distortion analyzer suggestions?

[GK]

Well, now the analysis of the  4 x 0.25W tones vs a single tone at 1W has been done (not just by me) you seem to be singing a different tune now to the "no it isn't" answer you gave below in an earlier post

It is much harder (on the amplifier) to communicate 1W of signal via two 0.5W tones than it is to send a single tone at 1W. it gets even harder on the amplifier if you try and send four tones each at 0.25W to make up the 1W of average signal power.

No, it isn't. Ironically you are (again) applying a common audiophile intuition. HD and IMD are both products non-linearity. If an amplifier has low measured THD across the full audio band it will have no IMD issues and IMD measurements will not be anymore "revealing" than the THD measurements. This fact has been established for about 50 years already. 
 

I think it IS harder on the amplifier to send 4 tones at 0.25W compared to a single tone at 1W.

You are welcome to argue against this but I somehow doubt that is going to happen now

It's been fun

    

Silly games now? Well your post didn't make clear distinctions between average and instantaneous peak power and in the context of comparative IMD and HD testing it didn't make any sense. I simply gave you the benefit of the doubt that you weren't trying to do something so silly as proposing a non-comparative IMD test condition to prove that HD testing is theoretically inferior.

I have been quite clear in the points that I have been making on the merits of HD analysis versus IMD analysis and have certainly not learnt anything new here in this debate with you. 









 

[G0HZU]

just one example; if you know the THD+N performance of some broadcast or telecoms audio equipment, a time-effective and routine THD+N evaluation will quickly tell you with a relatively high degree of confidence that the equipment is still functioning as the manufacturer intended.

....and have certainly not learnt anything new here in this debate with you.

How about a real world example that violates your first quote above?

I can see I'm going to have to be careful how I answer this but I can give you a real world 'system' issue that shows the benefit of two tone testing over single tone for an audio amplifier in a telecoms system. Please realise that this is 'just' one example where IMD testing is more revealing of a fault in the AF amplifier compared to a single tone test. I can give others if you dismiss this one.

Let's look at a system issue involving a SSB broadcast transmitter.

It gets returned with a fault report that says "extremely high distortion in AF stages. Audio unreadable". So technician #1 fires up the transmitter and does a single tone system test and looks at THD in the transmit AF stage with the transmitter running as well. The AF amplifier in the transmit stage passes with flying colours and gives the expected low level of THD distortion and the RF signal from the SSB transmitter is also fine.

So it gets put back into service as fully working. It immediately gets returned as faulty again with the same report and gets sent to engineering to have a deeper look.

A two tone IMD test is carried out and the problem with the broadcast is now easily apparent. Due to a fault in the audio amplifier (broken RF decoupling capacitors) the transmitted RF envelope of the SSB signal is able to get into the AF stage and be rectified (detected) in the faulty amplifier. This injects the crudely rectified audio into the AF stage. This sets up a potential feedback path and the system can go unstable and sound awful.

The system fault can't be replicated with a single tone test because the RF envelope from the transmitter is constant and so no audio is rectified to corrupt the AF stage.

I hope you will accept that this can happen and does happen in SSB transmitters. Sometimes it's down to a fault or even poor design inside the radio and sometimes it can be caused by poor aerial grounding allowing high common mode currents to travel back to the radio and upset its operation.

A single tone THD test is a waste of time here because it can't replicate the fault condition that causes distortion in the Tx AF stage.

[G0HZU]

Getting the thread back on track... If anyone is interested in the 2015's (fairly limited) capabilities I can show a screenshot of the little GUI I produced for my 2015 last year. I run it on a tiny netbook so there isn't much screenspace.

It has a marker function to allow each harmonic to be measured using the mouse. See below: You simply click on each green bar to measure the level of that harmonic in terms of -dBc.

Here's a link to show how I evaluated the THD function on my 2015 when I first bought it.

https://www.eevblog.com/forum/testgear/keithley-2015thd-initial-tests/msg313669/#msg313669

At the end of the day' it's still pretty blinkered in what it can measure but I am only really interested in seeing the levels of the harmonics and I'm not interested once the harmonics dip below about -65dBc. The THD figure is of little interest to me. I primarily use it to quickly look at the harmonic levels from signal sources as I vary the signal  level of the source. It's the quickest and easiest piece of test gear I have for this if I use this GUI.

What's nice is you can simply stuff a signal into the 2015 and get the screenshot result almost instantly despite the high number of SCPI commands I have to send it over GPIB to fill the screen with all the data.

[GK]

just one example; if you know the THD+N performance of some broadcast or telecoms audio equipment, a time-effective and routine THD+N evaluation will quickly tell you with a relatively high degree of confidence that the equipment is still functioning as the manufacturer intended.

....and have certainly not learnt anything new here in this debate with you.

How about a real world example that violates your first quote above?

I can see I'm going to have to be careful how I answer this but I can give you a real world 'system' issue that shows the benefit of two tone testing over single tone for an audio amplifier in a telecoms system. Please realise that this is 'just' one example where IMD testing is more revealing of a fault in the AF amplifier compared to a single tone test. I can give others if you dismiss this one.

Let's look at a system issue involving a SSB broadcast transmitter.

It gets returned with a fault report that says "extremely high distortion in AF stages. Audio unreadable". So technician #1 fires up the transmitter and does a single tone system test and looks at THD in the transmit AF stage with the transmitter running as well. The AF amplifier in the transmit stage passes with flying colours and gives the expected low level of THD distortion and the RF signal from the SSB transmitter is also fine.

So it gets put back into service as fully working. It immediately gets returned as faulty again with the same report and gets sent to engineering to have a deeper look.

A two tone IMD test is carried out and the problem with the broadcast is now easily apparent. Due to a fault in the audio amplifier (broken RF decoupling capacitors) the transmitted RF envelope of the SSB signal is able to get into the AF stage and be rectified (detected) in the faulty amplifier. This injects the crudely rectified audio into the AF stage. This sets up a potential feedback path and the system can go unstable and sound awful.

The system fault can't be replicated with a single tone test because the RF envelope from the transmitter is constant and so no audio is rectified to corrupt the AF stage.

I hope you will accept that this can happen and does happen in SSB transmitters. Sometimes it's down to a fault or even poor design inside the radio and sometimes it can be caused by poor aerial grounding allowing high common mode currents to travel back to the radio and upset its operation.

A single tone THD test is a waste of time here because it can't replicate the fault condition that causes distortion in the Tx AF stage.

Now you are changing the topic almost completely. Your original claim (specifically on THD Vs IMD testing) was that traditional THD analysis (using traditional analogue analyzers) in audio (as in 20 Hz - 20 kHz "HiFi") design is inferior to IMD analysis for reasons thus far unsubstantiated.

I specifically outlined the specific conditions that make your claim generally untrue, and in a previous post I even gave a specific example (bandwidth limited systems) where IMD testing actually has the upper hand over THD testing (measuring the linearity of Class D amplifiers at the higher audio frequencies).
In Class D amplifier operation the relationship between low THD and low IMD still holds true right out to 20 kHz theoretically, however the problem of measurement is a practical one as there is no viable way to measure the amplifiers THD performance prior to the output filter.

As for your SSB example, it doesn't violate my text that you quoted as, firstly, I specifically wrote: "a time-effective and routine THD+N evaluation will quickly tell you with a relatively high degree of confidence that the equipment is still functioning as the manufacturer intended" - to repeat - "high degree of confidence", not "exclusively, 100% for sure". Secondly I was talking about audio equipment - not transmitters. IMD is used almost exclusively for testing the linearity of transmitters, as, in a partly similar way to Class D in audio design, they fall into the category of bandwidth limited systems (which I mentioned previously), but that's a whole different topic again.   

I think that is about all that I am willing to contribute to this thread. I'm not interest in further "debate" where basic points go ignored and unacknowledged and the topic of discussion simply get changed to something completely different instead.

[G0HZU]

Actually what I think is happening here is a chronic misinterpretation of each other's posts.

Now you are changing the topic almost completely. Your original claim (specifically on THD Vs IMD testing) was that traditional THD analysis (using traditional analogue analyzers) in audio (as in 20 Hz - 20 kHz "HiFi") design is inferior to IMD analysis for reasons this far unsubstantiated.

What kicked off the IMD claim was your reference to a health check of a broadcast transmitter. My immediate thought process with respect to this is that (in my experience) the IMD test is a harsher SYSTEM test of the broadcast system.

I think you are only interested in the test of the amplifier in isolation. In my experience there are other things that can generate terms that look like distortion in an audio amplifier. See my example above.

So partly why this thread is derailing is because you are sticking to your idealistic opinion about the maths behind distortion and I'm looking at the system limits. eg system issues that can mean that IMD testing can be a harsher test.

I also proposed a multi tone test system using the same average signal power sent to the listener/load which you clearly misinterpreted.

Also, I think you need to lighten up a bit...


Also, I expect you to misinterpret this post as well

[GK]

What kicked off the IMD claim was your reference to a health check of a broadcast transmitter.

I think the case of "chronic misinterpretation" is on your side. I never said "broadcast transmitter". I said "some broadcast or telecoms audio equipment". There are plenty of examples of equipment under that category that are not actual transmitters (nearly everything from between the studio mic. and the exciter). Ask vk6zgo, he might be able to give you some examples.

I only cited such examples of "professional" audio equipment as you were at pains to inform everyone that you are not an "audiophile" and are not interested in the evaluation of extremely low distortion equipment.

Nighty night.



 

[G0HZU]

I think the case of "chronic misinterpretation" is on your side.

OK, but it was a reasonable assumption of mine that that radio is involved in telecoms equipment

But my point is that you are misinterpreting my posts too. 

Edit:
Rest of post citing example deleted as it isn't productive.