[R-390] Measuring Sensitivity

[n4buq at knology.net]

On Fri, 09 Mar 2007 11:38:06 -0500, Roy Morgan <roy.morgan at nist.gov> wrote :

> At 09:27 AM 3/9/2007, n4buq at knology.net wrote:
> >... I'm interested in measuring the sensitivity of my R390A and I have
> >a signal generator (GR1001A) that has a 10-ohm output.  I have an adapter
> >that houses a 40-ohm resistor that matches the output to a 50-ohm cable.
> 
> Barry,
> 
> The GR 1001A is one of my favorite generators, so I have some information 
> and ideas for you:
> 
> The manual for this thing is on BAMA.  Reading it and puzzling out what 
is 
> going on is a good idea. (Link is below.)
> 
> The 1001A has an odd set of output impedances.  "The output impedance at 
> the ATTENuator jack is 10 ohms for all except the 100 MV  MULTIPLIER 
> setting, where the output impedance is 50 ohms."
> 
> >When I feed the balanced input with this (40 ohm weries unit), there is 
an 
> >obvious mismatch between the 50-ohm cable and the input.
> 
> Yes, but it may not be severe.  The input impedance of the R-390A may 
vary 
> from about 50 ohms to about 700 ohms depending on frequency, tuning of 
the 
> input circuit, and component variations.  I quote from MIL-R-13947b, the 
> Military Specification for the R-390/URR ("non-A").
> 
> "3.13.3 Antenna input impedance.- The rated input impedance for the 
balanced
> input circuit shall be 125 ohms. In the range from 500 kc to 16 mc, the 
> measured input impedance shall not be less than 50 ohms nor greater than 
> 375 ohms; for the range from 16 mc to 32 mc the measured input impedance 
> shall be not less than 100 ohms nor greater than 700 ohms."
> 
> If you accept the fact that the input impedance actually does vary over 
> such a range, then you can do a number of things:
> 1) Make sure you can reasonably expect the actual RF voltage at the 
antenna 
> terminals to be what you think it is.  (This is relatively simple to do.)
> 2) Simply follow the manual procedure for sensitivity measurement, 
keeping 
> in mind that the maintenance manual procedure is meant to deliver radios 
to 
> service that are meeting some sort of minimum performance measure, not 
make 
> an accurate measurement of actual sensitivity.
> 3) Measure the input impedance of the radio and do a bunch of figuring.
> 4) Just wing it with informed engineering as Collins seems to have with 
the 
> S-line.
> 
> >...how do I determine how much signal is present at the radio for a 
given 
> >setting on the generator?
> 
> The 1001A generator will deliver to the output end of a (relatively 
short) 
> cable the voltage level indicated by the panel meter and cotrols if:
> - on the 100 microvolt setting there is only the cable
> - on other settings, the 40  ohm series resistor unit is in place.
> 
> NOTE: If you terminate the cable with 50 ohms, in either case above, the 
> actual voltage at the termination will be ONE HALF what the panel 
controls 
> indicate.
> 
> Realizing this, if you simply use the 40 ohm series unit, and feed the 
> R-390 balanced input (one side grounded) with the end of the un 
terminated 
> cable, AND you make the assumption that the input impedance if the radio 
is 
> 125 ohms or more, you will have somewhere between the voltage indicated 
on 
> the generator and half that.  This is equivalent to method 4 above.
> 
> Note: it is quite clear from the reports of a Collins List member, who 
has 
> spoken to at least one of the design engineers who was at Collins during 
> the S-line production period, that they simply ran a simple cable from 
the 
> generator output to the radio, and took the panel indications as the 
value 
> of the input voltage.  The generators they used were meant to deliver 
> indicated voltage when the output was terminated in 50 ohms.  This works 
> out pretty well, IF the input impedance/resistance of the radio is 
> approximately 50 ohms.  I don't yet have any idea of what the actual 
input 
> impedance of S-line radios is.
> 
> 
> >  I realize I need to make up a matching network at the
> >termination of the 50-ohm cable where it connects to the receiver, but 
not
> >sure how all these matching networks play into the signal levels between
> >the generator and the radio.
> 
> 
> Here is what I suggest, if you want to "Make sure you can reasonably 
expect 
> the actual RF voltage at the antenna terminals to be what you think it 
is."
> 
> Make up a voltage divider with a 50 ohm resistor and a one-half ohm 
> resistor in series from the input to ground.  The output is taken off the 
> one-half ohm resistor.  This will do a number of things:
> - It will present a 50.5 ohm load to the end of the cable from the 
generator.
> - It will divide the voltage at it's input by 100.
> - It will let the cable operate at 50 ohms, so there will be no or little 
> standing waves to bother anything.
> - It will reduce the effect of any possible leakage into the cable or 
> connections up to the antenna connector, either from the generator or 
from 
> broadcast band signals or whatever.
> 
> I suggest you make this thing up inside a twinax connector with a BNC 
> connector fitted into the back end.  This will provide a convenient 
> divider/load for use with all your R-390 radios.  The half ohm resistor 
> (two one-ohm units in parallel perhaps) go from twinax pin to pin.  
Ground 
> one of them (there is a correct one to be grounded). Feed the other from 
> the center pin of the BNC connector via the 50 ohm resistor.  The BNC 
> connector can be mounted into the cable clamping nut of the twinax 
> connector either by filing the threads and force fitting or soldering, or 
> by threading the clamping nut.  (I think it's a 3/8 by 32 thread.)  Allow 
a 
> flexible wire from BNC center connector to the 50 ohm resistor to allow 
for 
> threading the clamping nut into the connector body.
> 
> Now, if you have a GR 1001A, use the 100 micro volt setting of the 
> attenuator with no series resistor.  If you use other settings of the 
> attenuator, use the 40  ohm series unit. You have a generator system that 
> delivers to the antenna terminals one percent of the voltage indicated on 
> the panel.  If you have a generator that operates correctly with the 
cable 
> terminated in 50 ohms,  just use the cable from generator to voltage 
> divider unit.
> 
> The voltage at the antenna terminals will be one hundredth of the panel 
> indication.  The input impedance of the radio will not matter enough to 
> bother.  (The "output" impedance of the signal source is about one half 
> ohm, and the input impedance of the radio is from 50 to 700 ohms.)
> 
> >I assume that once this calculation is made,
> 
> I did just above.  It's convenient to divice the generator output 
> indication by 100.
> 
> >then when I feed the radio with a modulated AM signal with the BFO off 
and 
> >get -7V on the diode load, then I should be able to know the sensitivity 
> >level, correct?
> 
> Yes, correct. You will likely get DIFFERENT results if you use the method 
> in the manuals, with a URM-25, for instance.  It's likely the numbers 
from 
> the manual method will be from two thirds down to one half of the numbers 
> from using a low-output-impedance attenuator as I describe above.
> 
> The reason:  An unterminated generator meant to indicate correctly if the 
> cable IS terminated will deliver TWICE the indicated voltage at the end 
of 
> the cable.  The receiver most likely has in input impedance above 50 
ohms, 
> possibly three to 14 times as high, and so will appear more or less like 
an 
> open circuit to the cable.
> 
> A note on leakage:  If you have an EICO, HEATH or other similar 
generator, 
> the question is not whether or not it leaks, the question is how 
> much.  Almost certainly it is leaking and the leaked signal probably ADDS 
> to the signal from the end of the cable at the receiver input 
> terminals.  So the sensitivity readings you get will be misleading.  To 
see 
> what it takes to get a moderately low leakage generator, examine the HP 
606 
> generators: every lead into and out of the oscillator box is filtered 
> thoroughly, the box is cast aluminum with rf seals on every opening, the 
> attenuator is similarly constructed, and so on.  There have been reports 
> that the common HP 8640 generator leaks somewhat unless it's seals and 
> fastenings are cleaned and tight. To see what's going on with a 
generator, 
> put a nice 50 ohm load directly on the output connector and sniff around 
> with your properly running receiver set to full gain.
> 
> >The reason I ask is that I'm having to crank the generator up to around
> >10uV to do the "switch the modulation on and off to get 10dB S+N/N 
readings
> >and that seems high.
> 
> You may well have leakage from the generator.  The 1001A leakage spec is 
> "Fields at one mc are less that one microvolt per meter at two feet from 
> the generator."  Do test it and then take it mostly apart to clean, 
tighten 
> and reseal all and every place there might be leakage.  One report from 
our 
> now-absent Nolan Lee told how he had to replace the coax inside the URM-
25, 
> because it was leaky or lossy, and also the BNC-to-BNC cables that came 
> with the thing were so lossy as to completely foul up any measurements.
> 
> >  The radio "hears" quite well, but I'm wondering just
> >how sensitive it really is.
> 
> It sounds like really you don't know yet.
> 
> >The problem I'm having with the "modulation on, modulation off" method, 
is 
> >that the cable and generator make a very, very good antenna, especially 
on 
> >the BC band, making these measurements  quite difficult.
> 
> Terminate the INPUT end of the cable with 50 ohms. Then try to figure out 
> where the signals are getting into the radio.  The generator set to a 
high 
> level with a few-turn loop on the end of another output cable can help 
you 
> find the route of signals into the radio.  Bottom and top plates may 
> help.  De-Oxit on RF and IF connectors may help.  (Compare a Mini-BNC 
used 
> in the R-390A with a General Radio 874 locking connector to get an idea 
> what the differences are!)
> 
> When I replace line cords on receivers, I locate a shielded computer cord 
> (these are marked "shielded" and are a bit thicker than normal) and use 
> that.  Any R-390 which is running without its line filter may be leaking 
> via the cord.
> 
> Speaking of loops, GR made the 1001-P3 Voltage Divider,meant to inject 
the 
> correct CURRENT in series with a household radio loop antenna.  This 
thing 
> is, I think, a 50:1 divider, and if you find one it could easily be 
> modified to do the 100:1 division.  Just add a one-ohm resistor across 
the 
> little wirewound card resistor you find in there.
> 
> GR made a standard loop antenna (type 1000-P10) for use with the 
> GR-1001A.  It had the needed matching inside it, and was meant to create 
a 
> standard field for testing of loop-equipped radios.
> 
> Also useful at times is the 1001-P4 Dummy Antenna.  This thing was 
defined, 
> in the 30's I think, to simulate a wire antenna over the BCB-HF frequency 
> range.  The URM-25 manual and many other sources contain the details of 
the 
> thing (a little inductor, a resistor and two caps.)
> 
> >Any advice here would be most appreciated.
> 
> See the R-390A FAQ page for LOTS of manuals and such:
> http://209.35.120.129/faq-refs.htm
> Including the R-390/URR engineering report, the R-390 Cost Reduction 
> report, and also the Mil Spec, which is at:
> 
> Military Specification, MIL-R-13947B - Receiver, Radio (Radio Receiver 
> R-390( )/URR) 1960-Oct-26
> http://209.35.120.129/mil-r-13947b.pdf
> 
> 
> There is a document: sensitivity_alignment.pdf that is also on the Pearls 
> of Wisdom page
> http://209.35.120.129/Pearls/index.htm
> that is most helpful. It is also most long, but very good reading.
> http://209.35.120.129/Pearls/sensitivity_alignment(s).pdf
> 
> Hewlett Packard, now Agilent, published many tutorials, tech notes and 
such 
> on their generators.
> 
> The GR 1001A manual on BAMA is at:
> ftp://bama.sbc.edu/downloads/gr/1110a/
> (No home should be without a 1001A, and the manual.)
> 
> (Comments, corrections, and discussion welcome.)
> 
> Roy
> Who needs an output attenuator dual pot for his 1001A

Roy,

Lots of great info here (understand the updates to the 100 millivolt vs 100 
microvolt settings).  I bought some components and had intended to build 
one of the 100:1 dividers and never got around to it.  Looks like I need to 
do that as it will make the measurements a bit easire (I think).

What I'm still a bit confused about is the fact that the receiver is 
connected across the 1/2-ohm resistor.  Isn't the radio seeing a pretty bad 
impedence mismatch here?  I realize this makes the generator immune to 
impedence changes brought on by the receiver's input, but doesn't this 
mismatch have a detrimental affect on the radio's ability to "see" the 
voltage across the 1/2-ohm resistor?

Sorry if this sounds goofy, but I'm having a bit of a hard time 
understanding all the impedence values and their effects in this setup.

Thanks again,

Barry - N4BUQ