W1VTP class E transmitter harmonic test

[w1vtp]

I finally got around to test my new class E transmitter for harmonics.  The results are quite good, I think.  I have included the results of my Excel spreadsheet for your viewing pleasure

I think Steve, QIX has done a great job in design and Wayne SSJ a great job in its build.

Page 1 has the test results.  Page 2 shows the test hookup.  The Flex's second receiver did a nice job of isolating its RX 2 input from the 1 watt carrier output from the main input / output

Al

[N2DTS]

What about a 2 tone intermod test?

[w1vtp]

What about a 2 tone intermod test?

Don't have a good enuf AF oscillator(s) for that test.  I'd just be testing the distortion products of the AF Oscillator.  I'll work on it.  The Pulse Duration Modulator should be pretty good.

I seem to recall that Tom, JJ was looking for such an oscillator because of the same issue of the low distortion of the modulator. Wonder if such a circuit was ever found

Al

[K1JJ]

Don't have a good enuf AF oscillator(s) for that test.  I'd just be testing the distortion products of the AF Oscillator.  I'll work on it.  The Pulse Duration Modulator should be pretty good.

I seem to recall that Tom, JJ was looking for such an oscillator because of the same issue of the low distortion of the modulator. Wonder if such a circuit was ever found

Al

Hi Al,

Back last winter when I did the IMD testing on the class A driver / 8877 AB1 linear system, I bought an Elecraft 2T two-tone tester board.  It works very well and is clean enuff for testing even my 4CX-350J driver on ssb.   They are a simple kit board.  You need only one tone for the AM THD test, which can be toggled from the board.

I posted my one-tone and two-tone performance for the old 24 pill class e rig in the thread below. (THD and IMD) I found the performance to be about average, around -25db down 3rd at 100%. The tube rigs - both linear amps and plate modulated 4X1 rig were somewhat cleaner, between -30 to -50dB 3rd at 100%, depending on the lashup. The audio negative  feedback made all the difference on the 4X1 rig. I understand to add NFB to the PDM class e board would be quite a task.  But again, I was satisfied with the class e pdm modulator stock performance.

The second page has the best info:
http://amfone.net/Amforum/index.php?topic=33555.0


As a reference, here is the IMD testing thread for the linear amp system. I still get excited seeing -50dB+ 3rd order in the final tests.    It's hard to beat an optimized AB1 linear system for IMD performance.
http://amfone.net/Amforum/index.php?topic=33505.0


I'll be curious to see your IMD data.  Running it up to full power at 150% positive should show you worst case, full strap performance.

T

[N2DTS]

Would not a tone generator in a laptop work?
I used a free program (NCH tone generator) to check out the IM on my rigs.
Why would it be lower on a class E setup?
 

[Steve - K4HX]

Less linear modulation.

[K1JJ]

Sure, a computer tone program should work FB. It all depends on how clean the computer audio board is. You need something that will be -55dB 3rd or better.  This is easy to find in low level chips, but I don't know about the low end computers.  

I use an outboard 2-tone board for simplicity.


T

[W2NBC]

Hello Folks..

A tone test is a tone test PERIOD.. Any impedance higher than the source is FB, IF YOU HAVE THE GAIN.  even to "smug" Flex owners..

The distortion from online programs like Audacity etc. are NIL..

Sweep your junk with 30Hz to 10K.. see a roll-off? where, how much.. ? make notes..

got a hardware old buzzard audio generator ? same deal.. chances are the impedance will match just fine..


73,

Jeff W2NBC

[steve_qix]

There is one test that very few AM transmitters pass,  and that is the 400Hz triangle wave test.  If the rig can do this, it will almost always sound really good and be able to pass, unmodified, most waveforms.  Another good one is a square wave test.

Easy to perform these tests and a very good indicator of how good it will sound.

[N2DTS]

Is there a worry about the mod transformer with those tests?
I once blew up the mod iron in my 30K1 doing full power audio tests...

[W2NBC]

Reading through the Optimod 9200 manual and calibration using the built-in square wave generator (that I used once on the BC-610  ) they warn not to exceed 50% modulation during the test, and to let the transmitter cool down between short adjustments. Yes, absolutely the best and effective way as Steve mentioned to utilize a tone generator to see any problems in the waveform. The caveat is that some have lost mod transformers during the process and THAT would ruin your day..

 Out of the manual, and interesting is this line:

"Adjustment of the high frequency transmitter equalizer controls cannot be
done into a dummy load, because the transmitter will overshoot and ring
differently when loaded by the reactance of your antenna system."

Tests into a reactive load may change some results in all kinds of testing..

[Steve - K4HX]

ANd that's for BC transmitters. Most ham grade transmitter should not exceed 50% modulation when doing sinewave testing, except for very short periods.

[w1vtp]

There is one test that very few AM transmitters pass,  and that is the 400Hz triangle wave test.  If the rig can do this, it will almost always sound really good and be able to pass, unmodified, most waveforms.  Another good one is a square wave test.

Easy to perform these tests and a very good indicator of how good it will sound.

I can do that triangle test with my HP function generator.  Later some time

Getting back to the original topic of this thread: I think the Flex did a fine job with that harmonic test.  Testing the fundamental by toggling the 20 dB Kay step attenuator yielded VERY accurate results showing the Flex receiver can double as a lab instrument. -10 dBm is a mighty strong signal - almost 70 dB over S9

Thanks guys for all your cool comments.

Al

Al

[w1vtp]

ANd that's for BC transmitters. Most ham grade transmitter should not exceed 50% modulation when doing sinewave testing, except for very short periods.

Steve:  perhaps you recall my THD testing of my EICO 720 a few years ago.  That was done at 100% modulation at extended periods of time.  The old girl and the Bogen M60 and the Hammond audio transformer performed without a hitch.  Probably due to overdesign of the components.  I still haven't put that combo on the air

Al

[K1JJ]

Yes, it pays to be careful and not pop a mod transformer during tests. I've had it happen to me once when testing the 813s X 813s plate modulated rig. It was a beautiful 100 pound BC mod transformer.  When running a 1 KHz sine tone at 120% positive - limited to 90% negative, the input connection weld to the primary opened up.  I was able to pull the cover off and repair it. But learned my lesson.

As far as testing, there are many techniques and objectives.  A 400 Hz triangle test is certainly a difficult test for the "average"  transformer-type AM rig to pass. A 100 Hz triangle is even harder and was difficult for my own 4X1 rig, even with generous negative feedback.  The transformer-less class E PDM rig passed the triangles without much problem, of course.

But the AM purity test that takes no prisoners is putting a 1 KHz tone through and looking at the spectrum analyzer for THD and IMD.  The 3rd and 5th order IMD peaks, especially,  tell how much crud we are sending up the band.  As we have all seen, we can have a beautiful sounding transmitter on frequency, but the crud level takes out our band neighbors. Poor IMD or THD can be from poor design, poor adjustments or simply a problem with bad parts, etc.  But once we get the rig to pass the general sine and triangle tone tests and sound good -  THEN  look to achieve at LEAST -25dB 3rd order IMD or better numbers at 100% positive modulation.  

The alternative is fooling ourselves in denial if someone complains of splatter.   It's no fun to have a splatter problem and not know what to do about it. I've been there.  The spectrum analyzer test gives us the reference we need to proceed and improve it to acceptable tolerance. We now have a precise, quantitative method to compare apples to apples and see small incremental improvements.  A simple $25 SoftRock connected to the computer is all we need.  

The louder we are, the more important good IMD and THD numbers become.

T

[N2DTS]

There are a LOT of poor signals on the air.
A lot of people run old stuff which was never very good, and now its old.
Lots of people have no scope, mod monitor or sdr, and old gear.

I would find it very hard to operate without knowing what was going out on the air.