The AM Forum

I would appreciate recommendations on any comprehensive article(s) or book(s) covering measurement and specification of the sensitivity of hf receivers, and the related use of signal generators, "forcing the match" with attenuator pads, standard (dummy) antenna simulators, and so on.

Thanks, all!

-al hart, w8vr

Try to find a copy of the ARRL Solid State Handbood a good reference

There is no single source.

Some good reading:

Radiotron Designers Handbook 4th Edition

Tube era ARRL and West Coast (W6SAI and other editors) Handbooks

QEX and Communications Quarterly had excellent articles by Rhode and others

Tech manuals for actual receivers and signal generators went into great detail, especially military and HP manuals.

Hi Al. Forcing the match by adding pads etc... sounds like a bad thing, however it is not. Although today modern signal generators have a very respective output impedance, nearly 50 ohms, older units were not as well controlled. So a SOP is to add a 3 or 6 dB pad to the output of the sig gen. In fact some of the earlier sig gens had their internal variable attenuator calibrated to read open circuit voltage (uV).  The use of a 6 dB pad was appropriate as now the reading was true AT the Rx under test. If you plan on doing IMD, desense tests, etc... padding a sig gen is also appropriate as this minimizes intermodulation distortion created by RF level from sig gen 1 modulating the ALC of sig gen 2 and visa-versa. The famous HP 8640 permits you to turn off its internal ALC and avoids this problem. Hope this helps.

SOP with even the latest SG is to add at least a 3dB pad to both ensure a flat VSWR as well as SG to SG isolation. For NF measurements all loss is automatically cancelled during the cal process.

You could use a service monitor such as the HP 8920 or 8965/E6380A series which will calculate S/N, SINAD, or audio distortion directly.

Sensitivity in HF receivers is relatively unimportant, at least below 40 meters.  A microvolt or two for 10 dB (S+N)/N for AM is pretty good, and more than needed given atmospheric noise level.    A strong (say 10,000 uV) should give 45-50 dB (S+N)/N.  If not, you gots some phase noise issues with the LO.

Thats a relative statement ;D

With a low noise receive only antenna on those rare noiseless nights where you can hear a Siberian fart on 160, getting down as low as possible is a real benefit.  I aim for .25uV or better but Im chasing DX on CW, not farting into a D-104 on AM  ::)

Every manufacturer will have an alignment procedure for their radio. For the simple tube "All American Five" table top radio, there would be very simple alignment.
But something like any of our boatanchor communications radios might get a little complicated.
Just look at the test equipment 'recommended' for alignment of an R390A.
The other replies here would give you an idea what is involved for an alignment, but it won't give you the info to sit down in front of a Collins 75A and go at it.

To W4AMV:  I hate you!  :)  It kept me up half the night trying to prove your seemingly-outlandish claim that the generator's output meter could directly read the receiver-sensitivity voltage when using the 6 db. pad.

After a lot of frustration, and clouds of doubt over my abilities with 8th-grade algebra with the sad conclusion that old age had completely befuddled me, I finally figured out that receiver sensitivity is NOT measured by the antenna-terminal voltage at the receiver, but instead is the voltage applied to those terminals through an impedance equal to that of the receiver input impedance, and in a completely 50-ohm system is actually twice the antenna-terminal voltage.

Oh well, I guess I can sleep later.

And W4AMV, maybe I don't actually hate you, but I did emit a lot of harsh language that didn't specifically exclude you.  :)  But tnx OM, you taught me something!


-al hart, w8vr

Your welcome. The FM generators provided by the MEASUREMENTS CORPORATION were extensively used by the infamous "batwing" company "back in the days". SINAD meters and distortion analyzers and measureing FM sensitivity grew on you as you tried to wrangle every .05 uV of sensitivity out a NBFM radio. What FUN!!

YUP I remember the BatWing stuff and we were lucky to get .5 microvolt sensitivity..........NOW even a Kenwood commercial two-way radio is .2 microvolts.
Even Motorola has got it down to .2 most of the time.
We always set the squelch for a base station to open at 1 microvolt......to reduce the skip on the low bands during Fall/Spring.

I still have memories of my Measurements 80 drifty-assed signal generator. I hated that thing with a burning passion! ! ! ! After having it drift one day after aligning my low-band Pre Prog, I took it off of the shelf and smashed it to smitherines! Then threw it all the way across the basement and them picked it up again and body-slammed it into the concrete floor again!! I took out all of my frustrations on it!!
 
I wanted to make sure that it was well destroyed so I wouldn't be even be the least bit tempted to try to repair it.

2 days later I bought a Cushman service monitor.........................

I love vintage gear, but I despise vintage test equipment!

Wow Fred, you must have seen bogus stuff.  I never saw a Motorola that wouldn't do much better than that unless the receiver had trouble.  Usually they would make .2 for 12 db sinad and add the preamp kit and it would do .1.  It didn't make a lot of sense to get the sensitivity down that low on Low Band because of interference.  And they only suggested one in rare cases on VHF.

Their Motrac, Micor, Mocom 70, and even the small under dash stuff did fine.  Even the tube stuff that preceded it was good for .35 or better.

Yup many older mot stuff was easy to get .25 to .35uv. before it became junk.