S meter with version 2.3.7

  • 2
  • Problem
  • Updated 7 months ago
  • Solved
  • (Edited)
After installing Version 2.3.7 in my 6600M, the S Meter and pan display is out of calibration.  
Ex:  a -62dbm on the pan display shows an S9+10db on the bar graph.

The bar graph never indicates below S6 without an antenna connected. 

I performed multiple cold boots as well as a factory reset boot with no change.  

One of the impressive features of the Flex is their superb accurate display. 

 Finally, I had to  downgrade to Version 2.2.8 which returned calibration to normal.  

I was wondering if anyone else has experienced this problem.  

73,  Paul
Photo of Paul,  KB3ML

Paul, KB3ML

  • 16 Posts
  • 6 Reply Likes

Posted 8 months ago

  • 2
Photo of Richard

Richard

  • 93 Posts
  • 11 Reply Likes
Paul....  I went and checked it out and yes there is a strong hysteresis but the S Level does change on my 6700 with the lastest software.  Thanks for the heads up. I thought at firs that it was "Stuck" but no it seems to be working.
73,Richard
Photo of W8WD

W8WD

  • 23 Posts
  • 2 Reply Likes
Paul,
I had a similar observation. With a dummy load connected to the radio the spectrum scope at the tuned frequency was about -125dbm but the S-meter is reading S5.

Brian/W8WD
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 9153 Posts
  • 3493 Reply Likes
Making the RX filter width smaller will show the S meter reading decrease because as Bill indicated, the S meter is reading the integrated RF power within the receivers defined passband.
Photo of Edward Armstrong-WB2K

Edward Armstrong-WB2K

  • 25 Posts
  • 1 Reply Like
Hi Tim I noticed this last night and V2.37.110 not only does the S meter read higher the receiver is extremely noisy I tried playing with the filters does not get rid of the noise I have never seen this in any other release I took screenshots and I will send them to you.you will see the noise floor stays the same but the S meter will read higher on the new release with a lot more noise hopefully they will fix this soon because there is no way I can use the new release the way it is with the new release the 6600 is the noisiest receiver I have much noisier than even my flex 5000 So I went back to v2.28.109 and the radio is quiet again.
Photo of W8WD

W8WD

  • 23 Posts
  • 2 Reply Likes
Tim, Understood, thank you for your comments. Using a HP8656B generator I found that the panadapter is accurate, the reported S-meter reading don't follow the IARU table. The following is measurement data for your reference.




Thanks!
Brian/W8WD
Photo of Jay / NO5J

Jay / NO5J

  • 1621 Posts
  • 235 Reply Likes
Brian

Did you get those S meter reading using a 500 Hz filter bandwidth in SmartSDR?
IARU specifies a 500 Hz filter width. wider filters would read higher S units.

Just curious.

#FlexRadio IRC chat

   73, Jay - NO5J

Photo of W8WD

W8WD

  • 23 Posts
  • 2 Reply Likes
Jay, Thank you for your feedback.  I ran the test with 400HZ bandwidth using my Maestro, I will re-run the test at 500Hz and publish the results.

Brian/W8WD
Photo of Bill -VA3WTB

Bill -VA3WTB

  • 3178 Posts
  • 706 Reply Likes
As far as the S meter is concerned, keep in mind that the meter reflects the receiver bins being sampled all the time. Even without an antenna connected the bins are sampled within the radio and show a reading on the S meter.

I should also add that rolling back to an older version unless there is something preventing the radio from working is a BAD idea. It is always best to keep with the latest version.
(Edited)
Photo of Paul,  KB3ML

Paul, KB3ML

  • 16 Posts
  • 6 Reply Likes
This may be an easier way to explain what I'm seeing.  Yes, changing filter band width will change the noise floor but not calibration.  

With your Service Monitor, apply a -73dbm (50uv) level to the radio.  
        Software Version 2.2.8 shows:
        S9 and -73dbm on S meter
        -73dbm on pan display

        Software Version 2.3.7 shows:
        S9 + 10db and -62db
        -80dbm on the pan display

73,  Paul

        
Photo of Wayne

Wayne

  • 614 Posts
  • 86 Reply Likes
Good documentation Paul I think you made a good catch there.
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 9153 Posts
  • 3493 Reply Likes
I have consolidated the comments and entered this issue into our bug tracker as issue #6448 for further investigation.
Photo of Jim K4JAF

Jim K4JAF

  • 354 Posts
  • 104 Reply Likes
I see this has been declared "not a problem", has it not been issued a bug tracker number?
I checked my 6400M with the Elecraft XG1 signal generator and S9, 50 mV was right on.  
Of course using 2.3.7 version.  Only a 6600 issue?
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 9153 Posts
  • 3493 Reply Likes
That was an oversight that has been corrected
Photo of Rick  WN2C

Rick WN2C

  • 280 Posts
  • 33 Reply Likes
My 6400M seems to be right on too. I don't have signal source to verify but I do not see it on mine.Running stand alone. I wonder if it is the same on the second SCU?
(Edited)
Photo of Larry Benoit

Larry Benoit

  • 72 Posts
  • 17 Reply Likes
My 6600 was just returned with the PEN update. I ran the radio this morning (very poor conditions) with v2.3.7 and v2.1.33 on 40 meters just a few minutes apart. My noise floor is usually around S6 on an elevated 40m vertical.   With v2.3.7 and v.2.1.33 the noise floor was approximately -108 dbm on the vertical scale, but on the same frequency with no signal the S units are nearly 9 with v2.3.7 and approximatley 6.5 with v2.1.33.  I also noticed that while operating with v2.3.7 the vertical -dbm scale is at substantial variance (weaker) with the S meter, typically 15-20 -dbm. 



73,
Larry KB1VFU
Photo of Bill Roberts

Bill Roberts

  • 224 Posts
  • 45 Reply Likes
So, is this situation a bug (to be addressed later) or not?  My 6600 with no antenna or a dummy load connected shows an S-meter reading of 6 whereas with version 2.2.8 the reading was 4.  Yes, I have read previous explanation on why an S-4 reading is to be expected, even with no antenna connected.

Life sure has changed for us guys who have been licensed for almost 55 years when standard operating procedure was to disconnect the antenna and set the little pot for S-0.
Photo of Edward Armstrong-WB2K

Edward Armstrong-WB2K

  • 25 Posts
  • 1 Reply Like
Hi I don't usually post a lot here but a few of us have tested 6600s and forgetting about the  S meter and listen to the receiver SET the AGC T to 28 or 30 and listen and then go back to earlier version of smart SDR and you will here the difference 2.3.7 is much noisier you will see with the higher S meter reading comes noise.

I will wait until they fix this do not care a lot about the S meter but from a quiet receiver to what it is now Drives me up the wall like I stated much noisier than even a flex 5000.

Ed   WB2K
Photo of Bill -VA3WTB

Bill -VA3WTB

  • 3178 Posts
  • 706 Reply Likes
Nothing is changed on my 6500.
Photo of Rick  WN2C

Rick WN2C

  • 280 Posts
  • 33 Reply Likes
This seems like a 6600 issue. I wonder if those that have a 6600 have checked both SCUs to see if the noise level is as high on the second SCU.
Photo of Bill -VA3WTB

Bill -VA3WTB

  • 3178 Posts
  • 706 Reply Likes
We will know more soon as Flex is looking into this now.
Photo of bh8asz

bh8asz

  • 1 Post
  • 0 Reply Likes
Follow with interest
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 9153 Posts
  • 3493 Reply Likes
Official Response
This defect #6448, which only effects the FLEX-6600 and FLEX-6600M will be fixed in the next release of SmartSDR.
(Edited)
Photo of Wayne

Wayne

  • 614 Posts
  • 86 Reply Likes
didnt you mean 6600M Tim?
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 9153 Posts
  • 3493 Reply Likes
Yes, it was a typo that has been corrected.
Photo of Bill Roberts

Bill Roberts

  • 224 Posts
  • 45 Reply Likes
As always, a prompt and effective response.  Thank you.

73
Photo of Steve - N5AC

Steve - N5AC, VP Engineering / CTO

  • 1045 Posts
  • 1023 Reply Likes
Actually, this problem is ONLY on the FLEX-6600 and FLEX-6600M, not the FLEX-6400(M)
Photo of Steve - N5AC

Steve - N5AC, VP Engineering / CTO

  • 1045 Posts
  • 1023 Reply Likes
Official Response
Howdy -- here's the deal here: the S-meter is off by several dB in v2.3.7 for only the FLEX-6600 and FLEX-6600M.  We discovered the issue shortly after release and plan to issue an update soon to fix this.  It is only an offset, meaning the s-meter and panadapter will read higher than they actually are (I believe the amount is around 7dB), but it is only an offset in the amplitude and does not affect performance in any way.

Please remember that you cannot compare noise read on the S-meter in the slice receiver and the panadapter noise floor because of the different bandwidths of receivers used in the slice receiver (filter) and the FFT bin bandwidth (unless they are the same based on a zoom level).
Photo of Rex K0KP

Rex K0KP

  • 44 Posts
  • 3 Reply Likes
OK on the issue and the official responses.  I think what is needed is a software switch to allow operator selection of how the S Meter works and responds.  S meters have been around for many decades.  Irregardless of how Flex chooses to implement the S meter, please allow the end user to modify how it operates to make the S meter work as if it were in a traditional analog receiver.  50 uv makes S9, add 30 db of preamp and it should read S9 plus 30, with 0 db of preamp it should read S0 with no antenna connected.  It doesn't matter to me that the S meter circuitry is reading empty FFT bins or not, please make it so the S meter works like it does in a Collins, Kenwood, Icom, or Yaesu receiver.  It is nice to say this radio is the perfect spectrum analyzer and so forth, and that is indeed nice, just incorporate a switch setting so the S meter can also work as in a historical relative indication mode.  Thank you!  Hurrahs please!  ??
Photo of Bill -VA3WTB

Bill -VA3WTB

  • 3178 Posts
  • 706 Reply Likes
What you are asking is for Flex to make a meter to reflect the very inaccurate and vintage meters. So it has to fake what is reads.
Welcome to the high tech of Flex radio. The way the meters work is a reflection of how direct sampling works.
(Edited)