Understanding Third Order Intercept

This video provides a general technical introduction to the concept of third order intercept and how third order intercept measurements are made using a spectrum analyzer or network analyzer.
More information on Rohde & Schwarz signal and spectrum analyzers: www.rohde-schwarz.com/us/prod...
Need help with Rohde & Schwarz products & services?
Contact Us: www.rohde-schwarz.com/us/m4c/...

[@CuriousMarc]

Remarkable job explaining this. This is very clear but not dumbed down. It also connects theory with practice, extracting what matters in the real world from what does not. And it deals with non-idealities, once again not dumbing anything down. In other words, clear, short but complete and practical. That is no small feat.

[@pauldenisowski]

Thanks! Really appreciate the feedback!

[@lucgondim]

Remarkable comment also!

[@El650Jefe]

This guys videos are the best in all of KZbin. Absolutely beautiful vids.

[@naftilos76]

That was a treasure. Compressed knowledge and steps to be taken by highly professional people. Even though this is a marketing tool, it is still unbelievably precious for me and i guess for most people that are RF electronics enthusiasts! Thanks again.

[@pauldenisowski]

Thanks for the feedback!

[@MubashirMahmoodsial]

first time in my life i clearly understand the actual concept behind TOI. and now i will try 2 different signal generators for 2 tones as i was using single generator to produce 2 tones

[@ellisjsimon]

I know that I'm very late to the party, but this is a very clear explanation. Many thanks.

[@pauldenisowski]

We're happy for feedback anytime :) Thanks!

[@ElPasoTubeAmps]

Excellent - thank you.

[@mohamedaboelenein7727]

This video is sooooo good!, thanks a lot!

[@-TOLINSKI]

Thank you! This is helping me so much in wireless comms course

[@pauldenisowski]

Glad it's helpful - thanks!

[@wd8dsb]

Excellent video, best one I have watched related to TOI. Thanks.

[@aayushpatel2243]

Many doubts got cleared by watching this video , Thanks

[@pauldenisowski]

My pleasure - thanks for watching!

[@advanceddesignsystemadsrfi449]

best video/..... neat and clean

[@krish2nasa]

Excellent explanation. Thank you very much.

[@estevan_tu]

Excellent video, thank you !

[@metebalci]

All of these videos are very nice, thanks.

[@hubercats]

Excellent presentation. Thank you!

[@mcie6306]

Fanstastic video , thanks for that

[@pauldenisowski]

Thank you!

[@yangwang5188]

11:40 我认为IP3应该是最上面的蓝色线，基波应该是中间的橙色线。

[@pauldenisowski]

您对了!谢谢您

[@MrArnus91]

You're right! IP3 is the BLUE plot and the FUNDAMENTAL is the ORANGE one.

[@relaxingmusic0215]

Yes, you are right! based on the TOI formula, it make sense

[@dwarslopers]

Really great explained, as good as the quality of your devices. Many thanks!

[@pauldenisowski]

Thank you!

[@sajjsamm]

best educational video ever!

[@Zain15a]

Excellent!!!

[@arman3020]

thank you! it is very easy understanding!

[@AllElectronicsChannel]

Well done, thank you !

[@manfredgotzinger1258]

excellent!!!

[@bacnguyenkhac154]

thank for your lesson

[@samsureshj]

clear explanation ...awesome

[@dfpolitowski2]

Explanation helps with ham exams too.

[@pauldenisowski]

GL ES 73! Paul, KO4LZ

[@andyhinkel]

thank you! this helped me a lot!

[@anuph6068]

so good

[@MohaMmad-bp7pg]

Perfect! Very good presentation. Keep going please. These measurements examples are very useful. Do it for noise figure, spurious,phase noise, and .........please

[@pauldenisowski]

Thanks. Some of the presentations that are currently planned include the ones you listed (noise figure, spur search, phase noise) as well as things like ACLR, EVM, noise power ratio, group delay, CCDF, etc. There are also series on analog and digital modulation and power measurements that should be ready soon.

[@sarveshkale863]

@@pauldenisowski eagerly waiting for it sir ! continue the good work

[@sarveshkale863]

@@pauldenisowski eagerly waiting for it Sir ! great job

[@pauldenisowski]

By the way, the first of many phase noise presentations was just posted: see "Understanding Phase Noise Fundamentals" in the Test and Measurement playlist.

[@vk5fe943]

Good video

[@jafaraljafari4005]

thank you for this excellent video , very clear

[@pauldenisowski]

Thanks!

[@rfmonkey4942]

wunderbar Ja..

[@RWBreckinridge]

This is an excellent video! Very helpful! Also, the VO sounds really, really good- a quick pass with RX de-mouth click would take it from 99 to 100%!

[@pauldenisowski]

Thanks for the suggestion!

[@kunaldhamdhere6178]

Can't thank you enough for this :)

[@pauldenisowski]

Happy to help - thanks for the feedback!

[@superdtp1630]

Thanks

[@nabil1399]

Wow, just wow

[@KaptainLuis]

thaaaank you!

[@abdelkaderalicherif6439]

thank you

[@trunke1085]

golden

[@pauldenisowski]

Thank you!

[@brothercavil491]

Thank you for the video- as others have said, it's very clear and well presented. Can you explain how the frequencies of the two tones in the two-tone test are chosen? I know you want to avoid having one be a harmonic of the other, but other than that, I don't know what the constraints are, and I've seen different spacing used in different situations- it looks like you're using 30 MHz in your example at 8:04, but I've seen everything from 10 MHz to 100 before. Thanks again

[@uploadJ]

re: "Can you explain how the frequencies of the two tones in the two-tone test are chosen? " Short answer: Experience. Long answer: Determined by the DUT (device under test) such as freq operating range and so forth. Additional factors enter into the freq chosen, such as power supply bypassing and dynamic response, especially when freqs are close by such as in the audio range.

[@DeepakKumarSingh2015]

Nice #discreteoptical

[@theoryandapplication7197]

intresting

[@gill6335]

what frequency difference should be used between f1 and f2 for the 2 tone test? Is it dictated by some standard like 1MHZ or so?

[@thingsyoumissedinclass]

10:23 When you add the input attenuator, are you adding an attenuator between dut and Spectrum Analyzer (SA), or changing SA's internal mechanical attenuator? Would that make a difference?

[@pauldenisowski]

You want the attenuator to be before the first active device in the spec an. Many (most?) spec ans have an internal, user-controllable input attenuator so usually the procedure to add 10 dB of attenuation is simply done via the spec an user interface. The results would be the same if you were to screw an external attenuator to the spec an input port, although clearly this is slightly les convenient. [It should also be noted that instead of mechanical attenuators, instruments may also have internal electronic attenuators].

[@CAPSTONEization]

Question: What happens when the signal to the DUT is differential? In terms of Harmonics?

[@4DRC_]

Is IP3 ever considered in mixed signal audio frequency operations or is it primarily an RF measurement?

[@pauldenisowski]

Good question. Third order intercept is, in my experience, primarily an RF measurement. A common way of measuring linearity (or distortion) in audio applications is either SINAD or THD (total harmonic distortion). I've done videos on both of these, so please take a look!

[@thingsyoumissedinclass]

Could you elaborate on the method you use to determine whether the Spectrum Analyzer(SA) limits the linearity measurement?

[@pauldenisowski]

Sure. :) Essentially the question is whether the 3rd order intermodulation products are being generated within the device under test and/or in the spectrum analyzer. If the products are being generated external to the spec an, then an input attenuator will attenuate the fundamentals and the products equally. And since TOI/IP3 is calculated using the *difference* in levels between the fundamentals and intermod products, reducing them all by the same amount won't change the TOI/IP3 (assuming, of course, that you don't add so much attenuation that the products disappear under the noise floor ....). If the products are generated within the spec an, then an external attenuator will only reduce the level of the fundamentals. Recall that the amplitude of the third order products rises (or falls) much more quickly than the amplitude of the fundamentals (steeper slope). So attenuating the fundamentals at the input to the spec an will create a noticeable change in TOI if the products are being generated internally. Hope that helps!

[@SureshKumar-nk2ok]

sir we too use agilent network analyser but only 2 ports rf out and rf in ,how to use for measuring it.

[@efrembelay2657]

when i want to calculate the TOI manually, is it a must to change dBm to watt? or i will use as it is?

[@erfho8y]

Is this also relevant to VSTs in DAWs??

[@acn535]

excuse me, where can we get this amazing presentation?

[@pauldenisowski]

I don't think I ever made a whitepaper version of this - do you think there would be enough interest?

[@wyw4466]

How do you prevent the tons not mixing before it reaches the DUT?

[@pauldenisowski]

The section on source isolation provides some guidance. Generally, either attenuators or circulators can be helpful in preventing undesired mixing.

[@danganh870]

Thank you for the video. However, I still don't understand how can we detect the causes of IMD by applying an Input Attenuator? Can anybody explain that to me?

[@pauldenisowski]

Sure : ) If you look at the equation on slide 13, you see that TOI (or IP3) is a function of the power difference, P-delta, between the fundamental tones (Ptone) and the third order products (PIM3). In the case of external generation, both Ptone and PIM3 are all attenuated by the same amount, so P-delta stays the same and thus TOI stays the same. If the distortion is being internally generated, Ptone will be decreased (since the fundamentals passed through the attenuator) but PIM3 will be decreased even more: remember that the power in the products has a different (steeper) slope than the power in the fundamentals (see slide 9). A 10 dB decrease in the power of the fundamental tones results in a 30 dB decrease in the intermodulation products. So attenuating the fundamental means that P-delta will change and TOI will decrease *if* the distortion is being generated internally. We actually have an entire white paper: "Intermodulation Distortion Measurements on Modern Spectrum Analyzers" that covers IP3 measurements in much more detail. In particular, pages 15 and 16 mention how proper use of input attenuation can reduce the generation of intermodulation products in the first mixer in a spectrum analyzer. scdn.rohde-schwarz.com/ur/pws/dl_downloads/dl_application/application_notes/1ef79/1EF79_1E.pdf If you watch our video "Understanding Basic Spectrum Analyzer Operation" it also mentions the importance of properly setting the reference level to avoid overdriving the first stages of the spectrum analyzer and creating IMD within the analyzer. Being aware of the possibility of IMD and adjusting your settings / levels appropriately is important in making good spectrum analyzer measurements. Hope that helps!

[@danganh870]

@@pauldenisowski Thank you, Paul. Your explanation is really helpful.

[@pauldenisowski]

@@danganh870 Anytime :) Happy New Year!

[@chrisscott1547]

So, if I want my 3rd order IM products at least -30 dBc, my fundamental output signal cannot exceed IP3 - 10dB...

[@El650Jefe]

At 10:35 can someone explain why the fundamental Signals power remains essentially the same when 10db of attenuation is added? I would of thought their power levels would of dropped by 10db.

[@AzT3cMaSSaCR3]

To my best knowledge, the values "did" drop but the spectrum analyzer automatically offsets the value for you. By default the values you see on spectrum analyzer are with reference/calibrated to the input port of the analyzer. The analyzer has an internal attenuator that can be automatically or manually adjusted. By default the attenuation is set to auto.

[@zynthos9]

at 5:32 why does the line for Third Order Product not start at the intersection of the two axes, but rather later down the X axis?

[@pauldenisowski]

The third order product only appears above a certain input power level (X-axis). The more linear a device is, the higher the input power level that's required before third order products appear. All other things being equal, a third order product line that starts further down the x-axis will produce a higher TOI / IP3 value (a good thing). If the fundamental and third order product both started at the origin, then IP3 = 0 dB and the amplitude of the third order product would be higher than the amplitude of the fundamental for *all* input power values (a very, very bad thing).

[@pratwurschtgulasch6662]

i have only one question, why is your equipment so freaking expensive?