[TMC] GPR-90 RXD Tuning

[Richard Knoppow]

----- Original Message ----- 
From: <wb9tow at egr.msu.edu>
To: <jvendely at cfl.rr.com>
Cc: <wb9tow at egr.msu.edu>; "tmc" <tmc at mailman.qth.net>
Sent: Thursday, March 13, 2014 7:53 AM
Subject: Re: [TMC] GPR-90 RXD Tuning


> Very true John, Sensitivity was king in those days!
> But, it was no fun loosing a ZL yesterday AM on 40m CW
> due to front end over load by a station 5 kc away!
> Yes, 10 dB of attenuation might have saved the qso!
> 73
> Steve
>
>> Hi Steve,


     Its interesting that most, if not all, of the modern 
"rice boxes" have resistive attenuators on the input.  Can 
make a big difference.  The GPR-90 and some other receivers 
with two RF stages used only one of them on the broadcast 
band where signal strength and noise are generally high.
     One must distinguish between sensitivity as it is often 
defined and receiver noise level. Most of the old 
sensitivity specs were actually of overall receiver gain, 
that is, they were statements of the signal level needed to 
produce some audio output level. A great many older 
receivers will reach maximum audio output level with a 
couple of microvolts input. Noise was generally specified in 
terms of signal to noise for either a 30% modulated AM 
signal or for CW.  There are specs for both in the GPR-90 
literature.  Now, the above method of signal to noise 
measurement is useful for trouble shooting because it gives 
a reference point for "normal" performance but is very 
dependant on receiver bandwidth so can not be used for 
comparison of receivers of different designs. For that one 
needs to measure noise figure. Noise figure is a measure of 
the ratio of the actual receiver noise output to the output 
from an ideal noise-free receiver. In fact, TMC gives a NF 
number for the RXD, I think it is 6db, which is very good 
for an HF receiver. NF specs are usually given for UHF and 
microwave receivers but rarely for HF receivers because in 
common practice the noise from the antenna is much higher 
than receiver noise at least up to perhaps 20 mhz and often 
well beyond.
    Its interesting that the noise spec for the RXD and the 
90 are about the same. TMC made a lot of the grounded grid 
input stage which used a very low noise triode but the 
conventional pentode stage in the RXD seems to be about as 
quiet. It used what was probably the best tube available at 
the time.
    FWIW, I made signal-to-noise measurements on my AR-88F 
with both the original, modified AVC, and with the standard 
AVC.  The modifications that RCA used on this version were 
evidentlly intended to improve its performance in the 
triple-diversity system it was designed for. Mainly they 
much reduced the AVC "delay" in the original receiver so 
that the RF and IF stages run with lower minimum bias.  This 
does reduce the noise level a bit but also makes the 
receiver overload more easily.  The SNR measured with normal 
AVC was about what's shown in the handbook, about 2 uV for 
10db SNR with about 7 Khz bandwidth and about 1.5 uV for the 
modified bias.  The receiver used the best tubes available 
at the time it was made.  I have no means for measuring NF.


--
Richard Knoppow
Los Angeles
WB6KBL
dickburk at ix.netcom.com