Flex-1500 TX-broadband noise

Kjeld Bülow Thomsen

In the Dubus Magazine 2013-3 Leif Åsbrink, SM5BSZ writes about "Band Pollution from Amateur Transmitters". The conclusion of this article is, that the Flex-1500 is unsuitable for use on the Microwave bands with transverters.

I'm using two Flex-1500 on the Microwave bands and have now for a long time been looking for a response from FlexRadio on this issue. The reffered article has been mentioned on the Flex Yahoo reflector, but without any comments from FlexRadio.

The measurements made by SM5BSZ are easy to reproduce and shows that this is a real problem. The cause of the problem is analyzed and a solution proposed. So my question: Will FlexRadio support the solution to this problem?

I'm avare of several Microwave stations that has put their Flex-1500's aside as they do not want to pollute the bands with broadband noise.

The article can be found here: http://www.sm5bsz.com/dynrange/dubus313.pdf

Vy 73 de Kjeld/OZ1FF

http://www.oz1ff.dk

Tim - W4TME

Kjeld - Thanks for your question.  Here is our reply to Leif's atricle.

1. FlexRadio is very much aware of Leif SM5BSZ's report and his ongoing concerns regarding FlexRadio's Direct Conversion IQ-based software defined radios.  He has shared his concerns with us and in public for several years.

2. The "noise" he refers to is an artifact of the Delta-Sigma CODECs used to generate the baseband signals.  These CODECs were originally designed for the audio domain and attain very high dynamic range by modulating the noise outside of the audio passband.  In this case, it's approximately 300 kHz away from DC.  This noise would appear to look similar to the transmit phase noise of most all transmitters.

3. We believe Leif's measurements are basically accurate and have verified them in our lab using our test equipment.

4. Using Leif's own measurements, the worst case signal generated is > 82dBc, or 82dB BELOW the carrier.  The FCC and CE require all spurious and unintentional signals to be at least 43dBc.  We are 40dB BELOW the legal requirements so there is no issue there.  At 5 Watts full scale output for the FLEX-1500, the resulting signal is 0.0316 microwatts!  Even if you drive an amplifier to full legal limit with your FLEX-1500, the result would only be 10 microwatts.
 
5. While measurable using sophisticated test equipment, we do not believe that under normal "real-world" operating conditions this noise will ever be perceived by other operators on the band - other than those hams within site of your tower..... On a FLEX-5000A driving a full-legal limit PA, If the 5000A is transmitting on 14.349USB, a person trying to work CW at 14.001 a 1 S-unit rise in the noise floor at about a mile away.... BUT....  Only if you are in a quiet rural location that the environmental noise floor doesn't obscure it.

6. Time marches on and so does technology.   Our new FLEX-6000 Series uses Direct Up Conversion technology for generating the transmitted signal.  Because it's DUC, the only thing you need to worry about is the Phase Noise of the master oscillator.  In our case, the oscillator is better than -140dBc/Hz at 2kHZ from the carrier!  At > 50kHz from the carrier its better than -150dBc/Hz.  Very quiet indeed!  This just isn't a problem on the new radios.

The bottom line is yes, there is a lab measurable noise **** at about 350 kHz on the FLEX-1500, FLEX-3000, and FLEX-5000.  No, we don't believe it's a problem in the real world unless you are trying to run Multi-Multi or a stack of FLEX-3000/5000s at Field Day on the same band.  The new FLEX-6000 Signature Series employs a DUC transmitter and does not exhibit this noise.