[hpsdr] Comments on setting up PureSignal

[Brian Lloyd]

Now that people are experimenting with PureSignal with external amps, I
think I see a possible trend for an error that can be made that results in
less improvement than might otherwise be possible with better
configuration.

First, it is clear that there is a fair amount of leakage inside the ANAN
100/100D since it is possible to use PS without a sampler, just by using
the leakage to provide the feedback. What this says that the leakage is a
potential source of error to PS when trying to correct an external
amplifier using an external coupler.

When trying to correct an external amplifier it becomes very important to
make sure that the signal from the coupler is MUCH larger than the signal
from internal leakage in the 100/100D. After all, it is the signal from the
coupler that we want to correct, not the leakage. The way to do this is to
use as MUCH attenuation from the Hermes or Angelia stepped attenuator and
as little attenuation as possible from the coupler commensurate with not
doing any damage to the ADC or the stepped attenuator on the board.
Remember, the stepped attenuator attenuates BOTH the leakage signal AND the
coupler signal so by increasing the stepped attenuator attenuation and
decreasing the coupler attenuation, we increase the coupler-to-leakage
ratio.

Now, as I recall, best performance from PS comes when the signal arriving
at the ADC is just below clipping. That occurs at about -12dBm if I recall
properly. (If I am wrong, please correct me and adjust the numbers
accordingly.) Let us assume that we are using a 1000W external PA which is
+60dBm. That means we need a TOTAL of 72dB of attenuation between the
output of the external PA and the input of the ADC. Typically 30dB of
attenuation comes from the coupler. More attenuation comes from the stepped
attenuator. The rest of the attenuation comes from a pad between the
coupler and the input to the stepped attenuator. The combination of pad and
stepped-attenuator must add up to 42dB of attenuation. It is the ratio of
the stepped attenuator setting and the external pad which will determine
performance.

When setting the attenuation ratio we must be aware of two things: not to
overload the ADC, and keep the signal level below that which will damage
the stepped attenuator on the board. So we have two do-not-exceed values:
-12dBm for the ADC and +24dBm (about 250mW) for the stepped attenuator.
That suggests an optimal value of 36dB of attenuation from the stepped
attenuator and 6dB of attenuation from the pad between the coupler and
input to the stepped-attenuator.

OTOH, we can't get 36dB of attenuation from the stepped attenuator AND we
would be operating it at its absolute limit. If we set the stepped
attenuator to its maximum attenuation, -31dB, then we would need an 11dB
pad. That would give optimum performance with 1000W out of the external PA.
You can get this by using a stepped attenuator between the coupler and the
RX input to the radio, or you can use two standard in-line pads of -10dB
and -1dB. If you are using a standard -15dB pad you will need to decrease
the attenuation setting for the stepped-attenuator by 4dB to -27dB. This
will allow more leakage to affect the performance of PS but should still
result in good performance.

-- 
Brian Lloyd, WB6RQN/J79BPL
706 Flightline Drive
Spring Branch, TX 78070
brian at lloyd.com
+1.916.877.5067
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20140226/4955809e/attachment-0003.htm>