[Collins] 51S-1 and 6BF5

[Dr. Gerald N. Johnson]

On Sun, 2008-02-17 at 08:36 -0800, Ulrich N Fierz wrote:
> Hi all
>   the 6BF5 question keeps popping up and as it is more than 6 (?) years ago I worked on it and posted it to the CCA list I'll do it again here. The higher line voltages these days have certainly also something to do with the overheating of the tube - but that's not the whole story.
> 
Not everybody reads the CCA list.

> Here my edited original comment:
>   
> Hi 51S-1 owners:
> almost everything has been said but for the curious there is some
> additional food for thoughts: not only is the 6BF5 operated too 
> close to or outside the published limits in the 51S-1 (higher AC 
> voltage may be an additional cause) but the bias voltages on the 
> circuit diagram do not compute when you take the resistor ratios...

The voltages shown diagrams can be those measured with a VOM, not
necessarily a 100 megohm input impedance VTVM. When the grid resistor is
100k to 1 meg, the load of the meter is significant. And when th grid
resistor is that large, 100 megohm leakage from the coupling capacitor
can have a significant effect on the 6BF5 operating point. Then the tube
gain is increased with the higher plate current.

>  They will again be different when measured, as an overheated 6BF5 
> might (often) draw grid current.

The hot tube control grid will heat (and with age have been sputtered
with cathode material) and so have grid emission. That leads to grid
current that opposes the bias and so to greater plate current and
greater grid heating from the electron stream. In S-line receivers this
is also a problem with 6BA6 IF tubes that display it by driving the S-
meter negative by displacing AVC with that grid emission current. There
have always been those who advocate checking tubes for grid emission and
claiming faster equipment fixing based on grid emission than any other
tube test. There have been tube testers dedicated to grid emission tests
and few of the major tube testers, if any, have included a direct grid
emission test.
>   
> My solution was to leave the unbypassed 39 ohm cathode (AF and DC
>  feedback) resistor in place and pull the screen grid voltage 
> down by connecting a 27kohm (2W) resistor to gnd (there is a 
> convenient tie point on the turret nearer the p.s. circuitry). 
> This yields better performance and stability than increasing the 
> screen grid resistor only.
>   
Cathode bias resistors are good. They give inverse feedback stabilizing
the operating point of the tube and reducing distortion. Series screen
resistors are not beneficial to low distortion audio because they let
the screen voltage wander with signal swings. A good screen bypass helps
stabilize the screen voltage short term. Lowering the screen supply
impedance with the voltage divider also helps.

> I also replaced the original screen grid resistor with a higher wattage 
> type and the control grid coupling capacitor (which might or might not be leaky).

More than one ham rig screen resistor has been low on power rating. As
are the VOX bias resistors in the S-line transmitters. A coupling
capacitor with 100 Megohm leakage, leaks 1.5 microamps with 150 volts
across it. 1.5 microamps to the typical half meg output grid resistor
gives 3/4 volt change in bias. My tube manual says 7.5 volts is typical
for the 6BF5, so that leakage makes a ten percent change in bias. That
is significant. In receivers that used black beauties for that coupling
capacitor, their leakage can be 50 or 100 times greater, so using a
polyester capacitor (like an Orange Drop) in that location is important.
>    
>   With a good 6BF5 there is no grid current and the bias voltages 
> measured should track what is supposed to be there according to the 
> dividing ratios in the grid bias circuit (starting with
> approx -37VDC).

Early S-line receivers suffer from generating the bias voltages from a
20 ma selenium rectifier that ages poorly and these days is decades
beyond its end of life so the bias supply voltage is low. Then the back
leakage of that selenium has not been kind to the bias supply filter
capacitor. Early 512F-2 transmitter power supplies also have that
problem.
>   
> Now the 6BF5 is operated well within its limits - still getting 
> hot of course... However the tube was designed for TV receivers 
> and there usually had to work long hours, perhaps more then your restored 51S-1 ever will.
>   
> As I spent some time analyzing this matter armed by tube manuals,
> several 6BF5s by different manufacturers (which still is not an 
> expensive tube) and various pieces of commercial test equipment 
> I thought this might information be helpful and save others time better spent. 
> 
> Thank you for the bandwidth and again I hope this proves helpful 
> to the operators of e.g. 51S-1 receivers.
> 
>   73 and keep them working - a 6BF5 works just fine and Collins had a reason to use them!

Yah, lots of gain, which makes bias stability more critical.

>   Ulrich  hb9aik - vk6bqd 
> 

-- 
73, Jerry, K0CQ, Technical Advisor to the CRA
All content copyright Dr. Gerald N. Johnson, electrical engineer