Topics

Carrier suppression on uBITX v5 - -35 to -40 dB ok?

Rob French (KC4UPR)
 

For USB/LSB on the uBITX, what is a reasonable carrier suppression?  (and how is that properly measured, i.e. I assume it's carrier relative to peak?)

Background:  I should have left well enough alone, but... I've been using my uBITX for FT8 and PSK31, and had directly connected my uBITX mic input to my USB soundcard's headphone output, and my uBITX speaker output to my USB soundcard's mic input.  No special adapter circuits or anything.  Worked fine, bunch of good QSOs.  Fast forward, I was reading about interface circuits, and saw someone said that I MUST put a blocking capacitor between the respective inputs/outputs.  Well, I didn't do that.  Also read about someone saying that a few volts into the mic input could kill diodes in the balanced modulator, leading to carrier leak through.  So then I got paranoid and figured I should check my signal to make sure I hadn't done something like that.  Which brings us to now...

Test Setup:  uBITX v5 transmitting into a dummy load.  RTL-SDR dongle plugged into my PC, operating in direct mode.  "Tone generator": me whistling into the mic.  GQRX running on my PC to view the spectrum.  Just eyeballing the spectrum display.

Results:

(1) with background noise only ... noise floor ~ -75 to -80 dB on GQRX readout.  With PTT on, no tone, desired sideband ~ -75 to -70 db, carrier ~ -75 to -70 db (visibly distinct from desired sideband), opposite sideband absent.

(2) with a tone ... noise floor ~ -75 to -80 dB on GQRX readout.  With PTT on, "whistle" tone, desired sideband ~ -35 dB (peak), carrier seems to be unchanged (-70 to -75 or so), opposite sideband still absent (carrier kind of gets lost in the spectrum from my "whistle" at this point, but as I increase the volume from nothing, the carrier appears to be staying the same).

Does this all seem reasonable?  

Thanks!
-Rob

(PS - As an aside, I checked the DC bias on the USB sound card mic in and headphone out... mic in ~ 2V, which I assume it to drive electrets, and headphone out is 0V)

Tom, wb6b
 

Hi,

I just double checked the schematic. It looks like there is adequate DC blocking capacitors in the mike input circuit. Also, the mike input already has a DC voltage applied to power the microphone. So, unless the sound card doesn't like the DC voltages and causes distortion, it is likely not the source of the carrier leakage.

I'm not sure how much is good or bad, but the topic of carrier suppression and fixes has come up on this group. Hopefully, other may be able to shed more light on what level is acceptable.

I have had good success with connecting the high side of the volume control directly to the microphone input of a USB sound dongle. And the headphone output of the USB sound dongle directly to the microphone input of the uBitx. Some sound adaptors may work better or worse in this configuration, but the generic adaptor used seems fine in this arrangement. It is possible some sound adaptors might not like the DC voltage (but low current) the uBitx puts on the microphone input applied to the headphone output.

Tom, wb6b

Ashhar Farhan
 

The carrier to signal ratio needs to define the signal level as well. A 'Haaalllowww' is usually insufficient (though I use it all the time).
You could download a wav file of two tone and play that into the mic. That is the simpler way to do it properly.

- f

On Sat 17 Aug, 2019, 1:34 PM Tom, wb6b, <wb6b@...> wrote:
Hi,

I just double checked the schematic. It looks like there is adequate DC blocking capacitors in the mike input circuit. Also, the mike input already has a DC voltage applied to power the microphone. So, unless the sound card doesn't like the DC voltages and causes distortion, it is likely not the source of the carrier leakage.

I'm not sure how much is good or bad, but the topic of carrier suppression and fixes has come up on this group. Hopefully, other may be able to shed more light on what level is acceptable.

I have had good success with connecting the high side of the volume control directly to the microphone input of a USB sound dongle. And the headphone output of the USB sound dongle directly to the microphone input of the uBitx. Some sound adaptors may work better or worse in this configuration, but the generic adaptor used seems fine in this arrangement. It is possible some sound adaptors might not like the DC voltage (but low current) the uBitx puts on the microphone input applied to the headphone output.

Tom, wb6b

Gordon Gibby
 

​A significant number of sound interface systems will have a TRANSFORMER here and there.   It is important in most cases to provide DC isolation to prevent the transformer from dramatically changing the biasing conditions of the mic input amplifier of the radio to which it is connecgted.   This blindsided me once when I connected a homebrew isolator (with transformers) to an icom 2 meter rig to do packet -- and the signals sounded horrible.   You won't damage the transformer, but you can throw the linearity of the mic-amp stage out the window.   A Signalink in the same situation did FINE -- because it had capacitor dc-isolation.   I learned the hard way to DC isolate.


Cheers,

Gordon



From: BITX20@groups.io <BITX20@groups.io> on behalf of Tom, wb6b <wb6b@...>
Sent: Saturday, August 17, 2019 4:04 AM
To: BITX20@groups.io
Subject: Re: [BITX20] Carrier suppression on uBITX v5 - -35 to -40 dB ok?
 
Hi,

I just double checked the schematic. It looks like there is adequate DC blocking capacitors in the mike input circuit. Also, the mike input already has a DC voltage applied to power the microphone. So, unless the sound card doesn't like the DC voltages and causes distortion, it is likely not the source of the carrier leakage.

I'm not sure how much is good or bad, but the topic of carrier suppression and fixes has come up on this group. Hopefully, other may be able to shed more light on what level is acceptable.

I have had good success with connecting the high side of the volume control directly to the microphone input of a USB sound dongle. And the headphone output of the USB sound dongle directly to the microphone input of the uBitx. Some sound adaptors may work better or worse in this configuration, but the generic adaptor used seems fine in this arrangement. It is possible some sound adaptors might not like the DC voltage (but low current) the uBitx puts on the microphone input applied to the headphone output.

Tom, wb6b

Evan Hand
 

Rob,

To answer your question, -35-40 is what I have measured using an SDR Play RSP2 and the spectrum analyzer software (this is the only tool  that I currently have that can get a low enough resolution band width RBW to make the measurements). I did use a two tone generator on the mic input to verify the audio vs carrier.  Simple matter of turning on and off the tones while keeping the uBitx in transmit mode.

There are two possible causes for carrier leakage that I have fund:
1 - mixer diode issues as you have pointed out
2 - BFO out of alignment

I found that by changing the BFO setting I will get some carrier injected back into the signal, even if the mixer is OK.

I tested the mixer diodes by injecting an ac signal into the BFO transformer across the output coil coming from the BFO to the mixer and measuring the common connection of the mixer diodes with an oscilloscope to verify a half wave looking signal.  If they were bad, should not get a full wave rectified waver form. Not sure if it is the best way to verify the mixer diodes, however seemed to work for me.

NOTE: in doing the transmit test I first established the power level at frequency using CW and an SWR/POWER meter in the line between the rig and the dummy load.  I then adjusted the tone generator output to be at the same or slightly lower level, and that the distortion on the measured spectrum was minimal.  As Ashhar stated, there are audio programs or recorded wave files that can provide the constant 2 tone audio.  There are also good kits that can provide a very pure tone/tones.

Also of Note: I am not at my QTH so cannot get to the actual rig and verify the connection points for the mixer diode verification test.

FWIW

73
Evan
AC9TU

Don - KM4UDX
 

Tom+++

I built an little Easy digi interface kit with small transformers, DC blocking cap, to bridge the ubix audio i/o to the laptop audio i/o.  Then I realized by laptop audio i/o jacks were all broke, so I had to use a cheap external usb sound card just to gain good audio i/o jacks from the laptop. Going from uBITX audio i/o to EasyDigi i/o to cheap usb sound card dongle i/o to laptop usb port all worked 100% fine.

But.

Is the EasyDigi interface totally redundant given that I have to use a external sound card anyway?  I don't need the audio transformers for isolation -- the external sound card does that function, and nothing is bothered by the mic+v so  I don't need the blocking DC cap either.  Can I remove the EasyDigi from the audio i/o chain and go straight from uBITX audio i/o to sound card i/o??

Thanks oh wise ones for the insight and guidance!

Don