Elecraft

Measuring SSB RMS Power

Classic

List

Threaded

12 messages Options

John Oppenheimer

Measuring SSB RMS Power

Using DL1 with Oscilloscope to measure KX3 SSB RMS power:

https://www.kn5l.net/ssbpower/

Question, is this technique correct?

John KN5L
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]

Don Wilhelm

Re: Measuring SSB RMS Power

John,

That connection is OK if you realize that your scope is reading the DC
voltage developed by rectifying the RF from 1/2 of the load. You still
have to use the power formula that is in the DL1 manual.

I prefer to allow the 'scope to read the full peak RF voltage. Connect
the 'scope probe to the side of one resistor which is connected to the
BNC jack (one of the leads closest to the BNC), and the 'scope ground
probe to one or the resistor leads furthermost away from from the BNC jack.

Then use the formula to convert peak voltage to power (for a 50 ohm
load) which is V squared then divided by 400. Easy to do quickly with a
handy calculator.

73,
Don W3FPR

On 6/20/2019 9:38 AM, John Oppenheimer wrote:
> Using DL1 with Oscilloscope to measure KX3 SSB RMS power:
>
> https://www.kn5l.net/ssbpower/
>
> Question, is this technique correct?
>
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]

Bob McGraw - K4TAX

Re: Measuring SSB RMS Power

In reply to this post by John Oppenheimer

Here is a paper on the topic. This covers the correct procedure.

http://www.eng.auburn.edu/~troppel/courses/TIMS-manuals-r5/TIMS%20Experiment%20Manuals/Student_Text/Vol-A1/a1-12.pdf

73
Bob, K4TAX

Sent from my iPhone

> On Jun 20, 2019, at 8:38 AM, John Oppenheimer <[hidden email]> wrote:
>
> Using DL1 with Oscilloscope to measure KX3 SSB RMS power:
>
> https://www.kn5l.net/ssbpower/
>
> Question, is this technique correct?
>
> John KN5L
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
> Message delivered to [hidden email]
>

Elecraft mailing list

Re: Measuring SSB RMS Power

Don , That may be peak to peak voltage squared decided by 400 for a 50 ohm load.
Thanks 73 Ken K5DNL

On Thursday, June 20, 2019, 9:30:46 AM CDT, Bob McGraw K4TAX <[hidden email]> wrote:

Here is a paper on the topic. This covers the correct procedure.

http://www.eng.auburn.edu/~troppel/courses/TIMS-manuals-r5/TIMS%20Experiment%20Manuals/Student_Text/Vol-A1/a1-12.pdf

73
Bob, K4TAX

Sent from my iPhone

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]

Don Wilhelm

Re: Measuring SSB RMS Power

Ken,

The DL1 is a 50 ohm load. I did specify that the formula was applicable
for a 50 ohm load.

For other loads, the formula is Vpeak squared divided by 8R.

It saves the trouble of converting to RMS and then using the more
familiar Vrms squared divided by R.

If you wish to derive the formula yourself, you can do that, but
represent the RMS voltage as V time sqrt 2 rather then the approximation
of 1.414 (or 0.707) which is normally shown. The sqrt will cancel out
when it is squared.

73,
Don W3FPR

On 6/20/2019 10:50 AM, Ken Roberson via Elecraft wrote:

> Don , That may be peak to peak voltage squared decided by 400 for a 50 ohm load.
> Thanks 73 Ken K5DNL
>
> On Thursday, June 20, 2019, 9:30:46 AM CDT, Bob McGraw K4TAX <[hidden email]> wrote:
>
> Here is a paper on the topic. This covers the correct procedure.
>
> http://www.eng.auburn.edu/~troppel/courses/TIMS-manuals-r5/TIMS%20Experiment%20Manuals/Student_Text/Vol-A1/a1-12.pdf
>
> 73
> Bob, K4TAX
>
>
> Sent from my iPhone
>
>> On Jun 20, 2019, at 8:38 AM, John Oppenheimer <[hidden email]> wrote:
>>
>> Using DL1 with Oscilloscope to measure KX3 SSB RMS power:
>>
>> https://www.kn5l.net/ssbpower/
>>
>> Question, is this technique correct?
>>
>> John KN5L
>> ______________________________________________________________
>> Elecraft mailing list
>> Home: http://mailman.qth.net/mailman/listinfo/elecraft
>> Help: http://mailman.qth.net/mmfaq.htm
>> Post: mailto:[hidden email]
>>
>> This list hosted by: http://www.qsl.net
>> Please help support this email list: http://www.qsl.net/donate.html
>> Message delivered to [hidden email]
>>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
> Message delivered to [hidden email]
>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
> Message delivered to [hidden email]
>

John Oppenheimer

Re: Measuring SSB RMS Power

In reply to this post by Don Wilhelm

Hi Don,

On 6/20/19 9:21 AM, Don Wilhelm wrote:
> John,
>
> That connection is OK if you realize that your scope is reading the DC
> voltage developed by rectifying the RF from 1/2 of the load. You still
> have to use the power formula that is in the DL1 manual.

Well yes, power references are DCV^2 / 25

DdB ratios computed using voltage ratios in the form of
20 * Log(v1/v2). For example, 0 compression Pep/Pav dB =
20 * log(15.3/5.24) = 9.3. I do need to flip the table title to Pav/Pep
for proper negative dB values, Done.

> I prefer to allow the 'scope to read the full peak RF voltage. Connect
> the 'scope probe to the side of one resistor which is connected to the
> BNC jack (one of the leads closest to the BNC), and the 'scope ground
> probe to one or the resistor leads furthermost away from from the BNC jack.

Using detected RF envelope for oscilloscope average RF measurement at
each digitized point during long, six second, trace. This will result
with increased accuracy when measuring modulated speech.

John KN5L

> On 6/20/2019 9:38 AM, John Oppenheimer wrote:
>> Using DL1 with Oscilloscope to measure KX3 SSB RMS power:
>>
>> https://www.kn5l.net/ssbpower/
>>
>> Question, is this technique correct?
>>
>
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]

John Oppenheimer

Re: Measuring SSB RMS Power

In reply to this post by Bob McGraw - K4TAX

Hi Bob,

Document Table 1, Line 7. speech = V/5*2^.5. which is same as Vrms/5.
Using measured RMS ratios, Vrms=15.3, 15.3/5 = 3V, compared to measured
5.24V.

Speech is rather undefined, so not too worried about the discrepancy. My
speech with no pauses is ~= Vrms / 3.

Really what is being evaluated is gain resulting from speech compression.

John KN5L

On 6/20/19 9:29 AM, Bob McGraw K4TAX wrote:

> Here is a paper on the topic. This covers the correct procedure.
>
> http://www.eng.auburn.edu/~troppel/courses/TIMS-manuals-r5/TIMS%20Experiment%20Manuals/Student_Text/Vol-A1/a1-12.pdf
>
>
> 73
>
> Bob, K4TAX
>
>
> Sent from my iPhone
>
> On Jun 20, 2019, at 8:38 AM, John Oppenheimer <[hidden email]
> <mailto:[hidden email]>> wrote:
>
>> Using DL1 with Oscilloscope to measure KX3 SSB RMS power:
>>
>> https://www.kn5l.net/ssbpower/
>>
>> Question, is this technique correct?
>>
>> John KN5L

Bob McGraw - K4TAX

Re: Measuring SSB RMS Power

In reply to this post by John Oppenheimer

To evaluate speech use a pink noise source. That is the standard source signal for acoustic measurements.

Bob, K4TAX

Sent from my iPhone

> On Jun 20, 2019, at 10:34 AM, John Oppenheimer <[hidden email]> wrote:
>
> Hi Don,
>
>> On 6/20/19 9:21 AM, Don Wilhelm wrote:
>> John,
>>
>> That connection is OK if you realize that your scope is reading the DC
>> voltage developed by rectifying the RF from 1/2 of the load. You still
>> have to use the power formula that is in the DL1 manual.
>
> Well yes, power references are DCV^2 / 25
>
> DdB ratios computed using voltage ratios in the form of
> 20 * Log(v1/v2). For example, 0 compression Pep/Pav dB =
> 20 * log(15.3/5.24) = 9.3. I do need to flip the table title to Pav/Pep
> for proper negative dB values, Done.
>
>> I prefer to allow the 'scope to read the full peak RF voltage. Connect
>> the 'scope probe to the side of one resistor which is connected to the
>> BNC jack (one of the leads closest to the BNC), and the 'scope ground
>> probe to one or the resistor leads furthermost away from from the BNC jack.
>
> Using detected RF envelope for oscilloscope average RF measurement at
> each digitized point during long, six second, trace. This will result
> with increased accuracy when measuring modulated speech.
>
> John KN5L
>
>>> On 6/20/2019 9:38 AM, John Oppenheimer wrote:
>>> Using DL1 with Oscilloscope to measure KX3 SSB RMS power:
>>>
>>> https://www.kn5l.net/ssbpower/
>>>
>>> Question, is this technique correct?
>>>
>>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
> Message delivered to [hidden email]
>

John Oppenheimer

Re: Measuring SSB RMS Power

Hi Bob,

Viewing pink noise with an oscilloscope, it is continuous, where as
speech has breaks.

The web page is evaluating both expected power to a load and compression
gain. Using a little over six second recorded speech. Recording contains
fast spoken words, mostly numbers and a few "this is a test" etc, for
maximum power over time. Evaluating scope shoot, about 25 words during
six second recording. Not even close to pink noise.

John KN5L

On 6/20/19 11:27 AM, Bob McGraw K4TAX wrote:

> To evaluate speech use a pink noise source. That is the standard source signal for acoustic measurements.
>
> Bob, K4TAX
>
>
> Sent from my iPhone
>
>> On Jun 20, 2019, at 10:34 AM, John Oppenheimer <[hidden email]> wrote:
>>
>> Hi Don,
>>
>>> On 6/20/19 9:21 AM, Don Wilhelm wrote:
>>> John,
>>>
>>> That connection is OK if you realize that your scope is reading the DC
>>> voltage developed by rectifying the RF from 1/2 of the load. You still
>>> have to use the power formula that is in the DL1 manual.
>>
>> Well yes, power references are DCV^2 / 25
>>
>> DdB ratios computed using voltage ratios in the form of
>> 20 * Log(v1/v2). For example, 0 compression Pep/Pav dB =
>> 20 * log(15.3/5.24) = 9.3. I do need to flip the table title to Pav/Pep
>> for proper negative dB values, Done.
>>
>>> I prefer to allow the 'scope to read the full peak RF voltage. Connect
>>> the 'scope probe to the side of one resistor which is connected to the
>>> BNC jack (one of the leads closest to the BNC), and the 'scope ground
>>> probe to one or the resistor leads furthermost away from from the BNC jack.
>>
>> Using detected RF envelope for oscilloscope average RF measurement at
>> each digitized point during long, six second, trace. This will result
>> with increased accuracy when measuring modulated speech.
>>
>> John KN5L
>>
>>>> On 6/20/2019 9:38 AM, John Oppenheimer wrote:
>>>> Using DL1 with Oscilloscope to measure KX3 SSB RMS power:
>>>>
>>>> https://www.kn5l.net/ssbpower/
>>>>
>>>> Question, is this technique correct?

Jim Brown-10

Re: Measuring SSB RMS Power

Bob is correct to the extent that Pink Noise approximates the SPECTRUM
of speech, but John's is on target that speech is not continuous, but
highly variable in amplitude. For uncompressed speech, the average value
is typically as much as 20 dB below the peak value; compression can
reduce that to 10 dB or less. Highly sophisticated audio processing
that has been used by radio broadcasters for at least 40 years reduces
that ratio to 6 dB or less. "If the modulation meter moves below 100%,
it isn't loud enough!"

73, Jim K9YC

On 6/20/2019 10:13 AM, John Oppenheimer wrote:

> Hi Bob,
>
> Viewing pink noise with an oscilloscope, it is continuous, where as
> speech has breaks.
>
> The web page is evaluating both expected power to a load and compression
> gain. Using a little over six second recorded speech. Recording contains
> fast spoken words, mostly numbers and a few "this is a test" etc, for
> maximum power over time. Evaluating scope shoot, about 25 words during
> six second recording. Not even close to pink noise.
>
> John KN5L
>
> On 6/20/19 11:27 AM, Bob McGraw K4TAX wrote:
>> To evaluate speech use a pink noise source. That is the standard source signal for acoustic measurements.
>>
>> Bob, K4TAX
>>
>>
>> Sent from my iPhone
>>
>>> On Jun 20, 2019, at 10:34 AM, John Oppenheimer <[hidden email]> wrote:
>>>
>>> Hi Don,
>>>
>>>> On 6/20/19 9:21 AM, Don Wilhelm wrote:
>>>> John,
>>>>
>>>> That connection is OK if you realize that your scope is reading the DC
>>>> voltage developed by rectifying the RF from 1/2 of the load. You still
>>>> have to use the power formula that is in the DL1 manual.
>>> Well yes, power references are DCV^2 / 25
>>>
>>> DdB ratios computed using voltage ratios in the form of
>>> 20 * Log(v1/v2). For example, 0 compression Pep/Pav dB =
>>> 20 * log(15.3/5.24) = 9.3. I do need to flip the table title to Pav/Pep
>>> for proper negative dB values, Done.
>>>
>>>> I prefer to allow the 'scope to read the full peak RF voltage. Connect
>>>> the 'scope probe to the side of one resistor which is connected to the
>>>> BNC jack (one of the leads closest to the BNC), and the 'scope ground
>>>> probe to one or the resistor leads furthermost away from from the BNC jack.
>>> Using detected RF envelope for oscilloscope average RF measurement at
>>> each digitized point during long, six second, trace. This will result
>>> with increased accuracy when measuring modulated speech.
>>>
>>> John KN5L
>>>
>>>>> On 6/20/2019 9:38 AM, John Oppenheimer wrote:
>>>>> Using DL1 with Oscilloscope to measure KX3 SSB RMS power:
>>>>>
>>>>> https://www.kn5l.net/ssbpower/
>>>>>
>>>>> Question, is this technique correct?
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
> Message delivered to [hidden email]
>

k6dgw

Re: Measuring SSB RMS Power

Brown noise is supposed to be sleep inducing. The security folks used
pink noise in the plenum above the SCIF's to mask conversations. I
can't remember exactly what the spectrum of A-weighting was, I think it
had a dip in the mid-range.

The TV station I was employed by in college would "rent" us out to
smaller AM and FM stations who didn't have a resident engineer. The
standard AM configuration was an aggressive peak limiter followed by a
"Sta-Level" with a very long time constant. Result was the modulation
was pegged at 99.5% all the time, and it sounded that way too ... zero
dynamic range. Years ago, I looked at KFBK's waveform when the then new
Rush Limbaugh was beginning his bloviation career. It too was nearly
constant since Rush could inhale very very quickly and talked very fast.
[:-))

Non-classical FM stations [we serviced 4] used peak limiters only since
100% FM modulation is an arbitrary number [75 KHz deviation then]. The
one classical music FM station we had used no limiting/compression. Our
TV station used no limiter, NTSC aural deviation then was 25 KHz,
however our two network feeds sounded fairly well compressed.

I seem to get best audio reports, and recordings a couple of folks have
sent me sound best to me when my K3 compression reliably goes to 10 dB
with occasional peaks to 15 dB on the bar graph. Much more than that
and there's clearly room noise in pauses and the audio begins to distort.

73,
Fred ["Skip"] K6DGW
Sparks NV DM09dn
Washoe County

On 6/20/2019 11:15 AM, Jim Brown wrote:

> Bob is correct to the extent that Pink Noise approximates the SPECTRUM
> of speech, but John's is on target that speech is not continuous, but
> highly variable in amplitude. For uncompressed speech, the average
> value is typically as much as 20 dB below the peak value; compression
> can reduce that to 10 dB or less. Highly sophisticated audio
> processing that has been used by radio broadcasters for at least 40
> years reduces that ratio to 6 dB or less. "If the modulation meter
> moves below 100%, it isn't loud enough!"
>
> 73, Jim K9YC
>
> On 6/20/2019 10:13 AM, John Oppenheimer wrote:
>> Hi Bob,
>>
>> Viewing pink noise with an oscilloscope, it is continuous, where as
>> speech has breaks.
>>
>> The web page is evaluating both expected power to a load and compression
>> gain. Using a little over six second recorded speech. Recording contains
>> fast spoken words, mostly numbers and a few "this is a test" etc, for
>> maximum power over time. Evaluating scope shoot, about 25 words during
>> six second recording. Not even close to pink noise.
>>
>> John KN5L
>>
>> On 6/20/19 11:27 AM, Bob McGraw K4TAX wrote:
>>> To evaluate speech use a pink noise source. That is the standard
>>> source signal for acoustic measurements.
>>>
>>> Bob, K4TAX
>>>
>>>
>>> Sent from my iPhone
>>>
>>>> On Jun 20, 2019, at 10:34 AM, John Oppenheimer <[hidden email]> wrote:
>>>>
>>>> Hi Don,
>>>>
>>>>> On 6/20/19 9:21 AM, Don Wilhelm wrote:
>>>>> John,
>>>>>
>>>>> That connection is OK if you realize that your scope is reading
>>>>> the DC
>>>>> voltage developed by rectifying the RF from 1/2 of the load. You
>>>>> still
>>>>> have to use the power formula that is in the DL1 manual.
>>>> Well yes, power references are DCV^2 / 25
>>>>
>>>> DdB ratios computed using voltage ratios in the form of
>>>> 20 * Log(v1/v2). For example, 0 compression Pep/Pav dB =
>>>> 20 * log(15.3/5.24) = 9.3. I do need to flip the table title to
>>>> Pav/Pep
>>>> for proper negative dB values, Done.
>>>>
>>>>> I prefer to allow the 'scope to read the full peak RF voltage.
>>>>> Connect
>>>>> the 'scope probe to the side of one resistor which is connected to
>>>>> the
>>>>> BNC jack (one of the leads closest to the BNC), and the 'scope ground
>>>>> probe to one or the resistor leads furthermost away from from the
>>>>> BNC jack.
>>>> Using detected RF envelope for oscilloscope average RF measurement at
>>>> each digitized point during long, six second, trace. This will result
>>>> with increased accuracy when measuring modulated speech.
>>>>
>>>> John KN5L
>>>>
>>>>>> On 6/20/2019 9:38 AM, John Oppenheimer wrote:
>>>>>> Using DL1 with Oscilloscope to measure KX3 SSB RMS power:
>>>>>>
>>>>>> https://www.kn5l.net/ssbpower/
>>>>>>
>>>>>> Question, is this technique correct?
>> ______________________________________________________________
>> Elecraft mailing list
>> Home: http://mailman.qth.net/mailman/listinfo/elecraft
>> Help: http://mailman.qth.net/mmfaq.htm
>> Post: mailto:[hidden email]
>>
>> This list hosted by: http://www.qsl.net
>> Please help support this email list: http://www.qsl.net/donate.html
>> Message delivered to [hidden email]
>>
>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
> Message delivered to [hidden email]

John Oppenheimer

Re: Measuring SSB RMS Power

In reply to this post by John Oppenheimer

Searching email archives, found this comment from Lyle in KX3 list:

On 2/8/14 4:04 PM, Lyle Johnson wrote:
> The KX3 SSB compression algorithm is a bit different than the K3 for a
> number of reasons. Reducing the bandwidth, removing the very low
> audio frequencies, etc., are all part of the algorithm design.

Therefore, simple power gain may not be the full story. Per above, power
is concentrated within a narrower BW as compression is increased.

Though I wonder if TXEQ with significant low frequency attenuation
changes compression power gain.

For testing, KX3 TX EQ is set to "KX3 Optimal" in:
https://www.kn5l.net/KX3-TxEq/

John KN5L

On 6/20/19 8:38 AM, John Oppenheimer wrote:
> Using DL1 with Oscilloscope to measure KX3 SSB RMS power:
>
> https://www.kn5l.net/ssbpower/
>
> Question, is this technique correct?
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]