Phase shifters to correct phase error in IQ. How much do I need?
I am building this system http://www.microwave.gr/content/view/229/1/
The wideband 7-12GHz 90 hybrid should have accurate 90 degrees phase split, but the Narda datasheet says it can be varied within +-10 or so degrees throughout bands. Now that is too much for a phasing modulator I think.
Since I am using it somehow narrow-band in the amateur band, in order to correct the phase, should I purchase a phase shifter? http://www.jyebao.com.tw/pdfpath/PTS-A3A8-18-15-f.pdf
What I am thinking to implement is shown at the attachment.
The variable attenuator is used to compensate for the amplitude difference and the phase shift network is used to compensate for the shift difference. Dependent on the values, I may need to interchange their positions or connect them in the same channel.
The software I am using is able to correct any amplitude/phase errors in the audio I/Q signals
1. Will I be able to correct phase and amplitude errors with the attached schematic?
2. What about any imbalances at the combiner? Will I be able to correct them with the attached schematic?
To maintain the quadrature phases, you may use Polyphase Filter.
Search in internet, you'll find many articles about it.
You are talking about the I/Q audio signals. I have no problem with the I/Q signals, since their amplitude and phase can be accurately controlled by the DSP computer software. But one has to correct the RF portion as well and this is what I am talking about.
- - - Updated - - -
The link below is to a thread about generating 90 degree phase shift.
The idea of using an LC tank loop turned out to be viable. Different schematics are shown.
Not sure what values you need for 7 to 12 GHz, nor how closely it will hold to 90 deg. across the entire range.
https://www.edaboard.com/thread246154.html
Also look at related threads listed at the bottom of this page.