Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

donate to eham
   Home   Help Search  
Pages: [1] 2 Next   Go Down
  Print  
Author Topic: The IMD Debate: Please Contribute :)  (Read 7496 times)
VK3BL
Member

Posts: 1496


WWW

Ignore
« on: November 29, 2017, 12:20:31 AM »

IMD is such a loaded topic.  

I have personally measured my ANAN 100D at -55 IMD3 with the SDR Play (which is actually pretty decently calibrated), but had a friend down the road report my 'real' IMD3 as -30.

Why? Distortion was being generated in either his antenna system or my own.

I found evidence of intermittent arcing in an N Connector used outside, and promptly replaced it; the fact of the matter is I intend to move to 7/16 DIN for outdoor use when any length of coax feed line is subject to a vSWR above 5:1; it is very easy to overlook the fact that voice peaks in SSB can generate very high voltages on a transmission line, even if the loss is not worth worrying about due to the length.

In my case, I was feeding ladder line with a 10 ft section of LDF-450; I couldn't care less about the loss, but the vSWR was enough to intermittently cause arcing at the balun junction point N connector.  

Even at 400 Watts PEP on a 50 ohm line, a 1:1 vSWR will result in peaks of 141 volts.  Multiply this by 5 (vSWR 5:1), and you have 700 volt peaks; its easy to see how a dirty, broken, slightly wet or frankly cheap connector can cause rectification via arcing.

QRO and insufficient / poor choice of connectors should get equally as much attention in 'all knobs to the right' syndrome, IMHO.

On another occasion my 'monitoring receiver' was reporting an IMD3 of -20.  I tore my antenna system apart looking for a fault, only to find that the antenna attached to the SMA connector on the SDR Play had come loose... The second I tightened it back up, everything was rosy.  

There is a reason professional grade connector manufacturers offer specification sheets and torque calibrated tools.

We hams need to start paying attention to these things if we want to put out clean signals at high power; rather than just assuming 'all knobs are to the right' or 'tetrodes are always dirty'.  We can and should help each other.

The first thing we should start doing is STOP dropping the v from vSWR.
Logged

J.D. Mitchell BA  - VK3BL / XU7AGA - https://www.youtube.com/ratemyradio
N8FVJ
Member

Posts: 602




Ignore
« Reply #1 on: November 29, 2017, 06:20:28 AM »

You have to start out with a clean IMD transmitter. The new Kenwood TS-590SG is very clean at -42dB 3rd order. The Flex transceivers are also very good. Most Icom & Yaesu radios are poor in this regard. The IC-7300 is only -30dB third order.
Logged
VK3BL
Member

Posts: 1496


WWW

Ignore
« Reply #2 on: November 29, 2017, 12:24:32 PM »

You have to start out with a clean IMD transmitter. The new Kenwood TS-590SG is very clean at -42dB 3rd order. The Flex transceivers are also very good. Most Icom & Yaesu radios are poor in this regard. The IC-7300 is only -30dB third order.

Yes of course, such as one needs a fast car to win the Grand Prix.

My point really was you still need to learn how to drive fast.


Re: IC-7300.

I could get ~35dB down on peak voice tone in 16 sample FFT Averaged SSB use; thats pretty clean in my books.  The trick is, as Tom W8JI points out, don't boost your bass and treble.  On the other hand, no matter what I did, I couldn't get it worse than -30dB down IMD3. 

Bear in mind I am talking real world, peak voice tone vs out of band distortion product, not two tone, "lab testing".
Logged

J.D. Mitchell BA  - VK3BL / XU7AGA - https://www.youtube.com/ratemyradio
AC7CW
Member

Posts: 1099




Ignore
« Reply #3 on: December 02, 2017, 10:22:15 AM »

I never seriously thought about antenna/feedline voltage until I investigated magloops. Voltages can run in the thousands. Vacuum variable capacitors get onto the parts lists and thank goodness for Russian surplus on eBay...

95% of all failed electronics things in general are due to electromechanical failures. That points to the idea that the aforementioned need for attention to connectors is indeed a good thing, not to mention switches and anything outdoors in wind and weather.

I guess that if I'm on the air with my club remote station I can use my SDR radio's bandscope to check the signal quality... should be fun, no end to the fun in this hobby.
Logged

Novice 1958, 20WPM Extra now... (and get off my lawn)
DL8OV
Member

Posts: 800




Ignore
« Reply #4 on: December 07, 2017, 12:47:11 PM »

I think we may be getting lost in the abbreviation soup here.

IMD are the close in unwanted products, normally seen at your carrier frequency +/- your modulation frequency. If I want to check for IMD then my spectrum analyzer normally has a sweep width of 50 KHz. A signal with good IMD looks like a church steeple whilst a bad signal looks like an airplane wing.

If you want to built a transmitter with good IMD then you need a holistic approach and look at every step in the signal chain from the microphone to the low-pass filters, this takes time and it takes a lot of patience. After three years of work I have a transmit chain which I am happy with but I doubt that any of the big box manufacturers would bother.

Poor contacts on connectors or antenna switches will act like diodes and cause harmonics. This MIGHT affect your IMD but the results of a poor antenna connection will be seen over many MHz, I've personally seen 9th harmonics from a 14 MHz signal, that's nearly 130 MHz!

So, poor contacts, arcing, are we talking about harmonics or IMD here?

Peter DL8OV
Logged
VK3BL
Member

Posts: 1496


WWW

Ignore
« Reply #5 on: December 09, 2017, 04:09:05 PM »


So, poor contacts, arcing, are we talking about harmonics or IMD here?

Peter DL8OV

Both, if a poor antenna contact acts as a nonlinear system component, it by definition is capable of producing IMD.
Logged

J.D. Mitchell BA  - VK3BL / XU7AGA - https://www.youtube.com/ratemyradio
WB4SPT
Member

Posts: 593




Ignore
« Reply #6 on: December 15, 2017, 12:03:11 PM »

Even at 400 Watts PEP on a 50 ohm line, a 1:1 vSWR will result in peaks of 141 volts.  Multiply this by 5 (vSWR 5:1), and you have 700 volt peaks;

We hams need to start paying attention to these things...
The first thing we should start doing is STOP dropping the v from vSWR.

That's not how the math works on VSWR.  Or ISWR either.   Grin

And, "V" is capitalized...
Logged
AF6LJ
Member

Posts: 464




Ignore
« Reply #7 on: December 15, 2017, 12:42:15 PM »

This is an interesting thread.
Measuring IMD isn't as straight forward as most would like to think. SDRs may not be the ideal choice for those measurements and one must be careful using an analogue spectrum analyzer making such measurements as they can be overloaded well before first mixer overload occurs.
I've done plenty of measurements with analogue analyzers, enough to know that how the test environment is set up is way too important to take casually. I would assume that the same holds true for the modern digital analyzers. As far as SDRs goes...
To put it bluntly; making any claims based on what a computer A to D, or a 10 or 12 bot run of the mill ADC presents is scary to say the least. The A-D subsystems built into modern analyzers are far more trustworthy.
Logged

Take Care
Sue,
AF6LJ
G3RZP
Member

Posts: 28




Ignore
« Reply #8 on: December 16, 2017, 01:12:27 AM »

WB4SPT is correct.

Take a lossless 50 ohm line and leave the end open or put a short on it. Feed a 100 volt peak sine wave into it, and because the SWR is infinite, the 100 volts are reflected. If the reflected volts add in phase, you then get 200 volts peak. In other places along that transmission line where the forward and reflected volts are in anti-phase, you get zero volts.
Logged
VK3BL
Member

Posts: 1496


WWW

Ignore
« Reply #9 on: December 18, 2017, 03:04:06 AM »

WB4SPT is correct.

Take a lossless 50 ohm line and leave the end open or put a short on it. Feed a 100 volt peak sine wave into it, and because the SWR is infinite, the 100 volts are reflected. If the reflected volts add in phase, you then get 200 volts peak. In other places along that transmission line where the forward and reflected volts are in anti-phase, you get zero volts.

More than happy to be corrected, but my understanding of shorted and open transmission lines are that they represent unique cases.

With reference to this article: https://www.w8ji.com/vswr_reactive_power.htm, please flush out your critique; my experience as well as many others demonstrates rather plainly that it is not as simple as saying "Infinite:1 SWR = 2x Voltage, worst case", as implied by your critique.

Expressing mathematical constructs using plain English is always fraught with danger, but I posit thinking of '5:1 = 5x the maximum voltage' is more helpful than 'infinite:1 = 2x the voltage', even if the later is technically more correct, albeit limited to special cases such as shorted or open lossless lines of uniform impedance.




« Last Edit: December 18, 2017, 03:16:26 AM by VK3BL » Logged

J.D. Mitchell BA  - VK3BL / XU7AGA - https://www.youtube.com/ratemyradio
W9IQ
Member

Posts: 1845




Ignore
« Reply #10 on: December 18, 2017, 10:29:51 AM »

Quote
The first thing we should start doing is STOP dropping the v from vSWR.

The popularization of the term VSWR originates from the days when SWR was measured by probing the feedline with an RF voltmeter to find the minimum and maximum voltages. Had this been done with RF current meters, we would likely have been saddled with the term ISWR instead of VSWR. But the resulting value would be the same.

Today's ham SWR meters do not measure the VSWR - contrary to many fanciful, but erroneous, descriptions of separating the forward and reflected voltage waves. They do no such thing. Instead they sample the complex voltage and current at the point of insertion and then calculate (either in analog or digital form) the SWR for a given characteristic impedance.

So in many ways, simply saying SWR is a more nearly accurate description of what is being reported. Of course the debate can continue ad infinitum since SWR = VSWR = ISWR.

- Glenn W9IQ
« Last Edit: December 18, 2017, 10:34:05 AM by W9IQ » Logged

- Glenn W9IQ

I never make a mistake. I thought I did once but I was wrong.
W9IQ
Member

Posts: 1845




Ignore
« Reply #11 on: December 18, 2017, 11:05:07 AM »

The peak voltage that could potentially develop on a transmission line is given as:

Epeak=SQRT(Pforward*ZO*SWR)*1.414

Note that multiplying power times impedance and taking its square root is the equivalence of voltage. If the transmission line is too lossy or too short, this peak voltage may not develop but the formula accurately reports the worst (highest voltage) case.

- Glenn W9IQ
Logged

- Glenn W9IQ

I never make a mistake. I thought I did once but I was wrong.
VK3BL
Member

Posts: 1496


WWW

Ignore
« Reply #12 on: December 19, 2017, 11:29:57 AM »

The peak voltage that could potentially develop on a transmission line is given as:

Epeak=SQRT(Pforward*ZO*SWR)*1.414

Note that multiplying power times impedance and taking its square root is the equivalence of voltage. If the transmission line is too lossy or too short, this peak voltage may not develop but the formula accurately reports the worst (highest voltage) case.

- Glenn W9IQ

Thanks Glenn,

I believe even I can do that math Wink
Logged

J.D. Mitchell BA  - VK3BL / XU7AGA - https://www.youtube.com/ratemyradio
K6AER
Member

Posts: 4815




Ignore
« Reply #13 on: January 02, 2018, 08:20:57 PM »

You have to start out with a clean IMD transmitter. The new Kenwood TS-590SG is very clean at -42dB 3rd order. The Flex transceivers are also very good. Most Icom & Yaesu radios are poor in this regard. The IC-7300 is only -30dB third order.

Was that IMD of -42 dB at 100 watts out?
Logged
VK3BL
Member

Posts: 1496


WWW

Ignore
« Reply #14 on: January 02, 2018, 10:13:00 PM »

You have to start out with a clean IMD transmitter. The new Kenwood TS-590SG is very clean at -42dB 3rd order. The Flex transceivers are also very good. Most Icom & Yaesu radios are poor in this regard. The IC-7300 is only -30dB third order.

Was that IMD of -42 dB at 100 watts out?

ARRL PEP figures maybe... that would be -36dBc IMD3... but at 100 Watts out IMD5 is the dominant product so its likely the results were cherry picked and -6dB "PEP'd".

I doubt any 12v exciter is much cleaner than any other at 100W out; -30dBc give or take a dB with "optimisation" seems to be about the limit.  Thats ~0.15% THD in Audio terms, which is pretty clean in the scheme of things.

My own testing has shown you'll put out a NARROWER signal driving a linear at 25-50 watts with an exciter than you will running an exciter flat out!  The key is to keep IMD5 and higher order products down, which as long as you don't drive the exciter or linear flat out you will achieve!

The point of this thread though is that if your antenna system has issues such as arcing or is too close to metal gutters etc, you can pump it with -50dBc of pre-distorted signal and not achieve anywhere near that in practice.  Looks great going into a dummy load but your friends won't be as kind...


« Last Edit: January 02, 2018, 10:18:52 PM by VK3BL » Logged

J.D. Mitchell BA  - VK3BL / XU7AGA - https://www.youtube.com/ratemyradio
Pages: [1] 2 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!