[hpsdr] A new entry in the direct-sampling SDR field

[Nico Palermo]

> Very interesting... Do you know what noise figure you are getting in
> this configuration?

Phil,
if I made no coarse mistakes calculating it, the noise figure of the
THS4509 driver in a 200 Ohm balanced system should be around 7 dB for a
voltage gain of 12 dB. The 1:2 input transformer halves the effective En
of the op-amp (which is the major noise contribution) and has also the
advantage that less voltage gain is required from the op-amp to scale to
the required receiver sensivity. Less voltage gain, less noise compared to
the ADC quantization noise, better overall performance.

> That is close to like 1/3 more I think ($96 vs $66 going by Digi-Key
> prices).  Do you compare the LTC2206 16 bit part to the 14 bit part?

No, at least in terms of cost. When I discovered that the LTC2206-14 was
declared as a 76 dB SNR device (much better than Analog Devices and TI 14
bit competitors) I decided it was the better one for my receiver.
Theory says all us that one bit more should improve SNR by 6 dB. So why
should I have used a 16 bit ADC, two more than those I'm using, for just a
2 dB SNR improvement over a 14 bit device? Let ADC designers improve their
16 bit devices, then I will consider (and pay) them.

> Looks like a nice receiver.

Your also indeed.

73s Nico iv3nwv


 1182380358.0