WSJT-X db readings

WSJT 65 (and variants) Discussion
Post Reply
VK2FAK
Forum Diehard
Posts: 919
Joined: Thu Apr 15, 2010 6:28 pm

WSJT-X db readings

Post by VK2FAK » Mon Mar 03, 2014 4:52 pm

Hi all....


Listening to JT65 using WSJT-X....and was thinking what is the relevance of the db reading...?

I had a local station coming in at 20 over on the S-meter and the software was giving it a -1 db, the station being contacted showed nothing on my S meter and was given a -3 db on the software.......changing nothing, as in audio levels and such.....why that small of a difference..?

John
Regards

John Vk2FAK

VK2OMD
Forum Diehard
Posts: 1063
Joined: Tue Oct 10, 2006 8:34 am
Contact:

Re: WSJT-X db readings

Post by VK2OMD » Mon Mar 03, 2014 5:25 pm

VK2FAK wrote:Hi all....


Listening to JT65 using WSJT-X....and was thinking what is the relevance of the db reading...?

I had a local station coming in at 20 over on the S-meter and the software was giving it a -1 db, the station being contacted showed nothing on my S meter and was given a -3 db on the software.......changing nothing, as in audio levels and such.....why that small of a difference..?

John
You didn't mention in your post that the dB reading (yes, capital B because it stands for a man's name) is Signal to Noise Ratio, SNR in JT nomenclature, S/N also.

So, it doesn't directly measure signal strength in absolute terms, it cannot do that when it is connected to the audio end of the receiver.

What you might not expect is that a very strong signal (20 over on the meter) has such poor SNR... in the first instance, I would look for overload of your PC sound card or the like. If that clips the audio signal, it will create lots of harmonic and IMD products, and there is your noise increase.

Owen

PS: I should mention here that JT software tends to give S/N wrt noise in 2500Hz bandwidth, though it will be typically measured in much smaller bandwidth chunks, averaged and extrapolated to 2500Hz.
Last edited by VK2OMD on Tue Mar 04, 2014 7:35 am, edited 1 time in total.

VK2FAK
Forum Diehard
Posts: 919
Joined: Thu Apr 15, 2010 6:28 pm

Re: WSJT-X db readings

Post by VK2FAK » Mon Mar 03, 2014 6:31 pm

Hi all..

Interesting point Owen.....I have tried everything I can to change the dB value of the strong signal but it remains at -1....can't be overloading the PC audio input now.....

even to the point now where I see no signals on the waterfall besides that one local station.....but am still getting plenty of decodes from other stations...some at levels of -4 and -7 that are not visible...

John
Regards

John Vk2FAK

User avatar
VK3ZAZ
Forum Diehard
Posts: 800
Joined: Sat Dec 18, 2010 1:46 pm
Location: Hamilton Victoria Australia

Re: WSJT-X db readings

Post by VK3ZAZ » Mon Mar 03, 2014 6:38 pm

You have an issue

I can tune into RMV locally and get 13dB

If I disconnect aerial and make rx noisy it drops

Possibly not by any dB ratio

Absolute values seem tp be -29 -28 dB

But if I see those I am using a IEE preamp and NF of less than 0.2db and gain

I cant hear a -29dB signal I can see it

I can hear a -1 db signal

Just

I have no idea what it means but its not directly related to s/s IMO

More SNR which depends on your setup.

I can certainly print out what I cant hear often

2014-03-03 03:02 VK3ZAZ 50.294484 -29 0 QF12ag +43 19.953 VK5KAA PF94iw 421 262
2014-03-03 02:24 VK3RMV 50.294470 +12 0 QF12ah +37 5.012 VK3ZAZ QF12ag 5 3
2014-03-03 01:48 VK3RMV 50.294466 +13 0 QF12ah +37 5.012 VK3ZAZ QF12ag 5 3
2014-03-02 21:56 VK3ZAZ 50.294490 -29 0 QF12ag +43 19.953 VK5KAA PF94iw 421 262
2014-03-02 21:08 VK3ZAZ 50.294485 -27 0 QF12ag +43 19.953 VK2KRR QF34mr
Tread your own path :om:

User avatar
VK3ZAZ
Forum Diehard
Posts: 800
Joined: Sat Dec 18, 2010 1:46 pm
Location: Hamilton Victoria Australia

Re: WSJT-X db readings

Post by VK3ZAZ » Mon Mar 03, 2014 6:46 pm

Signal levels reported by K1JT for each WSPR decode are the measured ratio of signal power to average noise power, scaled to a reference bandwidth of 2500 Hz. Within broad limits, this ratio is independent of the measured Rx noise level.
Tread your own path :om:

VK2FAK
Forum Diehard
Posts: 919
Joined: Thu Apr 15, 2010 6:28 pm

Re: WSJT-X db readings

Post by VK2FAK » Mon Mar 03, 2014 6:58 pm

Hi all...

Maybe its my set-up.....but wsjt-x does not seem to show dB levels on PSK-reporter......or have I not clicked something...

John
Regards

John Vk2FAK

User avatar
ZL1RS
Forum Diehard
Posts: 493
Joined: Fri Feb 08, 2008 10:29 pm
Location: Bay of Islands
Contact:

Re: WSJT-X db readings

Post by ZL1RS » Mon Mar 03, 2014 7:23 pm

The software is reacting to the audio fed into the soundcard. The audio level from the receiver on stronger signals is affected by the AGC threshold and characteristic of the individual receiver ... the AGC creates a a sort of 'compression' effect on strong signals.

(Maybe I didn't explain that very well, but perhaps someone will put it better ... I keep thinking 'audio dynamic range', but that might get misconstrued).

73,
Bob, ZL1RS in the Bay of Islands at RF64vs
http://www.qsl.net/zl1rs

VK2FAK
Forum Diehard
Posts: 919
Joined: Thu Apr 15, 2010 6:28 pm

Re: WSJT-X db readings

Post by VK2FAK » Mon Mar 03, 2014 7:29 pm

Hi all...

That is fair enough....but what happens when you reduce input by say turning right back the RF gain...

But it gets back to the point, if this applies to every radio, there is nothing at all consistent about the dB readings..

John
Regards

John Vk2FAK

VK2OMD
Forum Diehard
Posts: 1063
Joined: Tue Oct 10, 2006 8:34 am
Contact:

Re: WSJT-X db readings

Post by VK2OMD » Mon Mar 03, 2014 8:51 pm

VK2FAK wrote:Hi all...

That is fair enough....but what happens when you reduce input by say turning right back the RF gain...

But it gets back to the point, if this applies to every radio, there is nothing at all consistent about the dB readings..

John
It might help your understanding if you used something simple like Spectran to look at a steady carrier, and did things like turn the preamp on/off, swith the attenuator in/ou, adjust the RF gain by a little, but a lot and observe what happens to the difference in dB between the signal and the adjacent noise (grass) on Spectran... that difference is the S/N ratio. Try this for signals with high and with low S/N ratio... what is different?

The S/N ratio you observe depends on signal strength you receive (antenna gain plays into that), and the background noise (your environment affects that a fair bit, but antenna efficiency also plays into it.

Does an inefficient antenna necessarily reduce S/N significantly?

Here is a little exercise to stimulate some grey cells.

http://owenduffy.net/blog/wp-content/up ... rab001.jpg

Above is a screen grab from W4HBK of me last Saturday. (Click on it for a larger image.)

What is the S/N ratio (use the red trace)?

The effective noise bandwidth of the spectrum buckets is 0.25Hz, what would the S/N ratio be if bandwidth was 500Hz?

If CW can be copied by ear with S/N down to about -6dB in 500Hz bandwidth, how much would the signal need to increase to be copyable by ear? Answers down the page.

Owen
























What is the S/N ratio (use the red trace)? 15dB

The effective noise bandwidth of the spectrum buckets is 0.25Hz, what would the S/N ratio be if bandwidth was 500Hz? -18dB

If CW can be copied by ear with S/N down to about -6dB in 500Hz bandwidth, how much would the signal need to increase to be copyable by ear? +12dB, 16x (80W since the tx was 5W).
Last edited by VK2OMD on Wed Mar 05, 2014 7:18 am, edited 2 times in total.

VK2OMD
Forum Diehard
Posts: 1063
Joined: Tue Oct 10, 2006 8:34 am
Contact:

Re: WSJT-X db readings

Post by VK2OMD » Tue Mar 04, 2014 8:54 am

VK2OMD wrote: Does an inefficient antenna necessarily reduce S/N significantly?
It is trick question... is the inefficient antenna at the tx or rx?

Receiver:

The answer is no . If you tune in a station on 80m using an efficient antenna, preamps on, RF Gain max, then insert the attenuator that most radios offer (somewhere 10-20dB), it probably doesn't degrade the S/N ratio enough to hear the difference, in fact it might even improve.

Try it! Hint: cover the S meter, you might be influenced by its readings... just concentrate on what you hear, readability.

So, this demonstrates that under some conditions, S/N might not be very sensitive to rx antenna efficiency, or receiver Noise FIgure or sensitivity.

Transmitter:

The answer is yes. An inefficient antenna will reduce the radiation power which reduces S in S/N.

Owen

PS: Efficiency is well defined in engineering as PowerOut/PowerIn, and an efficient antenna system is one where most of the RF energy captured by the antenna is delivered to the receiver, or most of the transmitter power is radiated... ie it has low losses.

Post Reply