, as with the Pixie. With hindsight I should have worked my way around the schematic, giving myself a fighting chance of being able to trace the various sections of the transceiver.


The power supply would have been a good starting point (top section of the schematic) but as usual hindsight is an exact science and I’m very much reliant on my visual checks and a degree of luck that this thing actually works. As this has taken several weeks, rather than hours to build, when I’ve had 5 minutes I’ve revisited my earlier work and double-double checked component placements.


Here’s the finished build prior to the final visual check and smoke testing. It’s a compact board and wasn’t overly taxing.

I’ve made the following modifications in line with the sellers suggestions to push the transceiver towards the higher power output which this board is supposed to be capable of –

Q1 has been substituted for the supplied 8050 signal transistor alternative, inductor L5 (Red T37-2 toroid between the DC power jack and the white BNC connector) has been wound with 15 turns rather than 16

I also added an application of thermal paste between the D882 Power Transistor and the heatsink to ensure maximum thermal dissipation.

Typically after finishing this kit, following the enclosed documents (as limited as they were) I found yet another bill of materials / component listing as a jpeg image on the original eBay listing.

This document makes a further modification suggestion of replacing CP9 (10uF electrolytic) with a 10 ohm resistor “if you do not like the delay when switching transceiver”.

As clear as mud! Something to look out for at the smoke test.

I eventually decided to leave the crystals alone on this one and kept with the stock 7.023MHz ones. I was going to look at installing SIP sockets on the board to make them removable but this would have meant drilling additional holes in the PCB to accommodate the mod and there wasn’t the space between the tracks. Also, with the potential for a FPARC ‘Pixie Off” I was keen to leave one of my QRP builds with crystals the same as the rest of the community, so at least I could see if it worked in the wild with similar devices available to test against.

Smoke testing needed a similar setup to the Pixie, against my FT857D. The Frog is capable of running on 9V with a claimed maximum power output of 2W and 3W if supplied with 12V.

A sensible place to start was with a 9V supply to make sure the mystical blue smoke wasn’t released as although I’m sure any of the readership hailing from France would be partial to a bit of “grenouille fumé” I didn’t want to see the fruits of several weeks delayed labour frazzle on the desk in front of me.


That all worked OK, so I moved up to 12V. With a supposed difference in output power as described, plumbing my MFJ- 971 into the setup would hopefully give some indication of output power and is a job for another day.

Comparing the output on Fldigi at the two voltages there is an obvious difference between the two signals.image

This is the output on a 12v supply. I’m not overly convinced with the purity of the signal and noticed a few harmonics when playing around. I’ve had this before with radios in close proximity when I was trying to get my SoftRock working. I’ve got a 40m filter which can be placed on the antenna output if necessary so there’s nothing too dramatic to worry about.

To be honest what do you expect for £9.49? The Frog generates its own side tone which is actually quite pleasant on the ear and the LED gives a nice indication between receive and transmit states which is a bonus.

The final stage is to put this thing in some sort of enclosure.

2016-03-11 12.25.34

Another eBay purchase eventually landed me one of these. OK it does the job but I’m not sold on the whole winged look. There will be a certain degree of modification to this but at least the board fits nicely on the internal rails and the predrilled holes are nearly in the right place!

Here’s the Frog alongside the Pixie for comparison.


As a “next stage” in building transceivers this was fun but I’m on the hunt for another challenge now!


I managed to box the Frog today which is a bonus as I’ve got a growing pile of things needing a case starting to clutter the place. Gentle application of a hacksaw removed the “wings” and we have a slightly more fitting case.


As such the finished product now looks like this



The hole in the top is for easy access to the variable resistor that allows the frequency of the oscillator to be pulled a few KHz either way.

Addendum 11/09/16

After an email from Rasmus Svensson who is also building a Frog, it was discovered that there are plenty of kits out there which have PCB’s which bear no resemblance to the supplied schematic of bill of materials. Rasmus was having a head scratcher over an elusive resistor which he’d noticed I’d got but he hadn’t. As such, for anyone else who is Frog’ing here are copies of the schematic, BOM and overview which came with my kit.

scan-160130-0002 scan-160129-0001 scan-160130-0001

Chinese QRP Transceivers #2 – Pixie v3

Well my eBay Pixie purchase eventually landed the other week. The delay in building it has been due to a slight absence of components in the kit which was annoying as my seller was the one rated by others on both eBay and YouTube. That said he no doubt just buys them in by the truck load wholesale so I’m not grinding an axe.

So armed with a few extra ceramic capacitors I decided it was time to fire up the soldering iron and see if this thing actually works.


You don’t get much for your £3-92 but with a little care and attention (it’s needed as this board is tiny and the positioning of the components is pretty tight in places) you get this.


Still armed with a suitable amount of scepticism I hooked things up for the necessary smoke test.


Here we’ve got everything ready to go on my desk. My capacitive touch key is plumbed into my K16 Keyer (still minus a case) which converts the ‘dit-dah-common’ wiring of the touch key to the ‘key-common’ input required by the Pixie as it’s geared for a straight key. The K16 also provides a sidetone so you can hear what you’re tapping out on the key. The Pixie is on a dummy load and my iPod speakers provide a powered audio output.

With my FT-857D set to 7.030MHz the battery was duly attached to the Pixie and . . . it worked.

(For the non radio crowd, the dummy load is a “fake antenna” which absorbs the generated radio signal to prevent it getting out into the real world just in case you’re accidentally producing anything nasty and illegal which would cause Ofcom to come a calling. The reason the FT-857D is receiving the signal without a real antenna on the Pixie is that both radios are about 6 inches away from each other and in a similar manner to placing an analogue quartz watch next to set of head phones and being able to hear a tick, you will pick up some localised radiation.)

You note the genuine surprise here as for £3-92 I really wan’t expecting much!

I even cranked up Fldigi and stuck it in CW mode to capture the moment.

Screenshot 2016-02-10 16.55.40

Here are my feeble attempts at a CQ call immortalised as a memento to this little transceivers inaugural outing!

The next step was to stick it on a real antenna. Unfortunately, as the waterfall in the screenshot above shows, the 40m band was very quiet at that time so there wasn’t an awful lot to hear, so I’ll give it another outing later when the band’s more active.

This Pixie is a V3 and I’m still waiting for the V4 to arrive via China Post.

I will add that I’ve pimped this Pixie by replacing the stock 7.023MHz crystal with a 7.030MHz. This will put it slap bang in the middle of the CW centre of activity in the 40m band.