[hpsdr] Strange Janus effect

[Phil Harman]

Hi Al,

This is what I suspect you are seeing.

There are two audio outputs from Janus, one from the TLV320 and the 
other from the PWM DAC.  We originally intended the TLV320 to drive 
headphones/speakers/line out and the PWM DAC to provide I and Q to the 
SDR Tx.

Very late in the design I spoke to Bob N4HY who let me know that he was 
considering altering PowerSDR to move the I and Q signals away from 
baseband to get rid of the low frequency noise that some sound cards 
produce.

This change would have meant that we could not have used the I and Q 
outputs to drive the Tx since they were intended to work at baseband. 
So we modified the FPGA code to send the Rx audio to the Janus I&Q 
outputs and the I&Q signals to the TLV320.

The PWM DAC that is now used for Rx audio uses about 11 bits as apposed 
to the 16 bits that the TLV320 uses.  So,if you reduce AGC-T (RF Gain) 
and increase the AF gain you reach a point where you starve the PWM DAC 
of bits and it starts to distort.  I can hear this happening if I set 
the controls as above.

There are a number of options (since we use FPGAs!)

1. Keep the RF gain at the level at which the band noise just does not 
activate the AGC - which is how it was intended to be used.

2. Change the Ozy FPGA code so that we swap the audio outputs - we will 
need you to run some test on the I&Q outputs to check the S/N, phase 
and amplitude balance etc is OK. This should be OK unless PowerSDR 
changes to non baseband I&Q signals some time in the future.

3. Wait for Penelope and/or Mercury since both have additional TLV320s 
which we can use for Rx audio!

Sorry for the long winded answer and hope this makes sense.

73's Phil....VK6APH



Quoting "a.groff" <k0vm at mchsi.com>:

> ***** High Performance Software Defined Radio Discussion List *****
>
> On Aug 30 I posted to the HPSDR mailing list
> -------------------------------------------------------------------------
>
> Subject: [hpsdr] Strange Janus effect...
>
> ***** High Performance Software Defined Radio Discussion List *****
>
> I am noticing a strange cross-talk ( ?? ) effect with Janus (& OZY ? )
> when used with SDR-1000 & PowerSDR  (v1.9.1 svn 1354, v1.10.2 ) that I
> do not notice with the D44 sound card...
>
> I notice this effect in the receive audio when I start to decrease the
> AGC-T ( RF Gain ) control in Power SDR.  As the received signal starts
> to disappear from audibility, it is replaced by very distorted replica
> at a reduced amplitude.  The effect was first observed in SSB modes but
> is also present it AM mode ( maybe more so. )
>
> Does you one else have this effect or do I have a problem with my
> Janus/Ozy/system ??  How would I start to isolate the problem?
>
> AL, K0VM


----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.


 1192152086.0