Fw: Fw: SB-301 Frontend tubes...sensitivity

[talen]

Ed,

This reflector certainly is informative and has a wealth of information
and I welcome more input/comments. This is interesting stuff.

The proceedure you outline is exactly the one which is documented
to determine the S+N/N sensitivity of a receiver with a fixed 10db
delta at the speaker and see what level of signal it took to achieve
that.....the lower the better.

I'm not trying to establish an absolute receiver sensitivity level, only
make "relative" measuremants. 0.2uV is as low as my generator
goes without external attenuators. So, what I'm trying to do is a
relative comparison to see which is better with a fixed 0.2uV input
at the antenna, make a change to the design and see if it improved.
I guess a better approach, even for a "relative" measurement would
be to insert a step attenuator and record the db attenuation while
maintaining a "constant" 10db S+N/N at the speaker.

I don't believe a stronger signal with increased noise means better
sensitivity. That's like adding another RF stage or an AF stage for
that matter and making the signal (and noise) louder. Seems like
what you would want for "sensitivity" in a frontend is response to
very low level signals and greater separation between the noise
and signal ......or a higher S+N/N ratio. Other things to consider
in a frontend would be image rejection, strong signal handling
capability, etc.

73s  Kees K5BCQ


----------
From: k6uuz at juno.com
To: talen at inetport.com
Cc: HEATH at LISTSERV.TEMPE.GOV; kk7ih at nvrams.org
Subject: Re: Fw: SB-301 Frontend tubes...sensitivity
Date: Thursday, August 10, 2000 12:24 AM

Kees;

I thought all night about your post and came to the following conclusion:

You stated that the S/N ratio stayed about the same but both signal and
noise increased.

The increase in signal represents an increase in sensitivity, even if the
noise increased too. What we are measuring here is the minimum signal to
achieve a given S/N ratio, not the S/N ratio it's self. Look for an
increase in gain, not an improvement in S/N ratio which may be fixed.

I don't know your set-up or experience, so I will assume the worst and
outline the procedure for measuring the gain and sensitivity of a
receiver.

The standard procedure for measuring sensitivity of an AM receiver is the
amount of signal from an AM signal generator modulated 30% by a 400 ~
tone, to achieve a 10 dB S/N ratio.

For SSB you can zero-beat the carrier or turn off the modulation and use
the loudest tone from an AM carrier.

To perform this you need a HF signal generator with a calibrated
attenuator (I use an army surplus SG85/URM25 which does a good job) and
an AC voltmeter across the speaker.

Put a 51 ohm resistor across the antenna terminals and set the volume for
some low value of noise as read on the meter. Never touch the volume
control again. record the meter reading as "A".

Remove the resistor and connect the signal generator. Tune the generator
(or radio) for the highest meter reading and adjust the attenuator for a
10 dB increase in the meter reading. Record the attenuator setting as
"B". This is the sensitivity of the radio as it sits.

DB can be calculated from the formula dB=20 X log of the ratio of the two
readings (in microvolts). If my math is correct this a change of 3.16. In
other words an increase in the meter reading from .1 volts to .316 volts
is 10 dB.

Now change the tube and realign the affected circuits. Take another 10 dB
measurement, same as above, using "A" as the base. Record the attenuator
setting as "C". This is the sensitivity of the radio with this tube in
it.

Do the same thing with other tubes you want to try. Record the attenuator
setting to achieve a 3.16 increase over "A", as "D", "E", etc. Each of
these is the sensitivity of the radio with that tube in it.

Conclusion:
Convert your attenuator settings to dB if required. Compare the settings.
If "C" is 2.7 dB greater than "B", that is the added gain that tube gives
you. If "E" is 1.2 dB less than "D" that is the loss between those two
tubes.

Ed Richards


On Wed, 9 Aug 2000 03:44:27 -0500 talen <talen at INETPORT.COM> writes:
> That's right, Ed. But just like on the SB-104(A) testing, I tweaked
> and
> peaked everything for the frequency under test. In this case the Ant
> and RF slugs and the Preselector, of course. The 6GM6 and 6EW6
> (especially the 6EW6) tuned up differently and showed more overall
> gain (making me think initially it was a hotter frontend) so I
> reduced
> the AF gain to maintain a 30mV Noise level. Turns out the gain was
> both signal and noise and the (Signal+Noise)/(Noise) ratio didn't
> change substantially from that of the 6BZ6 as indicated.
>
> The thing that surprised me was how closely together they measure
> on my TV-7 when set up for a 6BZ6. I don't get that........
>
> I don't know what the effect of a "new" 6EW6 or 6GM6 (with really
> high gm) would be because I don't have any.
>
> Also, all these measurements have been at 14.250Mhz only. I intend
> to take more data on 15m and 10m because 20/15/10 tend to be
> where sensitivity causes the most problem.
>
> 73s  Kees K5BCQ

Listserver Subscription:listserv at listserv.tempe.gov - "subscribe heath 'name' 'call'"
Listserver Submissions: heath at listserv.tempe.gov
Listserver Unsubscribe: listserv at listserv.tempe.gov - -"signoff heath"