[Amps] Checking for IMD

[Gary Schafer]

See below:

> -----Original Message-----
> From: amps-bounces at contesting.com [mailto:amps-bounces at contesting.com] On
> Behalf Of Tom W8JI
> Sent: Sunday, August 13, 2006 11:17 PM
> To: amps at contesting.com
> Subject: Re: [Amps] Checking for IMD
> 
> > So from the above the IM products from 300 and 3000 Hz
> > tones are:
> > 3rd = 5.7 KHz and -2.4 KHz.
> > 5th = 8.4 KHz and -5.1 KHz.
> > 7th = 11.1 KHz and -7.8 KHz.
> >
> > All of the negative products represent the products that
> > fall on the
> > opposite side band.
> 
> The extremes of IM3 product are the *highest frequency
> transmitted* plus 2.7 kHz and the *lowest frequency
> transmitted* minus 2.7kHz.
> 
> Consider a 7 MHz USB signal . The upper IM3 would be 7.003
> plus .0027 = 7.0057
> The lower IM3 would be 7.0003-.0027= 6.9976

Yes tom, you come up with the same numbers that I do. Plus 5.7 KHz and minus
2.4 KHz from the carrier frequency.

> 
> If we had a receiver at 7.0000 with EXACTLY the same carrier
> offset and filter BW the LSB passband on a sideband switch
> would be 6.9997 down to 6.997.
> The highest tone spread IM3 product from the TX would fall
> at 6.9976  and be covered. The lowest tone IM3 found on LSB
> could be well above 6.9997.
> 
> The passband of the receiver does not exactly align with the
> IM3 range because of the 300Hz lowest tone offset.
> 
> Most likely all the closer spaced IM5 and higher will fall
> within the passband, but none of the wider ones will.

We really don't care with this kind of test if they all fall in the pass
band or not. The whole method of using voice for IM measurements is not a
very precise one to start with.

The opposite side band tells us just as much as does tuning to another
frequency to look at IM products. Neither gets them all.

> 
> We really can't say this is any kind of IM measurement. It
> simply checks opposite sideband with whatever peak
> accumulated power the S meter is detecting. Any good AGC
> detects the peaks and holds the sample, but also has some
> tolerance to ignore very short momentary peaks. So the S
> meter is an almost peak, but not totally peak, indicator.
> 
> It definitely isn't anywhere near average unless the
> receiver has almost useless AGC.

It is as Tony said much the same measurement that is done on the wanted part
of the signal so the comparison of IM product levels to wanted signal levels
should be somewhat valid. Although there may well be less dense IM products
than there are of the wanted signal. 

> 
> This thing really needs some thought as to what to call it.
> 
> As for receivers, many are terrible for IM within the
> roofing filter BW. The FT757 is around 55dB IM3 DR with two
> pure tones. The Collins 75S3 around 60dB (no roofing filter
> so that is 3kHz spaced IM3). An R4C would be useless for
> this method. While fortunately they are mostly gone, it
> still illustrates we have to know the receiver.
> 
> The problem is IM3 DR is measured by audio and not by S
> meter reading. The fact the meter does not move at the
> bottom end pushes the lower end of threshold up. It isn't
> good enough to just hear the signal, now it has to move the
> S meter to a useful level to be a useful signal.
> 
> I'm not sure what that would do to the useful dynamic range,
> but at first glance it doesn't seem good. My 751A for
> example has a noise floor about 20-30dB below the lowest
> possible S meter reading. Any IM3 DR test I would do would
> look for IM3 products near that noise floor. An IM3
> transmitter measurement would require a starting level at
> least 30dB higher than noise floor. That will take DR away
> from the receiver.
> 
> Do you follow what I am saying Gary?

Yes Tom I understand what you are saying but I don't think that I have seen
a receiver that I can feed a signal generator into, hear the signal and be
able to increase the signal by another 30 dB before the S meter moves. That
seems like an awfully high agc threshold?

> 
> I sure hope people don't use this method over the air unless
> they fully understand what they are doing. It looks like the
> initial thoughts didn't even understand that S meters read
> the peak power, not the average.

I disagree. I would hope that more people would take a listen to the other
side band on signals. It is a quick and easy way to get some idea of how
clean the signal is. Not an absolute by any means but with a little practice
and common sense it could become a valuable tool in helping to clean up the
bands a little.

73
Gary  K4FMX

> 
> 73 Tom