The rig is a simple direct conversion transceiver based around the OXO transmitter by GM3OXX with an SBL1 or similar double balanced mixer based direct conversion receiver.
The design worked very well and allowed me to work lots of good DX on the 15m band using just a simple wire dipole. The RF gain control is right at the front of the receiver and was usually advanced somewhat to minimise AM breakthrough, although this was rarely a problem. Putting attenuation at this point is the best place in a simple direct conversion receiver. The passive filter after the mixer worked well and provided very useful levels of selectivity. A small high impedance crystal earpiece was used on receive.
The use of Neosid toroids was simply because these were available at the time. Today I would used T37-6 (yellow) cores instead. Also the PIN diode switch is a bit of overkill, again used as I had some. 1N4148 diodes will probably be as good or change the TX-RX break-in to use the simpler arrangement normally used with sub-1W QRP rigs.
My very first contact on this rig was around 4000 miles to an SSB station in Chicago. I only had a crystal in the SSB part of the band at the time. I heard him working someone so tail-ended on CW and got, "QRZ the station on CW". A solid QSO resulted, the first of many more once I got a 21.060MHz crystal. Offset between TX and RX was around 600-800Hz and so no special IRT offset was needed. The rig was in use for several years and was, at one time, the main rig used out of choice because it was small and a bundle of fun. Sadly the rig was dismantled to make other rigs.
I have included the original SPRAT article. These show the circuit, the description and parts list. Sorry about the hand drawn schematic, but this was the 1980s!