[Amps] Measuring IMD in a home workshop

[Jim Thomson]

Date: Wed, 22 May 2019 14:23:12 -0700
From: "Bob Wilson, N6TV" <n6tv at arrl.net>
To: Amps Reflector <amps at contesting.com>
Subject: Re: [Amps] Measuring IMD in a home workshop

<The Elecraft K3 also has a built-in 2 TONE generator.

<Switch to SSB mode
<Set *CONFIG:TECH MD ON*
<Set *CONFIG:2 TONE ON*
<Exit the menu and tap the *XMIT* button.  You can vary the amplitude of one
<of the tones using the MIC knob; the other is fixed.

<73,
<Bob, N6TV

##  Nice try...it wont  work !    The  IMD  from a K3  is an abysmal –24 db  below one tone....
which is on a par with  sweep  tubes from the  70s.   All  that the  2 tone generator thats  built into the
K3  will  achieve... is an easy method to measure the  K3s  IMD, thats  it. 

##  IE:  to test the IMD  from a 1.5 kw linear, the driving source has to be a LOT cleaner than the amp under test. 
You cant use the  2 tone  /  100  w pep out from a K3,  to test  IMD on a L4B amp with its  2 x 3-500Zs. 

##  something like an  ANAN  7000 / 8000, using pure signal will  work.  They are typ  -70 for  IMD3 + 

##  So will a yaesu  running Class  A.    Class  A  is very clean. –70  to –85  for  IMD3+ 

##  another method is  to use  two xcvrs,  each outputting a CW  cxr.... then combine the  2 x signals
with a hybrid combiner.   2  tone spacing would be equal to the  difference between the  2 x VFOs. 

##  My problem with the  2 tone method is  everything is In a steady state.... like  B+, bias, drive,  ALC, etc, etc. 
IE:  all loads are static, and not dynamic...like normal  ssb. 

###  Having said all that,  I have  two  FT-1000-MK-V’s.     I TX  on one of them.... while listening a bit off freq  on the
2nd xcvr.   I then  toggle the  TX  between Class  AB.... and the  cleaner  class  A..... and the  difference in imd / splatter
while listening on a 2nd  xcvr is blatantly obvious.   I tune  2nd xcvr just off freq enough, so no splatter is heard.... while using 
Class  A  on  tx.    Then swap to  class  AB... and yoza.... splatter galore.  

##  Unless the xcvr is something like the  ANAN, using  Pure  signal,,,, or Class  A  with the yaesu,  its all a moot point really. 
Any decent linear  TUBE  amp  will have far better  IMD  vs the  typ  xcvr... like the K3..and  every other  xcvr out there. 

##  Most  SS  xcvrs  will have cleaner  IMD  when the PO  is reduced a bit, and depending on drive required for  the linear amp
under test, that could easily affect results.   Big difference between  100w drive to a 2 x 3-500Z  amp... and  40-50  watts  into
a  8877

##  SS  1.5 kw amps  are a whole  different ballgame.   IMD is typ not  good, on older models..when run at max rated power. 

##  still another method is to monitor the Tx  ssb  signal on a spectrum analyzer...or sdr rx.   Assuming a semi clean xcvr,
you can then see how wide the  ssb  signal is...when barefoot..or with external linear amp on.  

##  yaesu, kenwood,  icom,  elecraft, all need to get with the program, and implement some form of  either analog or digital 
pre-distortion techniques in future  xcvrs.   Then the sample can be taken from the  1.5 kw  amp itself... fed back to the xcvr,
the the  pre-distortion  concept  will correct for non linearity  in the external amp as well.   Folks using  ANAN  xcvrs   driving
external  linear amps  are all having good success.  IMD on the external amp is much improved. 

##  Big difference between an  ANAN   7000 / 8000  with its  -80 db  imd  vs  -24 db  imd  from a K3.   Apples +  Oranges.   

Jim   VE7RF