> -----Original Message----- > From: flexradio-boun...@flex-radio.biz > [mailto:flexradio-boun...@flex-radio.biz]on Behalf Of Christos > Sent: Saturday, January 03, 2009 3:33 AM > To: FlexRadio biz > Subject: Re: [Flexradio] carrier on ssb > > > It is not actually a carrier but LO leakage from the commutator > ic FST3253 and there is no adjustment you can make to null it on > this implementation. > I use sdr1k with the 100w PA and I have measured this leakage to > be -20dbm @ 160m (70db below 100w)and 0dbm @ 10 m (50 db below > 100w but only 30 db below 1w). > While you transmit on ssb it is on your vfo frequency and at > am/fm it is on vfo - if. > You can also trace it during reception but with much lower level. > > 73 Christos > > > _______________________________________________ > FlexRadio Systems Mailing List > FlexRadio@flex-radio.biz > http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz > Archives: http://www.mail-archive.com/flexradio%40flex-radio.biz/ > Knowledge Base: http://kc.flex-radio.com/ Homepage: > http://www.flex-radio.com/ >
If it's exactly at the LO frequency, it's really both, as it should be considered an "RF carrier" (since it is) that is the direct result of "LO leakage" through the QSE. All QSE's have it, to a lesser or greater degree. SoftRocks, David Brainerd's QSE, the SDR-1000, and others have it, and it's measurable. I assume the Flex-5000 still uses a QSE, so it will also have it. Newer switch IC technologies, and careful design may have allowed them to improve on the leakage from the SDR-1000, however. The phase and amplitude adjustments of the I/Q signals can adjust the RF image level, but not the LO leakage, because they adjust only the audio/IF IQ signals, and are not in the LO carrier path. Since the LO is usually (always?) a set of digital signals that turn on and off the QSE switches themselves, adjusting them is difficult. The real issue may be how much power you are actually transmitting. Just as Christos says above. If you are running a SoftRock barefoot at one watt (+30dBm), and the LO leakage is 60dB down, you are radiating a signal at the LO freq at roughly -30dBm (.001 milli-Watt) in addition to your wanted signal at about +30dBm. It will probably not be heard too far away. But, if your are running 100W (+50dBm), the leakage is now around only -10dBm (0.1 milli-Watt), possible to hear much further. It's not always easy to get 60dB carrier suppression on some QSEs. More like 45dB on some (which is only 6 S-units). That would mean that your 1W rig (+30dBm) is putting out a -15dBm (0.032 milli-Watt) unwanted carrier. If you move to a 100W (+50dBm) rig, your unwanted carrier is now +5dBm!! That's 3.2mW. People have (rarely) worked great DX on HF with 3.2 milli-Watts. If (for example) you are being received at ten over S9, your LO leakage is S6 at that same location. Also, keep in mind that no simple filtering (LPF, BPF) can reduce either the LO leakage or image signals, since they are so close to the wanted signal. Running a KW (+60dBm) is even worse. With 45dB leakage, you are actually transmitting a LO leakage signal of +15dBm, or 32 milli-Watts! If you used a KW with 60dB leakage, you would be transmitting 0dBm (1 milli-Watt) of leakage signal. The LO leakage comes from many factors, such as: How balanced the resistance, impedance, and levels are around the QSE switch (we may have some control over this). How balanced the switches themselves inside the QSE IC or IC's are (no control, other than change the part). How balanced is the transformer used? number of turns, how even is the the "twist" between the two primaries, etc. Are the QSE clocks to the switches exactly 50% duty cycle, and no clock jitter? More? If you really wanted to, I think you could fine-tune some of this leakage out, but it would take some work, and your improvement would be limited. And yes, LO leakage can often there on receive as well. Even though your QSD is on receive, barring an active device in the receive path between the QSD and antenna, the LO leakage will be there. You are essentially transmitting a low-level signal to the antenna through the QSD switch. One reason why QSD/QSE rigs may have a problem at multi-rig operations. The good news is that any transmit amplifiers are not in circuit during receive, so the leakage is at a MUCH lower amplitude. I can easily hear the LO from the SDR rig on my old Icom 720, if I tune slightly off the SDR's frequency (Rx freq plus or minus I/Q IF freq). I was reminded of this just yesterday, while I was working with the SDR-1000 on 40, and happened to turn on the Icom (each radio using separate, perpendicular dipoles). Putting an active buffer in between the QSD and the antenna helps here. One common aftermarket board is the Z10000 from Clifton Labs. Not needed as much on the SDR-1000 with the RFE board (it has a receive buffer/preamp), but recommended for SoftRocks and Brainerd receiver boards, especially if they are used to tap an IF of another rig for a panadaptor display. At least I think the above is correct. I'm not in league with the geniuses that show up here (NOT KIDDING!), but I've had a few years experience now. Please correct me where i am wrong.... Thanks for listening. Terry WB4JFI _______________________________________________ FlexRadio Systems Mailing List FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archives: http://www.mail-archive.com/flexradio%40flex-radio.biz/ Knowledge Base: http://kc.flex-radio.com/ Homepage: http://www.flex-radio.com/
