S Meter Reading after Service

James Whiteway

Burt, if you're that unhappy with the Flex, maybe the K4 will end up being what you want.
Being a teacher I would have thought you would be more open minded and receptive to new things.
While I understand how you feel about having an S Meter that reads zero with no signal present, at the same time, I also know that the technology behind SDR radios is radically different than the radios I have used in the past.( and still use some I cannot seem to let go of)
But, technology moves on and with it, things like signal measurement change with it. It doesn't make it less enjoyable for me. I like the challenge of learning new things. Keeps my 68 year old mind active.
James
WD5GWY Oh, and I speak for myself. As I don't presume to know what other hams want or need.

VK7WH Winston

I totally agree with you, James.

I’ve also got an old mind,I will be 80yo next year, and one of the many reasons I purchased my 6700 was to help keep my mind active. I am also very happy with an S Meter that is correctly calibrated.

Good DX and 73 Winston.

Bill -VA3WTB

What about the RF energy in the rceiver? were did that go? It's there!!
But of course with the antenna connected that inboard RF is not shown.

I know you had a back and forth discussion with Gerald, and he said they are not going to break the meter to make a few happy.

N2WQ

This debate sounds very much like how some drivers blindly follow their GPS, never looking out the window, and not asking themselves the basic question “Does this make sense? Am I going where I want to go rather than where the GPS is taking me?” Nobody here is arguing the technical basics of why the S-meter reads whatever number it does when the antenna input is shorted. The question some of us are debating is if it should. Just because the AD reads its own noise level it doesn’t mean that we the humans should take it at face value and not apply some common sense. The S-meter is meant to measure signal strength, not internal noise. For the record, I know my math and DSP algos, having created several DSP platforms, including my favorite system Hera for extracting human voice from a very noisy environment. Think extracting single person’s voice from the crowd at a large cocktail party. In 1987! I am not intimidated by the bullies who casually throw techie jargon left and right, trying to impose their personal views as the Bible.

Michael N3LI

So here's the basic question - Why is Flex completely wrong with their implementation of an S-Meter? 

I'll never be as smart a guy as you, but to my admittedly poor intellect and technical acumen, a signal strength meter is a rather poor, almost useless example of a measurement instrument. Are you as adamant about the implementations of S-Meters in Japanese equipment? The measurements show their accuracy is nil. Often calibrated only at S9. Wildly different levels for  differnet S-Units. This is not just me parroting something I've heard - this is actual measurements by me. Enough to convince me that the S-Meter on legacy radios is pretty worthless. 

About the only function they have is reporting zero on a dummy load, like some people demand. 

I look at S-Units as an extremely rough, innacurate and often wildly wrong talking points. About as useful  as a contest exchange of 5 of 9. 

So I use dbm, or if pressed, I use as narrow a bandwidth as I can and get an S-number that agrees with the dbm scale. I want to see the noise that is present. I want to know what level it is. 

How do we subtract the noise from the instrument, but have the local RF noise show up yet maintain accuracy?

Tell me why exactly Flex and the rest of us are wrong. Tell me why a legacy radio that adjusts the higher noise floor to zero is correct/

I hope you don't consider disagreement as bullying. If so, let me know, and I'll stop.

Dawg Fan

Exactly my comment....I use my dummy load brain and my ears...I hardly ever look at an S meter...Don't need it.

Dawg Fan

OMG,  just pick an S number and use it, the importance of the signal report is not near as important as making the contact.....It doesn't matter if your work history is an an astro physicist instead of some electronic technician, you are placing way way too importance on a signal report...**** IN THE SEAT TIME AS A HAM MEANS MORE THAN SOME WORK HISTORY....I don't believe there is anything wrong with the radio....Just give them an S7 or S9, pick one and move on...

KY6LA_Howard

@N2WO

Why is it so difficult for you to understand the COMMON SENSE of the Frozen Thermometer Paradox

Mercury Freezes at -38F.
If you put a mercury thermometer in -50F then it will still read -38F

The Same is true for an S-Meter... 
If the radio (as you can see from the charts below) can only hear -115dBm or S2 then common sense should tell you that it should not display -127dBm or S0


Further COMMON SENSE... S0 does NOT MEAN the total absence of signal but it means that the signal level is -127dBm... as agreed by the ITU international Definition.

So if shorting the antenna would give you S0...it means that you have -127dBm signal

But clearly you do not...

A truly accurate S Meter is definitely not a market killer for Flex Radio as their sales prove... so the market has decided

Like I predicted many years ago that every manufacturer - even the Japanese would ultimately make a SDR

I will also predict that now that they have the tools, then every manufacture will also start producing accurate S Meter instead of Fake News S Meters

Brian Denley KB1VBF

But the real issue is this:  ALL new SDR xcvrs available now are basically spectrum analyzers.  That’s not going to change.  They comply with the published standard for S meters.  The older xcvrs do not!  So this in not a Flex issue: Elecraft, ANAN, Flex, Perseus, etc. all comply with that  current standard.  So what are they to do?  Not comply with the standard?  There is no going back.  What would we go back to?  I guarantee you that the new Elecraft K4 will match Flex on S meter readings.

Brian KB1VBF

N2WQ

I think there are two parallel themes here that are getting mixed and contribute to some of the confusion.

The first one is that regardless of the temperature, S5 for a shorted input is just wrong. -97 dbm is a lot of signal that can’t be attributed to the frozen thermometer.

This leads to the second debate, which is what the S-meter should be reading. Or more precisely, what should S0 be. This is where the debate becomes philosophical.

Rob Sherwood lists the 6700 having a noise floor of -118 dbm with preamp off. One of the arguments here is because the radio can’t hear anything weaker than -118 dbm the S meter should not be showing anything less than the S value matching -118 dbm.

However, as the preamp is dialed up 10, 20, or 30 dB, the 6700’s noise floor drops and the radio can in fact “hear” -127 dbm or S0. Therefore, the S meter should be able to read S0.

Personally I don’t use the S meter and don’t care about its reading. What bugs me is when those who express a reasonable contrarian views get crucified for not following the party line. A good, civilized debate is healthy, but not when it turns into name calling.

Technology is amazing and I am thrilled how it creates innovation and competition in the ham market.

Burt Fisher

I don't get why anyone would not be friends over an opinion related to ham radio, in fact even an opinion related to politics. I have a friend whose politics are way to the other side of mine. We debate the issues, not the person. Many times I see the accusation of troll, simply for expressing a different opinion on many different subjects. 
I find Flex users to be the most intellectual of hams but somewhat stuck in their beliefs.,

Rolf Auras

After reading this interesting discussion I would like to check my understanding without diving in a pool of bins: When listening to the phones I hear three types of noise: 1. the signal 2. atmospheric noise 3. receiver internal noise A Flexradio shows all three on its S-meter, e.g. an Elecraft only 1 and 2. Is this view acceptable? Rolf, DJ1WT

Brian Denley KB1VBF

Elecraft and Flex S meter reading are the same.

Brian Denley KB1VBF

So is ANAN

Brian Denley KB1VBF

 Rob Sherwood on “S-Meter Accuracy” 6-OCT-2016 

While no Japanese rig has a properly tracking S meter, the following do: 

  Flex 6000 series 

 Apache ANAN 

 Most K3 and K3S, though occasionally an individual radio cannot be calibrated properly. This includes the option to have the reading correct regardless of preamp or attenuator setting. 

 Orion II once Ten-Tec updated software to allow S meter calibration. Orion II does not correct for preamp or attenuator selection. 

 Eagle is good, except the S meter is microscopic and I think it quits at S9+30 dB. The Eagle reads correctly regardless of preamp or attenuator selection. 

 Hilberling S meter is quite good. 

 ADAT ADT-200A is very good but ergonomics in general are absurd. 

 Elad DUO S meter is outstanding 

 Perseus S meter is outstanding.

Michael N3LI

Seems like some pretty good company our Flex's find themselves in. 8^)

Michael N3LI

Hi Rolf,   "diving into a pool of bins"  Now that is the best statement in this whole thread!

How about if we look at it as the Flex radio determining Signal strength at the noise floor by virtue of bandwidth. If you are using a wider bandwidth, like SSB, it is seeing more peaks in the noise, which tends to make the reading higher. When reduced to a much narrower bandwidth like say CW 50 Hz, there are a lot less peaks to measure.

Then when a signal comes in that occupies the bandwidth, be it CW or an SSB width you are listening to,  you get a good reading of signal strength.

I managed to get through that without invoking the "B" word! 8^)

Rolf Auras

Brian,

as Joseph and other reported their Flex show S 5/6 and no antenna connected. The Elecrafts show S 0 under the comparable conditions...

Bob Gerzoff

I'm a statistician so I have to ask:  If the peaks (and valleys)  are a random phenomena due to noise then from my perspective, the expected (average) noise level in any bandwidth would be same.   A wider bandwidth would see more signals, and hence have a higher reading.  So I'm thinking the higher readings are  due to some type of sum of the signals within in the passband.  Am I thinking right?
73, Bob
WK2Y

Geoff AB6BT

Here's some info regarding noise and some of the different types. If the noise is white then there is equal power in any specific bandwidth, say 1 kHz whereas pink noise has equal power per octave.

https://en.wikipedia.org/wiki/Colors_of_noise

Brian Denley KB1VBF

Rolf:
Not true.  Look at the xcvr s-meter evaluations by Rob Sherwood.  NOW that doesn’t mean that Joseph’s radio is OK.  It would have been great if Flex had given him their measurement when they returned his xcvr.  Other than that, comparison with another calibrated SDR xcvr or trying a new location are the only ways I can think of to isolate the problem.  Where did Joseph say he lived?

Brian. KB1VBF

k3Tim

Bob,

You are correct.  For example, a 500 Hz BW measurement taken at the low end of the 40 meter band should show a similar reading reading at the high end (assuming atmospheric noise only in the passband).  As the BW increases the amount of noise within the passband also increases. The energy within the passband is used to find the PSD. That is, the FFT bin are used in the summation to find power spectral density. This power can be converted to a dBm reading or voltage (based on 50 ohms).  An interesting link:

https://en.wikipedia.org/wiki/Spectral_density

That's my take on the S-Meter inner-workings. 

_..--
k3Tim/7

Rolf Auras

Brian,

I think you will agree, that a S 5 (no antenna) needs an explanation. If Flex confirms that this is correct and my explanation is wrong (internal noise) we have to ask ourselves how to deal with it. 

The relationship of bandwidth and noise is understood however a S 5 with no antenna is an useless information.

Rolf, DJ1WT

Mike-VA3MW

If you look at the original notes from both Steve Hicks and Gerald, it has been explained.  

Lasse Moell

With no antenna attached, turn on MAX pre-amp gain and you WILL see the S-meter drop! The explanation why has been given over and over again.
To get really low reading, decrease the IF BW to 50 Hz. I guess it might be hard to grasp the relation between receiver noise factor, IF filter BW and S-meter reading.

Rolf Auras

Taking Josephs posts I'm convinced he tested all these settings and he did NOT see the S-meter drop. Or does his Flex behave differently? Joseph please give us a comment on this. Thanks, Rolf, DJ1WT

Bob Gerzoff

Oh no.  Now I have to remember my calculus. ;-)  Thanks for the info and thanks to Geoff too.
73

Michael N3LI

Listening to radios with no antenna within itself is pretty useless information.

Now the idea that a S-meter should read S-0 when on a dummy load or shorted isn't correct unless you place some fudge factor that adjusts the ambient dbm to the value of 0. When things are quiet at my place and on a dummy load, I'm at around -123 dbm or so. WIthout the load and on an antenna, I'm around -120 dbm

I can get a reading where the S-number matches the dbm and S-number agree when the radio is on a dummy load. Which is around .5 to 1. This determines that my Flex is calibrated and operating accurately.

But I have trouble wrapping my head around the concept that we not only need to have purposefully inaccurate S-Meters, but use an incorrect way to measure S-Value on the Flex in order to say that the meters are inaccurate.

Michael N3LI

Looking at the posts, he was complaining about the readings on SSB, which indicates the radio was functioning well.

When he was using a narrow bandwidth, the S-level dropped appropriately.

Read Brian's posts on Sherwood's comments on meter accuracy. Flex is in the elite class of radios, and no Japanese radios track properly.

I've yet to see a good reason for dropping the performance. Probably the best is Burt's comment that "Joe Ham" wants it that way. Still not a good reason. These things are Ferraris, it makes no sense to try to turn them into Toyota Corollas.