Loading ...
Sorry, an error occurred while loading the content.

Re: Building the Ensemble RXTX: testing the PA stage

Expand Messages
  • warrenallgyer
    Suggestion: check the polarity on C12. The behavior you describe at R26 is not right. Any differential behavior among the four op amps will cause exactly the
    Message 1 of 38 , Jan 1, 2013
    • 0 Attachment
      Suggestion: check the polarity on C12.

      The behavior you describe at R26 is not right. Any differential behavior among the four op amps will cause exactly the symptoms you are showing; modulation by the input audio.

      Warren Allgyer - W8TOD

      --- In softrock40@yahoogroups.com, "warrenallgyer" <allgyer@...> wrote:
      >
      >
      > One of us is confused and there is a high probability it is me. I have been thinking you were using the 10KHz tone on the "Line Out" input to the RXTX. If you do this it should result in a CW carrier at full power, 10 KHz from the CFGSR indicated frequency. It will be either plus or minus 10 KHz depending on the setting of the IQ jumper. If the IQ phase and gain balances are off then you will see the lesser image 10 KHz on the other side and it should null out with gain and phase adjustments. The resulting carrier should be pure, single frequency with no sidebands. There should be no output at all on the CFGSR frequency.
      >
      > Somehow your 10KHz tone is modulating the CFGSR frequency instead of just mixing with it. So what you appear to have is an AM signal centered around the CFGSR frequency with appropriate 10 KHz sidebands, instead of a CW signal at CFGSR +- 10KHz.
      >
      > I used that tone generator some time ago but I don't remember the application. I think for such tests you will find that Rocky 3.7 has much more capability and it integrates the PTT, frequency control, and tone generation in one package. As a bonus, Rocky includes a two-tone test generator that allows true IMD measurement.
      >
      > Warren Allgyer - W8TOD
      >
      >
      > --- In softrock40@yahoogroups.com, "Orin" <orin.eman@> wrote:
      > >
      > > --- In softrock40@yahoogroups.com, "warrenallgyer" <allgyer@> wrote:
      > > >
      > > > Orin
      > > >
      > > > --- In softrock40@yahoogroups.com, "Orin" <orin.eman@> wrote:
      > > >
      > > > > It's a direct connection through a 20dB attenuator (I tried without the attenuator and got the same results - the SA attenuator will take 1W input MAX).
      > > >
      > > > My spectrum analyzer is cheap and it does not have the close in resolution of the HP but I use a direct connection via two 30 dB attenuators to put a minus 30 dBm or so signal into it and into an RXII with HDSDR for the close in measurements. On HDSDR I can see from -30 down to the noise floor at -120 and I see no 10 KHz spikes.
      > > >
      > > > >
      > > > >
      > > > > >
      > > >
      > > > > I clipped R42 - I saw no change other than a reduction in power out.
      > > > >
      > > >
      > > > Since I don't know where the 10 KHz IMD is coming from I don't know whether the driver is the culprit or not; I suspect it is not. The effect of clipping R42 will be on a normal two-tone audio test looking at distortion products within the SSB bandpass. You will indeed see a power drop as you are decreasing the gain of the driver. But you will also see a dramatic improvement in IMD on the two-tone test.
      > > >
      > >
      > >
      > > I'm thinking the problem is most likely the modulator. I put the RF probe across C20 and see 10kHz spikes there too. I added the picture to the others. Here is one:
      > >
      > > https://picasaweb.google.com/lh/photo/4tgOStSenFkGIWGVJjZXyNMTjNZETYmyPJy0liipFm0?feat=directlink
      > >
      > > I poked around with the DVM and 'scope. Something throws the IQ balance way out as I had to set it to 0.929 in the M0KGK IQ generator to reduce the image. With the IQ balance set to 1, signals at R26, R25, R28 and R27 are the same, so the op-amp circuits seem to be OK.
      > >
      > > One thing that I noticed was that the DC voltage at R26 would decay with PTT off whereas at the other resistors, it would hold steady. I don't see C12 as being leaky as that would tend to increase the voltage on C16. So either C16 or the '3253 are leaking when PTT is off.
      > >
      > >
      > > > > I have implemented the T6 cross-over mod. The carrier leakage is -23dBm with no input signal (about 50uW). I used a fancy RF probe at L1/C21 and it was -37dBm there. I should try again on the other side of C21
      > > > >
      > > >
      > > > So you are seeing a -23 dBm carrier leakage at the output of the RXTX with the rig in receive mode? I just connected mine directly to the spectrum analyzer and I see carrier on 21.020 at -39 dBm with the RXTX in receive. 16 dB is a big difference.
      > >
      > >
      > > No, -23 dBm carrier in transmit mode with no input signal. I do see a carrier in RX mode, but at -36 dBm.
      > >
      > > Orin.
      > >
      >
    • Orin
      ... Oops, I should have said R25... ... You have it right. CFGSR indicates 21.1 MHz, 10kHz single tone into the Line Out input on the RXTX. The image does
      Message 38 of 38 , Jan 1, 2013
      • 0 Attachment
        --- In softrock40@yahoogroups.com, "warrenallgyer" <allgyer@...> wrote:


        Oops, I should have said R25...


        >
        > Suggestion: check the polarity on C12.


        ...which would make the associated capacitor C11. It's OK. I even pulled it and tried another. And for good luck, I swapped C15 and C16. No change in behavior.


        >
        > The behavior you describe at R26 is not right. Any differential behavior among the four op amps will cause exactly the symptoms you are showing; modulation by the input audio.
        >
        > Warren Allgyer - W8TOD
        >
        > --- In softrock40@yahoogroups.com, "warrenallgyer" <allgyer@> wrote:
        > >
        > >
        > > One of us is confused and there is a high probability it is me. I have been thinking you were using the 10KHz tone on the "Line Out" input to the RXTX. If you do this it should result in a CW carrier at full power, 10 KHz from the CFGSR indicated frequency. It will be either plus or minus 10 KHz depending on the setting of the IQ jumper. If the IQ phase and gain balances are off then you will see the lesser image 10 KHz on the other side and it should null out with gain and phase adjustments. The resulting carrier should be pure, single frequency with no sidebands. There should be no output at all on the CFGSR frequency.


        You have it right. CFGSR indicates 21.1 MHz, 10kHz single tone into the "Line Out" input on the RXTX. The image does null depending on the IQ phase/gain settings.


        > >
        > > Somehow your 10KHz tone is modulating the CFGSR frequency instead of just mixing with it. So what you appear to have is an AM signal centered around the CFGSR frequency with appropriate 10 KHz sidebands, instead of a CW signal at CFGSR +- 10KHz.
        > >
        > > I used that tone generator some time ago but I don't remember the application. I think for such tests you will find that Rocky 3.7 has much more capability and it integrates the PTT, frequency control, and tone generation in one package. As a bonus, Rocky includes a two-tone test generator that allows true IMD measurement.


        Unfortunately, Rocky isn't playing nice with my soundcards on Windows 7. I'll try some more if I get chance tomorrow night.

        Orin.
      Your message has been successfully submitted and would be delivered to recipients shortly.