ADC overload myths debunked

Gerald-K5SDR

Note that the lab measurement that will tell you performance in this area is Reciprocal Mixing Dynamic Range (RMDR).  This will usually be the limit of performance for single tone blocking dynamic range in most all modern ham receivers that I am aware of no matter the architecture.  If RMDR is less than BDR, noise will dominate to cover up weak signals.  You can directly compare these measurements in QST reviews at 2 kHz, 5 kHz and 20 kHz offsets from the carrier.  RMDR for the FLEX-6500 and FLEX-6700 is 116/117/124 dB respectively.  For the FLEX-6300 it is 116/117/121 dB respectively.

Ken - NM9P

I have really enjoyed this thread.  It has been very informative, and it adds to my appreciation of my 6500.  I really appreciate that several of the staff at FRS, including Lead Engineers, Customer Experience Managers, and even the President of the company are willing to take the time to converse with their end-users.  How cool is that?  Awesome! 

Bill -VA3WTB

Oh, don't all the other company's talk to their customers directly? LOL

KD4HSO

Wow, thanks for the detailed responses guys; your are true engineer's engineers!  I have lost hair over TCXO for narrow band microwave comms for just what you stated.  Ended up characterizing several manufacturers SMT TCXO for phase noise below 10 Hz; they never publish that and they are not all equal, and some use digital vs analog temp comp (been burned by digital with phase hits over temperature).

I assume you'll be using ultra-low phase noise optical oscillators for the 7000 series ;-)

http://www.oewaves.com/technology

KD4HSO

To be fair, Elecraft does, but that's all I know of.

Peter K1PGV

What Ken said!

I don't even *understand* the details of Steve and Gerald's posts, and I *still* enjoyed reading them.  As Ken so aptly said "it adds to my appreciation"... and I love hearing about the factors that are germane to areas of engineering other than my own (OS software). 

Thanks guys!

Peter
K1PGV

EA4GLI

Well... maybe the others do, but in Japanese and we just don't know about it....

Walt - KZ1F

To be sure, more Japanese can speak fluent English, or American in Sarah Palin's world, than there are Americans that can speak Japanese. However, if the goal is to have no one besmirch FRS or their products, people here shouldn't be casting aspersions on other vendors or their products.

Bob - W7KWS -

Steve, Your comments on ADC overload are helpful & I've watched the video of your Chicago TAPPR presentation. The simulated signal traces that Gerald did added a lot of clarity. However, decimation remains a vague mathimatical concept for many of us. Could you please give us one of your very good, non-math explainations on how decimation equates to more ADC sample bits? This would go a long way in closing the loop for those I'm discussing this topic with. Thanks!

Steve-N5AC

Bob, ok here goes:

Suppose that you have a device that samples at 80sps (samples a second) and is an 8-bit device.  Now let's say that we want to decimate down to 40sps.  We take samples 1 and 2 and turn them into a new sample 1A.  Let's say the samples are each equal to 1 (binary 00000001).  In this case, our sample 1A could be set to 1 as well.  Now we look at samples 3 and 4 and they are equal to 1 (binary 00000001) and 2 (binary 00000010).  In this case, we have to pick either 1 or 2 or something in between.  So we decide that we will add one bit BELOW the decimal point and set this bit to a 1.  The resulting binary representation would be 00000001.1.  The bit just to the left of the decimal point is the "1s" bit and the bit just to the right is the "1/2s bit" so the value of this is 1.5.

This continues and the net effect is that we pickup a 1/2-bit of resolution each time we decimate, but the sampling rate is lower by a factor of two.  After this first decimation, we now need to carry 9-bits instead of 8 and we've added 3dB to our dynamic range because we can now instantaneously represent 8.5-bits of information instead of 8.  In the FLEX-6000, we pickup about 56dB of dynamic range going from 245.76Msps to 24ksps.

Does that help?

W1RE

Excellent explanation Steve.

Thanks!

EA4GLI

Walt, 
Was you reply directed at my statement? Because if that is the case I think you misunderstood my point.

I meant to literally say what I wrote, that maybe there are forums in Japanese from Yaesu, Icom and Kenwood where the owners "talk" to hams and we just don't know about it because of the language barrier. I know for a fact that a lot of people in Spain are not very aware of the ins and outs of SDR because there is very little information in Spanish. many people can understand basic English, but when it gets technical it is difficult to follow if it is not your mother tongue.

Tim - W4TME

Good point.  Those of us who are monolingual native English speakers sometimes miss this subtle, but important point.

Bob - W7KWS -

Thanks Steve. I'll probably look to you later as I try to determine the impact, if any, of the samples being taken at different instants & then combined. You have given us plenty to chew on for the time being. Hopefully more of this will sink in before I see you in Dayton. Best regards!

Barry N1EU

Oy:  http://mailman.qth.net/pipermail/elecraft/2016-March/229303.html

Dan -- KC4GO

W4TV isn't more that 25 miles from me and the only band he might get -130 signals is 10 or 6 meters 
 "Try copying a -130 dBm signal among multiple -30 to -50 dBm (S9+20 to S9+40 dB) amateur signals on 160 with a dozen AM signals also 
present at -30 dBm to 0 dBm at the receiver input.:" is in his dreams  This morning it's -100 with -10dB att. Flex 6500 Non resonant antennas.
When the band is open (160 the noise it much higher)   

Steve-N5AC

He's literally wrong "on every count."  How can the guy with mathematical models, actual lab tests and on-air results be accused of hand waving while the guy saying "I doubt" presents himself as the expert?  The FLEX-6700 is literally top rated by everyone that has reviewed it (including all he's mentioned).  I predict he will still be saying direct sampling doesn't work long after all HF receivers are direct sampling and the companies holding on to superheterodyne are out of business.

It's also interesting to me that he's not disclosed his conflict of interest (I trust my interest is well known): he sells the MK2R that glues multiple superhet radios together for SO2R which, coincidentally, is no longer required with the FLEX-6700 that can do SO2R with a single radio.  The "before and after" pictures of stations that have removed two superhet radios with all the "glue" and replaced with a single FLEX-6700 is striking.  The resulting station costs roughly half as much, is simpler to operate, looks cleaner, is easier to configure and has better performance.

QED does not stand for "keep arguing if you don't believe the proof."

Steve-N5AC

btw: You totally beat me to the deal of the century on the 6 radios.  Let me know if you actually receive them ;-)

EA4GLI

Are you guys talking about the ones from the KB1 in MA?

Barry N1EU

yup  ;-)

EA4GLI

Looks like a fake to me.... let me know how it pans out.

Barry N1EU

Sal, we're just having fun

EA4GLI

Sergey R5AU

LOL

Dan -- KC4GO

I got my long wire to tune to 1.850 and got a noise level of -67 S9+ (From 6500 S-Meter) in Casselberry Fl 12.64 miles from Geneva, Fl Location of W4TV

Bob-N4HY

I know the OEWaves people. They are friends of mine. Lute Maleki and Alfred Pais have done a great job. To finish the work of applying the whispering gallery mode resonators to Flex, well, ithe might currently cost about 1/3 of the entire value of Flex just for NRE and no one could afford the radio if the NRE were donated!

Walt - KZ1F

Sorry Sal. No, that was not directed at you. I just now saw your question to me. My comment was more an observation that, I believe this country is unique on the planet in the percentage of its population unable to converse in any native tongue other than 'American', to once again jab at Sarah. I try to but I don't get to converse much to German folk. I do try to use Italian when conversing with Enzo though.

W9OY

W4TV I think owns Microham and makes SO2R "glue" so when he talks about "talking book" it is actually he who is "talking his book".  He lives about 15 miles from me and I've talked to the guy and he is definitely a legend in his own mind and rabid on the K3.  

As far as 160 noise goes, I get -103 dBm on my dipole and -108 dBm on my vertical with WBN turned on but I live in a area of very low noise and my neighbor to the east is the Atlantic ocean.  His QTH is pretty rural as well so he may also have a pretty good noise floor, but I'm not sure what he is running for antennas

Craig K9CT

I used a Flex 6700 in the ARRL 160 contest. I did very well...the only people ahead were multiop or east of me by at least 800 miles. I did something very unusual in that I did SO2R on one band and had the second receiver on the band for copying while I transmitted on the run frequency. Full duplex operation. The RX antenna was 1100 feet from the TX antenna. I had another RX antenna on the run slice and that antenna is only 300 feet away...and it is not disconnected while TXing. At no time did I have IMD or burn up my radio. I don't have a noisy location but the band has way more noise than necessary for using a preamp. I insert attenuation all the time to max the dynamic range.

Next week I am selling every one of my microHam MKIIs. I don't need them any more.  Poor Joe....

73, Craig K9CT

 

W9OY

Craig your comment on the contest is telling regarding the future of SO2R! 

I also use the built in attenuation in my Flex for 160  Worked P40L tonight barefoot and he's S9 here, and I can hear 3XY1T so the band is pretty good

73