Power SDR Software

[pw fallon]

OK your SDR SW gurus I have an interesting issues that I need some guidance with.

I am currently using a Softrock card configured to sample the 455KHz IF on my 75A4. I use PowerSDR V2.5.3 with my EMU-0202 sound card. The PC is a high performance HP laptop with I5 processor running windows 7. It works great with one exception.

To T/R the Softrock/EMU combo I short the sound cards inputs. When I go from xmit to rec I get an image of the desired signal that is 14KHz higher then the frequency of interest . Since this is an IF application the Softrock LO frequency has been set to 462KHz and the image is 2x the difference between the LO(462KHz)  and the IF(455KHz).

Now this is where is get interesting.  If I do nothing and just watch the pan-adapter the image component slowly reduces in amplitude and eventually disappears into the noise floor. This "correction' will take between  10-60 secs to occur and depends on the signal strength (faster signal have faster correction). Clearly the SW is doing some sort of adjustment and once it is finished it works great (until the next time I go to xmit….)

I have tried countless combinations of sound card, audio and DSP settings to see what effect they may have on this issue. The only setting that I have found that has any impact is to adjust the Phone Buffer Size (RX) under the DSP tab.  It does not seem to matter whether the initial setting is (512, 1K, 2K…) , just switching between them does. Once I switch the buffer size (for ex. 512 to 1K)  the image disappears immediately.

All suggestions??

ps I dont see this if I run HDSDR….

Joe (PW)

[w3jn]

I'm wondering if instead of shorting the input signals you're swapping I and Q, or possibly just shorting one of the input channels.  Perhaps a DC transient is cornfusing the program; as I recall powerSDR has some sort of automagic image rejection algorithm.

Me, I'd short the output (or disconnect the speakers, or route the audio to a dummy load or something) and see what happens.  Better way to do it, you can monitor your transmitted signal on the waterfall but not have any output noise.

[N2DTS]

On my standalone sdr receivers, i short the computer audio line out and short the antenna input.
I can not hear anything, but I can see my TX signal.
Instant recovery.

With an IF receiver, just short the audio output.

[steve_qix]

Hmmmmm.. shorting the inputs...  Does this mean you are shorting the softrock output too?

It *looks* like a software problem, but may be a hardware issue if you're shorting the softrock output.  Even though I'm sure the devices are designed to work into a short circuit, chip heating, etc. will have effects on the output level, prop delay (phase), etc.

What happens if you disconnect the softrock output (but otherwise leave it on running and without a short circuit)?  Do you still have the problem? 

Could also be a DC level shift that takes a long time to settle. 

The soundcard is almost certainly AC coupled. The softrock output may be TTL (0 to +5).  This will cause a DC shift, that will need to be integrated by the DC blocking caps in the sound card input (or any DC blocking cap in the softrock output).

I would not short or disconnect the softrock between Rx and Tx.   In my shack I have a softrock, and when going from Rx to Tx, a relay AT THE RECEIVER ANTENNA INPUT shorts the input and disconnects it from the feeder coax, another relay shuts off the speaker, and also switches the audio input going to the main receiver audio system from the selected receiver, over to the mod monitor output (so the mod monitor now comes through the headphones rather than the receiver).

But otherwise, everything is still operating normally.  I have always found this the best way to mute and protect the receivers.

Regards,  Steve

[WD8BIL]

There is a "keying" line on the softrock that is jumpered during assembly. Did you try un-jumpering and muting with that?

[wa3dsp]

Curious if you ever tried SDR-radio software. It is great for RX. I used powerSDR several years ago but I find the V2 sdr-radio to be great.  I always had image and adjustment problems with powerSDR but never with sdr-radio.  It is written by Simon who writes Ham Radio Deluxe.

http://v2.sdr-radio.com/

I never did IF input but I have used a softrock with both powerSDTR and sdr-radio. Currently I am using an Afredri on the Internet as a remote receiver with sdr-radio and it works great.

 

[KQ6F]

Sounds like what you're seeing is a feature called WBIR (Wide Band Image Reject).  The very early PSDR versions had software controls for manually tuning out amplitude and phase differences between the I and Q signals inherent in the hardware and sound cards.  The controls were very touchy and were best used in conjunction with a signal generator for Rx and a spectrum analyzer for Tx.  People complained and so they dropped the manual controls and came up with WBIR.  But it never worked very well IMHO.  For one thing it required large signals to trigger the algorithm - greater than about -50dBm.  And it took a long time to beat down the image - up to a minute depending upon the signal strength.  And of course if you changed frequency (with a new image), the process had to begin again.  And finally, I have seen many instances where the image started out unnoticeable, but because of low signal strength the algorithm began indiscriminately wiggling the phase and amplitude controls causing the image to actually increase.

I kinda thought they have dropped WBIR but apparently not....

[pw fallon]

Hi Folks,

Thanks for all your comments. By modifying the T/R switching to deal with the audio output of the sound card and the RF  input to the soft rock I should be able to work around this issue

Steve during XMIT the input to the sound card is shorted and the output of the soft rock is open circuited. i probed around and did not see any funky levels, dc offsets, etc during the T/R cycle.

Perhaps I have run into a "behavior"  of the WBIR feature.  What I found rather interesting was the immediate effect that the changing the buffer size has on cleaning up the image….

BTW I have also tried SDR-COM which does a nice job on supporting my sort of receive centric application. I believe that it does not exhibit this image issue but I want to confirm this. HDSDR runs image fee in this application although its' lack of a true sync detector makes it not my 1st choice.
 
Joe

[N2DTS]

What is a 'true' sync detector?
HDSDR has ecss, which is sync detection.

I think I like sdr-radio better though.

[WBear2GCR]

Wondering if there is any reason to not mute the Softrock, and merely drop a suitable attenuator (pad -if one is required after the T/R relay switches) on the rcvr antenna side and continue to monitor the transmitted signal via the receiver? Since there is a propagation delay in the computer processing, I guess one would need to mute the speaker.

I successfully monitor my xmit audio with much the same method using an R-388, unmuted. The resulting "S meter" reading is approximately a 20dB over S9. Sounds good this way. No feedback until I get very close to the monitor speaker which is ~7-9ft away. Using a D104 which is not a terribly unidirectional mic...

[w3jn]

PW, try CuteSDR.  It has a few irritating idiosyncracies with its menu system but in Sync AM mode you can select USB to the right audio channel and LSB to the left channel, and adjust the passband of each sideband independently.  As I recall PowerSDR will do the latter but will not do the independent sideband thing in Sync AM.

[pw fallon]

I tried the Cute SDR SW but I don't believe that it will support a Softrock IF application.

IMHO the ECCS detector function in HDSDR sounds exactly like what you would expect if using a product detector to receive AM. Compare it to how the Sync detector function works in PowerSDR, SDR-Console, etc and I  think you will notice a significant delta in the relative performance.

PW

[W1AEX]

Yeah, I couldn't get CuteSDR to function with the Si570 USB in my new little USB SDR toy. It seems to look exclusively for the SDR-IQ firmware "handshake" before it will run. Joe, the version 2.x series of SDR-Console suggested by WA3DSP has the cleanest implementation of synch AM that I have seen. The AGC performance is also outstanding. I honestly like the way it behaves better than most versions of PowerSDR. From reading the thread I'm not sure if you tried that version or the earlier 1.5 version that is not in the same league (in my opinion).

Hey, if you want to hear your own voice in version 2.1 of SDR-Console, check out this link:

http://www.youtube.com/watch?v=cBrgdo1fm24

I have not seen the "recovery" behavior that you described at the top of the thread, but what I do is similar to what Brett does. I use several relays tied into the Viking T/R that disconnect and ground the ANT of the SDR and open up the input side of the computer speakers. No problem recovering after a transmission.

Rob W1AEX

[pw fallon]

Thanks for the link Rob.

I agree with your comments regarding the sync detector behavior in SDR-Console v2. It seems to have a very nice balance between the loop lock behavior and it's gain management.

[W2VW]

Can any of you suggest a circuit that will short out somebody elses transmitter while I transmit?

[W2NBC]

I'm sorry.. did you type something? 

[K1JJ]

Can any of you suggest a circuit that will short out somebody elses transmitter while I transmit?

This service is available for a price.  Just axe these FB HAMs: