Tweet
If you want to use full wave synchrounous demodulation in an IB the most usual approach is to have differential receiver outputs, and then use a dual phase clock and two SPST switches for each channel - or use a single phase clock with one SPDT switch for each channel.
Either how you do it means your PCB layout will be substantially more complex than if you were to use half wave detection.
However, there is another way. It could fairly easily be applied to one channel of, say for instance, the all-metal channel of a Tesoro Lobo. Doing only one channel would minimize the additional work involved with routing the board because the modification could be localized to one relatively small area of the PCB.
It is desirable to start with a MD that outputs a fairly clean transmit reference voltage. Not intending to slight the Tesoro Bandito, but its clock signal is perhaps too un-symmetrical for best results with this mod. 50/50 clock symmetry is ideal, of course.
This circuit is based on an application taken from TI INA105 datasheet called "Digitally Controlled Gain of ±1 Amplifier". If you clock it at your received frequency it becomes a fullwave synchronous detector. I figure I'll add this to the All-Metal channel of my LOBO clone that's in work. Simple enough.
Preferably you'll use an opamp with respectable slew rate; just how fast should depend on your frequency of interest. When the amplification factor changes from + to - the output signal is driven to the opposite polarity, and the quicker, the better. The LT6220 is rated at 20V/us and I'll use it at 20kc. It's none too fast, but only draws about ±1.2ma.
Either how you do it means your PCB layout will be substantially more complex than if you were to use half wave detection.
However, there is another way. It could fairly easily be applied to one channel of, say for instance, the all-metal channel of a Tesoro Lobo. Doing only one channel would minimize the additional work involved with routing the board because the modification could be localized to one relatively small area of the PCB.
It is desirable to start with a MD that outputs a fairly clean transmit reference voltage. Not intending to slight the Tesoro Bandito, but its clock signal is perhaps too un-symmetrical for best results with this mod. 50/50 clock symmetry is ideal, of course.
This circuit is based on an application taken from TI INA105 datasheet called "Digitally Controlled Gain of ±1 Amplifier". If you clock it at your received frequency it becomes a fullwave synchronous detector. I figure I'll add this to the All-Metal channel of my LOBO clone that's in work. Simple enough.
Preferably you'll use an opamp with respectable slew rate; just how fast should depend on your frequency of interest. When the amplification factor changes from + to - the output signal is driven to the opposite polarity, and the quicker, the better. The LT6220 is rated at 20V/us and I'll use it at 20kc. It's none too fast, but only draws about ±1.2ma.
Bit pants.
Comment