Fw: [Collins] Stage gain

[David Knepper]

Seminal piece

David Knepper - W3ST
Secretary to the Collins Radio Association (CRA)
Publisher of the Collins Journal
www.collinsra.com
CRA station call - W3CRA
----- Original Message ----- 
From: "Dr. Gerald N. Johnson, electrical engineer" <geraldj@isunet.net>
To: "David Knepper" <cra@floodcity.net>
Cc: <w8krr@att.net>
Sent: Saturday, March 01, 2003 10:50 PM
Subject: Re: [Collins] Stage gain


> With a radio down 30 dB, I'd be suspicious of screen bypass capacitors.
> I'd look for low screen voltage on every stage. If there are still old
> oil filled paper capacitors, I'd scrap them all and put in orange drops.
> Don't bother testing, just replace them all. You accomplish the task
> faster and won't have to do it again. It would be good to check
> resistors, especially screen and cathode resistors for values out of
> tolerance.
> 
> Remember that stage gain is directly proportional to screen voltage so a
> little loss in lots of stages can add up. Then I'd check tubes in a tube
> checker. If its one tube that should show up really bad.
> 
> There are two ways to get stage gains. One is with a sensitive RF
> millivoltmeter and a standard signal level at the antenna. Trouble is
> that the meter probe detunes each stage. A sensitive wide band scope,
> like my Tek 475 does well, but detunes each stage. Neither the RF
> millivoltmeter nor the wide band sensitive scope are common tools in
> service shops. Those are more like laboratory tools. A more repeatable
> technique is to drive grids with a low impedance signal generator,
> driving for a certain level detected audio or S-meter reading. Then the
> low impedance of the signal generator swamps the grid circuits and all
> the detuning is ignored.
> 
> How ever the stage gain is measured, every stage (except the diode
> detector and maybe a triode mixer) should have gain. The IF stages
> should have the most gain per stage, then the RF stage(s), then the
> mixers. Gain is harder to measure around mixers because of the local
> oscillator signal being easily coupled to the RF millivoltmeter.
> 
> If the loss of gain is based on S-meter indication, then I get as
> suspicious of AGC time constant and bypass capacitors as I am of screen
> capacitors. It takes very little leakage (a few megohms) to reduce AGC.
> In nearly all receivers using vacuum tubes, the S-meter is in the screen
> circuit of an AGC controlled IF stage. Essentially the S-meter circuit
> is using the IF tube as a simple VTVM. Low screen voltage on that stage
> can make the meter read low.
> 
> I don't see much advantage showing waveforms. There should be only two
> waveforms in an AM receiver with a modulated signal generator. Before
> the detector, its the AM signal at various levels. After the detector
> its the modulating audio. Any deviation from these two waveforms means
> distortion at the speaker. Its not like all the various waveforms in a
> TV. Not at all.
> 
> One other thing to check is the RF gain control. If its not getting down
> as low in value as it should, it will keep receiver gain down. Presuming
> its in the RF/IF stages' cathode. Otherwise it can be accomplished (as
> it is in S-line) by injection DC bias into the AGC line. If the RF gain
> control is in those cathodes, often the mute switch is in that circuit
> too and it needs to be checked.
> 
> It might be possible to compute the stage gains, though one would have
> to presume plate load impedances. Stage voltage gain is predictable as
> Gm * Rl. That is transconductance times plate load resistance. Mixer
> tubes are often rated for conversion transconductance (generally quite
> low).
> 
> 73, Jerry, K0CQ
> -- 
> Entire content copyright Dr. Gerald N. Johnson, electrical engineer.
> Reproduction by permission only.