Forum

Anyone out there have a link that explains EVM from a "graphical" perspective?

Every site that I can find talks about the definition of EVM from a vector stand point, and the various calculations that can be performed, but they don't include diagrams to go along with them - other than maybe a single point.

I know that people really familiar with these graphs can tell an awful lot just by looking at them. eg phase error, compression, noise, etc.

What I'm looking for is a set of diagrams showing EVM diagrams, and the  problems that create the patterns on them.  One set might have increasing amounts of phase shift, for example   - a Java simulator would be fantastic.

I have access to an Anritsu MT8860B which shows the graphs, and the calculation results, and I'd really like the knowledge to tie it all together.

Thanks in advance.

Not sure it this will help.  BTW, I believe most or all of the Anritsu's are supported in Labview.

http://zone.ni.com/devzone/cda/tut/p/id/650

http://zone.ni.com/devzone/cda/tut/p/id/3652

http://zone.ni.com/wv/app/doc/p/id/wv-689

Thanks for the links Jon.   

I already have the mathematical formulas for EVM, and I'm a visual kind of guy.  What I was looking for was a video, or simulation, showing (say) vector points showing up on an EVM diagram and relating their pattern to the problems, or lack thereof, that they are in the hardware. This may be exactly what's on the video link you posted, but I can't view it from work - I'll try at home tonight.

In the meantime -  yes, both Anritsu Test (Wi-Fi and Bluetooth) sets I'm familiar with can be run with Labview.    We use the programs written by Anritsu, as we don't have the resources available to write our own.   They do supply the source for both  programs if you should care to modify them.  The Wi-Fi unit is the MT8860C, and the Bluetooth tester is a MT8852B.

Anritsu supplies two programs with their MT8860(B/C) units.    One is called LANLook, the other is LANTest.   Lanlook is meant for R&D, or QA testing, and runs one test at a time.  Lantest is more for Go/No-Go testing using user controlled scripts, and you can esentially test everything b & g in one shot.  I usually run LANlook.

LanLook lets you run almost every IEEE 802.11 test possible.   For example on 802.11b it can measure Transmit Power Levels, Power Level Control, Spectrum Masks, Center Frequency and Chip Clock Tolerances, Power on and off Ramp up/down times, Modulation Accuracy, Receiver Senstitivy, Max. Input level, and Adjacent Channel Rejection.  For 802.11a/g it can measure Non-Adjacent Channel Rejection, Symbol Clock Tolerances, Center Frequency Leakage, Spectral Flatness, Constellation Error (EVM) and the "g" analogues of the "b" measurements of course.   They now have an "n" option available, but I am not familiar with that version.

Their programs are sometimes a little quirky, but they produce rock solid results.   The unit is perfect for people who don't already have RF Amplifiers, Spectrum Analysers, and RF Power Meters, etc. or who are not RF Engineers (like me).  It gives you everything except a PC, in a single box.  

Some of our departments have all that equipment, but I'd have to travel 3,000 miles to use it.

Depending on exactly what I'm doing, I can use mine with an ETS-Lindgren RF Enclosure at my lab desk, in one of our walk-in RF Screen rooms, or in the companies 3-meter compact Semi-Anechoic chamber.  LANlook can run in Simulation mode too, without a connection to a MT8860, if you should want to install it and see what it can do.   

Thanks again.

I remain envious of your lab. That is probably the crayiest question I've heard on the forums.  You should post something once you get the answer you're looking for.

That's what I plan on doing.

Don't be too envious - there are things like Network Analysers that I don't have local access to that would sure make life easier.  Yah, I could rent it, but then it has to be approved etc and usually won't get done fast enough.

I watched the wv-689 video link above, and thought it was well done.  I understood /E/ already.

It's still not what I'm looking for.  I want to see all the problems with dots (sample points), not just phase error.    And what causes it/them.

Thanks a lot.   I'll post more as  I find it.