开云体育

I noticed that my uBitx was outputting a significant carrier frequency in SSB mode when examining my signal using an SDR today. I had originally calibrated the BFO to 11.055.500 by listening to other stations, and it sounded good. I then tried re-calibrating the BFO, however by the time the carrier was appreciably reduced, the resulting signal was pushed a noticeable amount away from the carrier, which is also no good.

Next, I decided to try inverting the side bands to see if that would help (see /g/BITX20/message/75648 for details on why this happens. Jerry also has a write up here: /g/BITX20/message/44515, and presumably more write ups exist). I was pleasantly surprised to discover that the filter appears to have a much steeper cutoff when used in this mode. At 11.059.900, the carrier was reduced to -35dB (relative to the signal of me blowing into my mic - just looking for orders of magnitude here), while not being shifted far from the desired carrier.



I then switched to USB (well, the software thinks it's USB, but it's actually LSB because of the BFO cal), and was again pleasantly surprised with similarly low carrier.



Has anybody else noticed asymmetric carrier suppression when tuning their BFO? There are several threads in here that touch on the topic of carrier suppression, such as /g/BITX20/topic/32898523 and /g/BITX20/topic/30942464, and plenty of other threads where folks have flipped USB/LSB due to "bad" BFO calibrations. Assuming I change the software to correctly label the side bands, is there a downside to using the 11.059.900 BFO calibration that I'm not aware of?


Reed

Interesting observations.? My first thought is that maybe blowing into the mic produces mostly high frequency audio, so the inversion would put the high frequency closer to the carrier and produce the effects you noticed.? I would conduct the test with a two tone audio input and see if the effect remains.? There are both kits and free PC based two tone audio sources available.

Currently dealing with wife's medical issues, so not sure when I will get an opportunity to test on my v5 (same basic circuits and crystal frequencies).

Just a thought.
73
Evan
AC9TU

One other thought (getting old, and sometimes it takes a little time to get all the memories organized) is that the balanced mixer could also be an issue.? There was a prior thread last year on the v4 about carrier suppression when I was testing using an RSP1 SDR setup.? It was suggested that maybe the BFO mixer was not balanced, as a balanced mixer should not allow the carrier through (supposed to sum out the BFO which is the carrier in this case).? I had tried to verify with my testing, however could not.

Will monitor the thread for other's input.

Again, very interesting observation.? Thank-you for posting!
73
Evan
AC9TU

Hi Evan,

I went ahead and played some tones. 440Hz, and a 1kHz-2kHz sweep. Looks like in both BFO tuning cases, the 440Hz signal is roughly the same magnitude as the carrier, and both BFO tunings resulted in a -70dB average reading for the sweep, although the 11.055.500 had significantly more 440Hz.

I also tried tuning my BFO to 11.053.550 (as shown in my original post), and as you predicted, there wasn't any power when playing the test tones, so it must have purely been the high frequencies of the white-ish noise going through in that image.

Next, I put a test sweep from 300-3000Hz (2s sweep period) into the radio, and found that my output audio signals were approximately the same with the BFO calibrations set to 11.054.700 and 11.060.000. However, at these settings, the carrier was 5dB higher (aka less suppressed) when using the 11.054.700 value. Both of these settings reduced output at 600Hz to around -15dB vs tuning the BFO for maximum pass of that frequency.

Ultimately, it seems that my radio just isn't capable of allowing 300-600Hz fully through without also letting significant carrier through, regardless of BFO calibration direction. I decided to set my BFO to 11.054.700 for now, to suppress the carrier a reasonable amount at the cost of sounding tinny. I can't help but feel that there's got to be a hardware mod solution to this...

On a related note, the v6 schematic says that the filter has a 12MHz crystal for X7. My board's crystals are all stamped with 11.0592, though, so this appears to just be a labeling failure, and not an actual part swap. Ashhar - you might want to fix this.



Reed

Hi Reed,

For SSB use the filter shouldn't be counted on for much more than about 6 dB of carrier rejection and the carrier/BFO should be located at that point in the filter. The mixer if balanced correctly would be responsible for most of the rejection and 30-40 dB would be a pretty good value range, with a total rejection of around 36-46 dB.? For most sideband filters the 6 dB point would be located so that the voice frequencies (in the flat portion of the passband) would start somewhere between 200 and 300 Hz. Depending on the width, the high end of the filter frequency at -6 dB would be around 2.4 to 2.6 KHz. i.e. 2400 Hz - 300 Hz = 2.1 KHz, a favorite for the Collins "S" Line.

I'm very impressed with the crystal filter in these little rigs, as I mention in my first posting, here.?

If your SDR will allow you to collect peak hold information it would be interesting to see how your breath and any other noises you can make fill the passband. Such an exercise would provide a very accurate depiction of the filter response. I would sure be interested in seeing what that looks like, if you can pull it off.?

In the bottom display in your posting and depending on the video response of the display and image, what is indicated is a carrier suppression of around 37 dB (77 dB - 40 dB = 37). Not all that bad! I have to go with that since I don't know how you normalized the display.

I can't say that much would be gained by reversing the sidebands, as you suggested, it's likely that only a small change in the fidelity would be impacted.

Thank you again for your assistance the other night.

--

Hi Dan,

Unfortunately I couldn't figure out a way to get SDR# software to do a max-hold function. I agree that would be a useful datapoint. The AirSpyHF+ I'm using isn't calibrated that I'm aware of, so I just use it for relative measurements.

I don't speak at a -40dB level - that's just me blowing hard into the mic, so a best case scenario for signal. With a normal speaking voice into the stock mic, it's more like 5dB of suppression, as shown below.


I've also noticed that when in CW mode, if I set the TX->RX delay to be long, there's still a diminished (38dB vs full) but very present (30dB vs noise) carrier:


Got any tips on how I might balance my mixer(s) better, or is that a fool's errand?


Reed

Reed,

I have seen the same reduced signal in CW mode during key up just prior to time out.? That has been with both my v4 and v5 rigs.? I have always assumed that it was going back to TX in SSB mode for the short interval and some noise was in the audio section causing the small carrier.? It may be the same BFO bleed through as in the SSB audio.

As to "balancing" the BFO mixer, the term is used when you have discrete diodes, and can involve bias resistor adjustment, or diode selection.? The Matched diodes in the BAT54SL chip do not have balancing capability.? The only thing you could check is if both diodes are working.? Unless you remove the chip, static tests will not work in the circuit.? I did check with an oscilloscope and verified that I was seeing both diodes rectifying the BFO.? What I assumed is that the imbalance is due to slight miss match of the output to the crystal filter pad.? In my research on line into balanced mixers, the balance only works if impedances are matched.?

Just suggestions and observations on my part.? I may be wrong and any of the above points.
73
Evan
Ac9TU

Hi Evan,

The remaining signal is the "balanced" version the first IF mixer. The CW key up/key down routines either intentionally unbalance, or try to restore balance to, the IF mixer, but they don't change to SSB mode until TX times out. The other two mixers are turned off.


Reed

Reed,

Thank-you for the explanation.? I should have gone back to look at the code to verify what I had said.??

73
Evan
AC9TU

I agree with the technique, in fact an antenna noise bridge might be just the thing. I don't know if Reed's setup has the dynamic range for just an antenna connection. Reed?
--

Is this a V6? If so I think a CW carrier replaces the BFO during transmit while in the CW TX mode, provided I read the circuit and theory of operations correctly. ?
--

If you have a wide filter AM position on your regular rig you could even feed the static from an open frequency into the mic and see what happens.