Oscilloscope and Modulation Meter for SDR# MEASURES AM MODULATION ASYMMETRY

[IN3IEX]

Hello all
Why not use a 10$ DVBT dongle for TX monitor?
SDRSharp is a software SDR that can be used to receive from 26 to 1700 MHz with a cheap DVBT dongle.
HF can be received with an upconverter or whith the "direct sampling" mod.
Please note that SDR# can operate with many different receivers.

http://sdrsharp.com/   (there is an installation file)
http://superkuh.com/rtlsdr.html#directsample
http://uk.groups.yahoo.com/group/SDRSharp/
You also need to install Microsoft NET framework 3.5 (already installed in windows 7)

Bobrich wrote a nice oscilloscope plugin for SDR#.
I added three more controls to it.
I attach the modified dll and a screenshot.
Give it a try.
Please rename the .wav file to .zip and unzip it.

To observe SSB patterns or AM envelope just use USB or LSB, select about 20 kHz filter and tune the carrier to filter center.
CAUTION: Please use AGC decay longer than about 200ms to ensure constant receiver gain within the measurement time window. Propagation fading may also affect the measurements.
WARNING: Use latency >=100 ms in the Audio Panel. It is necessary to have enough samples in the audio buffer to properly use the Decimation control.

EDIT: AUTO AM% measurement ADDED.
EDIT: Separate AM% measurements for positive (above envelope average line) and negative (below envelope average line) peaks, two methods available. Positive/Negative peak ratio is also calculated: AR%= 100 * (AM%+ / AM%-).
EDIT: Troubles with the renamed wav file? Go here in the file section:
         http://uk.groups.yahoo.com/group/SDRSharp/messages
EDIT: Show envelope Peaks added

DISCLAIMER: This software is experimental. At this stage of development measurements are not "continuous", only parts of the audio buffers are analyzed, completeness depends on the audio buffer length settings and on the decimation control.
For complete/continuous and/or certified measurements please use a different solution.

Best regards

Giorgio

[N2DTS]

Someday, someone will write or include a mod scope in some of the SDR software that runs on popular sdr receivers like the sdr-iq or qs1r.
Flex made a bodge of it in AM, it showing only the audio, not any carrier...

As it is, running the sdr allows me to see overmod when the signal goes wide..


I wish more AM'ers would use an sdr, at least as a band scope, as people are often off frequency or have screwed up signals.
40 meters seems to have a hole at 7290 at night, and last night I noticed a long qso with various people operating on 7286 with a strong wide broadcast station on 7280 with audio out to 7285. 1 KHz away, when 7290 was open and clear...
Very weak narrow broadcast on 7295 was not bothering anything, its so usefull to be able to see what is going on...

Most of the time I use the homebrew RX for audio, and home brew tube rigs for TX, but hate operating without the sdr.

[IN3IEX]

If you have sdr-iq you can use sdr# and the scope plugin.
Let us know how it works.
Giorgio

[N2DTS]

I have not tried sdr#.
Something new to play with I suppose...

[IN3IEX]

Yes.
The relevant informations are:

1)Most SDR (the softwares) have good spectrum analyzers, but no oscilloscope.
2)SDR# is very efficient and in the middle of its development.
3)SDR# has a plugin structure that enables easy addition of functionalities.
4)SDR# "now" has an oscilloscope plugin that can be used to evaluate AM modulation % and clipping events.
5)Combined time domain, spectral analysis and real time monitoring with the same software.
6)Band recording capabilities for offline analysis.

The software is free.
The minimalist hardware is 10 $ !

[N2DTS]

Where is the oscope plug in?
What is it called?

Know offhand what to pick to use sdr# with an sdr-iq?

[IN3IEX]

Hi, download IN3IEXScopeViewZIP.wav  that is attached to my first message.
Rename it IN3IEXScopeViewZIP.zip and then unzip it.
You will find the dll and instructions.
I renamed it because it is not possible to attach zip files to our messages......

The best way to get a positive result is to download SDR# from here: http://qsl.net/e/ea1kr//sdr/rtlsdr/
You will find a version of SDR# with all plugins.
THEN overwrite the original oscilloscope dll with MY DLL (the one attached to my first message here).
Alternatively follow my instructions and edit the configuration file ( SDRSharp.exe.Config )with notepad.

BEFORE installing SDR#  install Microsoft .NET framework 3.5  (not 4.0 or 3.0 it must be 3.5) You can have different .NET frameworks installed with no problem, the .NET number is NOT a version number.
Top left, between stop and select buttons there is the possibility to choose sdr-iq. I do not have sdr-iq, therefore my knowledge ends here...

Giorgio

[KD3CN]

Steve, W3EEE did a nice write-up on the RTL-SDR:

http://www.w3eee.com/

I've been playing with one for a couple weeks on VHF/UHF with SDR#

73, Karl

[IN3IEX]

Interesting news.
Scopeview now has an algorithm to measure peak modulation % of AM signals automatically.
It is not perfect but works.
I updated the first message with the new files.

Giorgio

[IN3IEX]

Hi, I slightly improved the AM % measurement algorithm and updated the first message.
Please test it.

Interesting news.
Scopeview now has an algorithm to measure peak modulation % of AM signals automatically.
It is not perfect but works.
I updated the first message with the new files.

Giorgio

[IN3IEX]

This is a video showing how it works:

http://www.ing.unitn.it/~fontana/T4XB%20IN3IEX.avi

The transmitter is a modified Drake T4XB, microphone Collins MM-1
http://www.ing.unitn.it/~fontana/HiFiAMT4XB.pdf
http://www.ing.unitn.it/~fontana/TrueAMT4XB.pdf

The mike level was set very high in order to see the audio clipping (but no zero envelope clipping).

Giorgio

NOW Obsolete, See below.

[IN3IEX]

Updated dll in the first message.
Measurements of the peak asymmetry of the modulation are now available.
Positive peak % and negative peak %.

regards

Giorgio

[IN3IEX]

You have probably seen the two horizontal lines that keep track of the maximum and the minimum of the envelope.
After adding AM %+ and %- someone asked for a line showing the average value of the envelope inside the measuring time window.
Here it is.
I also added a new measure: AR% that is asymmetry ratio in percent.
IF AR%=100 the positive and negative peaks are fully balanced respect to the average. If AR%>100 positive peaks are larger than negative peaks, if AR%<100 negative peaks are larger than positive peaks (independently of modulation depth).  
The file in the first message is now updated.

EDIT: Troubles with the renamed wav file? Go here in the file section:
         http://uk.groups.yahoo.com/group/SDRSharp/messages

Giorgio

[IN3IEX]

Show Peaks added (nice).
Now the plot is green (or deep blue).
File in the first message updated.
...

[IN3IEX]

I added another method to measure negative and positive peaks.
This could be a nearly satisfactory version.

This is part of the attached manual:

Two cursors are available to measure the ratio between two amplitude levels.
The ratio is displayed in percent; it is also converted to AM modulation percent.
If AUTO AM is selected, "Ratio %" is controlled by cursors, "AM%" is determined automatically by an algorithm.

AM%=100*(max envelope - min envelope)/(max envelope).
AM% cannot exceed 100%.

The asymmetry of the modulation can be evaluated by two methods:

%+ %- Method box unchecked (standard method):
%+=100*(max envelope - average envelope)/(average envelope).
%-=100*(average envelope - min envelope)/(average envelope).
With this method %- can never exceed 100%, %+ can exceed 100%

%+ %- Method box checked:
%+=200*(max envelope - average envelope)/(max envelope).
%-=200*(average envelope - min envelope)/(max envelope).
With this method both %+ and %- can exceed 100%.
With this method AM% = (%+ + %-)/2 is the average of %+ and %-.



The "positivity" of the modulation may be defined by an Asymmetry Ratio AR% defined by:
AR%=100*(max envelope - average envelope)/(average envelope - min envelope) that is applicable to both methods.
IF AR%=100 the positive and negative peaks are fully balanced respect to the average. If AR%>100 positive peaks are larger than negative peaks, if AR%<100 negative peaks are larger than positive peaks (independently of modulation depth). 

White orizontal lines show max envelope and min envelope. A yellow horizontal line shows the average value of the envelope.

Holdoff and HOLD button can be used to slowdown display or hold a waveform for analysis.

Decimation has been optimized for envelope display.

Show Peaks may offer a better view of the envelope.

To display the envelope of a signal that has for instance 10 kHz bandwidth, like an AM commercial station, you must select LSB or USB with 10 kHz filter and tune the whole spectrum of the station in the filter passband. The carrier of the station will be at filter center frequency and the scope will show the sine signal of the carrier. You cannot listen to the audio of the station with this configuration, you will listen to a 5 kHz sine signal (the carrier) plus some noise due to the sidebands.
This configuration is similar to the configuration used for DRM or digital modes via Virtual audio cable and external decoders.   

[flintstone mop]

Someday, someone will write or include a mod scope in some of the SDR software that runs on popular sdr receivers like the sdr-iq or qs1r.
Flex made a bodge of it in AM, it showing only the audio, not any carrier...

As it is, running the sdr allows me to see overmod when the signal goes wide..


I wish more AM'ers would use an sdr, at least as a band scope, as people are often off frequency or have screwed up signals.
40 meters seems to have a hole at 7290 at night, and last night I noticed a long qso with various people operating on 7286 with a strong wide broadcast station on 7280 with audio out to 7285. 1 KHz away, when 7290 was open and clear...
Very weak narrow broadcast on 7295 was not bothering anything, its so usefull to be able to see what is going on...

Most of the time I use the homebrew RX for audio, and home brew tube rigs for TX, but hate operating without the sdr.

It will take a while before folks will understand what they are seeing in the display. And I can understand that computers aren't everybody's friend. OR "computer here is a WIN 95 and I am happy with it" There are probably still AMers on the air just looking at the meters on their table-top TX, or worse yet, a converted broadcast TX, and asking the Ham op on the other end 'How does this sound?"
Youtube has some nice videos showing the use of these tools and what you will see when everything is goodness.
The AM mod monitor from QIX Steve opened my eyes after I installed it on my computer. I never realized how much I was over-modulating and cutting off the carrier. The O'Scope is not accurate for that at all.
I'll use my SDR as the RX when the battle conditions are present. The tuner-uppers and close-by SSB is pretty much eliminated with proper adjustment of the bandpass and the DSP. Almost arm chair copy on any signal!!
Fred

[IN3IEX]

Before starting developing this software I did not fully understand the significance of positive and negative peaks of voice signal as shown in the envelope of AM for analyzing the quality of a transmitter.
With this kind of instruments it is possible to make comparisons between transmitters and circuital solutions and tune them for the better.

My two trasmitters are my home made plate and screen grid modulated 100W http://amfone.net/Amforum/index.php?topic=30021.0 and a modified Drake T4XB with screen modulation of the 6JB6s (cited above):
With normal voice modulation (decimation control full right to get the maximum number of samples of voice) I have:

Parameter            100W h.m.        T4XB                  
 
AM%                   98                   96
Positive  pk%        150                 133
Negative pk%        97                  92
Asym Ratio %        154                144

Take into account that the 100W h.m. is under its special AGC and is actually making almost no audio distortion and no splatter, the T4XB is slightly clipping the audio and is pushed quite hard with the mike gain (clipping is audio high level). Both do not show envelope zero clipping, the T4XB makes some splatter only because of the audio clipping and high audio bandwidth. The T4XB is partial controlled carrier, on the other hand this software processes independent bunches of 100ms of signal and the varying carrier level is not a problem.
With these settings both give very good reports, but as you can see the home made is better.

Giorgio