The polyphase network is a combination of resistors and capacitors which can convert the 4-phase audio from a Quadrature Sampling Detector into single sideband output. Low noise LM4562 op-amps are used in instrumentation amplifier configuration to provide excellent common mode noise rejection in the final difference amplifier. The PCB size is 37 x 27mm and is designed to plug into the receiver module kit. The intended application of this module is in CW, QRSS and WSPR receiving equipment where the important frequency range of interest is 700 to 1800Hz. If you have a stereo input soundcard then you can process the I-Q directly from the receiver audio output, without needing this polyphase kit.

Assembly instructions

Printed instructions are NOT supplied with the kit. You can download the kit-building instructions for your PCB from the links below. Check carefully the PCB revision you have. PCB Rev 2 started shipping on 19-Nov-2019. There are four additional 47-ohm resistors in Rev 2 and the PCB layout is slightly different. 

CLICK HERE to download the kit-building instructions for the polyphase network module kit, PCB Rev 2. 
CLICK HERE to download the kit-building instructions for the polyphase network module kit. 
CLICK HERE to download the FRENCH kit-building instructions (many thanks to Gilles F1BFU for the translation!)
CLICK HERE to download the ITALIAN kit-building instructions (many thanks to Roberto IZ7VHF for the translation!)
CLICK HERE to visit the receiver kit page and download instructions for the receiver module kit

Kit photographs

The following measurements show the audio response and the unwanted sideband cancellation, for the audio range 0 to 3000Hz. The intended application of this module is in CW, QRSS and WSPR receiving equipment where the important frequency range of interest is 700 to 1800Hz. 

Component matching

Pim PA2PIM has commended on this design. Pim is an expert on polyphase networks and has an extensive web page on the topic, that includes theoretical description and practical explanation, and an excellent spreadsheet for modelling polyphase networks. Pim says:

"For squeezing the most out of the delivered Capacitors measure them and group them in set's of 4 with the least variance within a set.. Although definitely no rocket science there is a Capacitor selection worksheet in the mentioned spreadsheet to speed that up. http://antennoloog.nl/data/documents/Polyphase_PA2PIM_2.4-20060324.zip (I made it when I got handed 250 Capacitors, not 16) Once sorted put the best matched set at the output of the network, i.e. C13-C16 and then work backwards. See par 8.5 in: http://antennoloog.nl/data/documents/Understanding_and_designing_Polyphase_networks_V4.0.pdf The component values you picked are capable of 70+ dB suppression. With the 5% Capacitor tolerance and no alignment feature it is an interesting set of design choices. Comparing theory to the results you showed that it works out very well."

I agree with everything Pim says. Matching the capacitors as closely as possible in sets of 4, should improve the performance of the network (unwanted sideband suppression). The polyphase kit produces good sideband rejection anyway, and is designed with no adjustments to make, so that it is easy to build. But if you have the equipment to measure the capacitors, then by all means, get some more dB out of it!