Phase Shifter Experiments
Playing With Novel Ways To Obtain A 90 Degree Phase Shift For Single-Signal Work
Introduction:
For some time I have been interested in phasing type single signal receivers. This has been on ongoing thought process since the 1970's. Hans G0UPL recently re-invigorated my interest and thinking with his phasing receiver.
Hans uses an active AF phase shifter, made up of RC networks and op-amps. My thinking is that this has got to be way too complex for what is accomplished.
Back in the 1970's era I built a phasing SSB exciter which used simple L-C and C-L networks for phase shifting in both AF and RF sections. The result worked but performance was not spectacular, so that idea was shelved for many years.
Recently I purchased a copy of Experimental Methods in RF Design (ARRL) and re-discovered some long-forgotten phase shifting schemes. These are the over-coupled transformer type design which seems to provide adequately accurate phase shift over enough frequency range to be useful in my designs. This section of my web pages is an experimental evaluation of this technique to determine if it is really workable, and if so, how good is it, what are it's limitations, and what are the design parameters that I need to be aware of when using these in my homebrew designs?
Circuits evaluated in this page are both AF and RF, with the AF circuit being first because I have more questions about it's potential performance.
========== AF Phasing Experiment =========
This is the basic design for a simple AF phase shifter. While I can find no available documentation that comes out and says it directly, the premise seems to be that as frequency deviates from the design center, complimentary phase shifts will occur in L and C legs to keep the output phase relationship relatively accurate over a couple of octaves. With this in mind, I have constructed this circuit, driven it with a variable frequency sine wave, and monitored the relative phase relationship between the two outputs using a dual-trace oscilloscope. The scope A-lead monitors the +45 degree output and the B-lead monitors the -45 degree output. Sweep synchronization is taken from the A-lead input.
[ Scope Image Capture Goes Here ]
From expiriments with this simple circuit, the phase seems to be holding quite well to a 90 degree difference over 300 to 3200 Hz. Obviously this is not an extreemly accurate measurement down to +/- one degree or so accuracy, but the scope does indicate no wild phase deviations over normal audio ranges used for communication receivers and transmitters.
----------------------------
The purpose of this circuit is to determine if the phase shift is still accurate when two back-to-back transformers are used instead of one. The scope trace bvelow shows the result:
[ Scope Image Capture Goes Here ]
From this part of my experiment it seems that the results with two transformers is consistent with that of a single transformer.
-----------------------------------
Where I am going with this is probably obvious to most readers by now:
This would be the layout if a DPDT switch were inserted in the circuit immediately prior and used to reverse phasing of the audio feeding the second transformer. This is intended to provide a convenient way to flip the phase relationship between leading and lagging for changing sidebands in a phasing receiver or transmitter.
[ Scope Image Capture Goes Here ]
This result indicates that switching the audio between transformers is a viable way to swap phase relationships.
========== RF Phasing Experiment =========