[Amps] Measuring IMD

Mon Nov 4 14:16:08 EST 2013

Ive used two methods over the years to feed QRO power into a SA.

One is to start with a directional coupler that has 20 dB or so at the 
frequency of interest. More is better. The piece of coax thru a toroid 
wrapped with wire in a lot of amps, ATU's, wattmeters, etc is a directional 
coupler. I used a surplus HP coupler that covered 100-250 mc (that alone 
shows its age) that was a freebie at a company thowing stuff away.
It was fairly bulky, passed 2500W+ with ease and had about 33dB attenuation 
from the forward port on 20M. The reverse port has a generic 10W attenuator 
screwed on to the N connector. Actual attenuation is not important, you just 
want a sufficiently reduced signal so you can use small (inexpensive) in 
line attenuators to get to a safe level for the SA as well as provide it 
with a solid 50 Ohm load.

As mentioned, the coupler directivity is frequency dependent.

The other is the Bird style plug in sampler which offers 50dB attenuation 
25-1000MHz, and unspcified below but rated only to 500W.
I use one for convenience in testing transmitters and still use a pad on the 
SA.

There are several other RF samplers on the market that while not designed 
for it can be used at HF.

Carl
KM1H

----- Original Message ----- 
From: "Robert Groh" <rgroh at swbell.net>
To: "Dave Trainor" <dave at n8zfm.com>; <amps at contesting.com>
Sent: Monday, November 04, 2013 10:11 AM
Subject: Re: [Amps] Measuring IMD

Dave,

You definitely want to reduce the power input to the spectrum analyzer down 
to something in the 0 dBm to +10 dBm level. I am not familiar with the HP 
8924() family but I have had, in my prior life as a radio design engineer (I 
am now retired), extensive experience with spectrum analyzers, IMD 
measurement, etc.

The required power attenuation can be obtained in a number of ways - use a 
'probe' (e.g. like Bird wattmeter's do), use small capacitor for decoupling, 
use a power attenuator, etc.

The probe, capacitor, etc techniques are frequency dependent (e.g. the 
output of the sampler decreases as frequency does down) but that won't 
affect the accuracy of your IMD measurements because your are measuring IMD 
over a very small frequency range (i.e. a couple of kHz).

If you use a good power attenuator (e.g. Bird) then you don't have to worry 
(much) about frequency variations at all. As a side note, you can pick up 
high power 50 ohm attenuators (on eBay, etc) for not a lot of money - I have 
a 40 dB, 250W Bird attenuator I got for $60 or so - a 40 dB attenuator is a 
darn decent load all by itself and the 250W rating is a continuous duty 
rating - the peak rating is determined by the voltage breakdown of the 
connectors and load which is quite a bit higher than 250W (figured that out 
once but have forgotten the details).

If you designed the system for (example) 2,000 watts and want a maximum 
input signal to the SA of 0 dBm then you need 63 dB of attenuation (i.e. 
2,000 watts = +63 dBm = +33 dBw).

As a side note, I would worry a bit about the 'linearity' of the Heathkit 
load when being pounded as you are talking about doing.

Much more to say but I won't burden the forum with it at this time. If you 
have any other questions, you can contact me directly and I'll be glad to 
give you some more info. You can also call via landline if you want (my 
address is in the Callbook)

73
Bob Groh, WA2CKY

________________________________
 From: Dave Trainor <dave at n8zfm.com>
To: "amps at contesting.com" <amps at contesting.com>
Sent: Sunday, November 3, 2013 8:51 PM
Subject: Re: [Amps] Measuring IMD

So here is a newby question, Let's say I have a HP8924C SA and it has a max 
of 60W input, and I want to check IMD on a 1.5KW amplifier. Do I do this 
with a T connector in the line to the dummy load? Or do I need to simply 
pick off a sample of the RF with a mica cap from the dummy load? I have one 
of the old "Cantenna" dummy loads I use and could add something to pick off 
a sample. What I don't want to do is somehow burn out my analyzer by making 
a silly mistake. I usually use it for repair/alignment work on VHF and have 
never tried to measure IMD on an HF amp but since I am in the process of 
collecting the parts I need for a tetrode amp, measuring the IMD to be sure 
it's clean is now something I'd like to do, and I am not quite sure how to 
safely measure that amount of RF, or if picking off a sample would not be a 
valid testing method.

73 - Dave N8ZFM

-----Original Message-----
From: Amps [mailto:amps-bounces at contesting.com] On Behalf Of peter chadwick
Sent: Sunday, November 03, 2013 4:26 PM
To: amps at contesting.com
Subject: Re: [Amps] Measuring IMD

The reference given to Pappenfus gives the equations for determining the 
level of 3rd and 5th order products from the curvature of the line.

Noise testing is good, provided you are sure about the peak to average 
ratio: to that end, noise with a notch in the middle of the AF passband 
leading to measurement of the IMD power in that notch is good. The technique 
was used over 50 years ago in measuring FDM telephone systems, and was 
applied to multichannel HF SSB tx's.

The question at the end of the day is 'How much use is the information about 
more accurate levels of IMD measurement except as an academic exercise?' 
More to the point is the IMD power in a SSB bandwidth at various offsets - 
as SM5BSZ has measured. Plus the fact that the vast majority of amateurs 
neither understand or care about how bad their tx is to the last few dB!

73

Peter G3RZP
_______________________________________________
Amps mailing list
Amps at contesting.com
http://lists.contesting.com/mailman/listinfo/amps

--
This message was scanned by ESVA and is believed to be clean.
Click here to report this message as spam.
http://h0stname/cgi-bin/learn-msg.cgi?id=EB92528C44.DA4AC

_______________________________________________
Amps mailing list
Amps at contesting.com
http://lists.contesting.com/mailman/listinfo/amps
_______________________________________________
Amps mailing list
Amps at contesting.com
http://lists.contesting.com/mailman/listinfo/amps

-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 10.0.1432 / Virus Database: 3222/6307 - Release Date: 11/04/13