[hpsdr] QSD gain and noise.

[Phil Harman]

In reply to Ahti and Frank,

Ahti, from memory the NF measurements you suggest are a little high but in the right ball park. The problem of connecting the Janus balanced inputs directly to the QSD caps is noise pick up since this is done at a high impedance level.  A little (low noise) gain after the QSD capacitors would allow Janus to be driven from a low impedance source and help with the noise pick up. 

Alternatively, we placed a header on the Janus board that would allow a little daugher board to be plugged directly in. The board could contain a QSD and the capacitors feed the balanced input over very short leads. 

It seems we still have a NF problem with the QSD  in that it rises with increasing frequency. The Flex5000 figures I have seen indicate that if we measure the NF on 20m then it increases by 10db on 10m and a futher 10dB by the time we get to 6m.

I may be mistaken but I have yet to see a QSD that does not have this problem.  In which case we may want to investigate using say an H mode mixer followed by a 1dB effective NF audio stage.  This has the potential of not requiring a pre-amp on any of the HF bands. I recall seeing an H mode mixer that had I and Q outputs but can't seem to find the circuit. To optimise the H mode we need to change the bias point of the switches as we change frequency but an I2C driven DAC could do this easily.

Some experimentation required!

73's Phil...VK6APH 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20070923/640a3a76/attachment-0004.htm>