Gnat-1: 1 transistor transceiver from SPRAT 138

Am I alone in struggling with this design from SPRAT 138?

Today I did a breadboard version for 80m on the bench. On TX I had no problems (as long as I grounded the transistor's emitter directly on TX), but on RX the performance was diabolical and no better than a crystal set with injected carrier. I cannot quite fathom out the function of the transformer T2 or why the key is grounding the secondary of T2. As described in the article I would have thought T2 could be removed and simply key the emitter to ground on TX and decouple a larger (variable) emitter resistor on RX.

Has anyone had success with this design on RX? It may well be me!

Gnat 1 transceiver - anyone made/used one?

Has anyone here built and used a Gnat-1 single transistor transceiver? This was in SPRAT earlier this year and it was designed by Chris Trask N7ZWY. Although I've seen a few pages with photos of copies, I've not heard how well it works and what anyone has worked with one. The picture on the right is from LY3LP's website and shows his version for 20m.

Anyone built the Gnat-1 micro-transceiver?

Has anyone reading this blog built the neat little GNAT-1 transceiver featured in SPRAT 138 (Spring 2009) by Chris Trask N7ZWY? This looks a very elegant design making good use of the small number of parts needed. I must have a go. Unlike the FETer and similar micro-transceivers it doesn't need complex switching and it produces a decent RF output.

Chris tells me he hasn't actually air-tested the circuit (surprisingly), so I'd like to hear how people get on. It is important to keep the key leads short Chris advises.