[HBR] Taming 6EH7 IF strip?

[Walt Hutchens]

Ian posted:

> I recently made a start on a 455kHz IF strip / product detector / BFO. The
> lineup is:
> - mechanical filter on the input, matched to 50 ohm output of the RF
> section
> - 2x 6EH7
> - 2x Miller IFTs (the 'interstage' type)

The 6EH7 is a fine tube but with very high gain and gain into the low VHF,
it requires some care.   These are more or less random thoughts.

You don't say if both ends of this strip are terminated:  That would be a 50
ohm resistor on the input and maybe 100k on the secondary of the output
IFT.  

200V seem a bit high for the plates; these tubes are just fine with
125-150V.  However since they're pentodes there won't be much change in the
gain.   Just don't exceed recommended operating power level.

Bypass caps at 455kcs should be 0.1 mfd at least for cathodes.   The usual
0.01 cap has a reactance around 40-50 ohms at that frequency, right?
(mental math -- not accurate but right order of magnitude)  So it's not
really much of a bypass for a low impedance or high RF current circuit.

I assume that B+ and AGC lines are decoupled.

Don't use an unbypassed cathode as any RF voltage there will couple to other
places via the filament leads.   150 ohms is usually fine and bypass the
whole thing.   

Do bypass both cathode leads and use a common ground point per stage.

Use a 100 ohm resistor with a 1/8" lead on the tube end to couple to the
grid of each stage;  this kills the Q at VHF.    While you don't think you
have a VHF oscillation, there can be one that kicks off the MF oscillation.
This resistor is ALWAYS needed with these tubes.    It costs you nothing in
desired gain because the MF circuit has no current flowing in this lead.

As a matter of general practice the screens should be on voltage dividers
with (say) 100k to ground from each.   This prevents the screen voltage from
going up so much with AGC action and thus improves the AGC.

Tube shields usually are NOT required except possibly for the first IF and
mixer to prevent pickup of stray noise at the respective frequencies.

Check the stage voltages against the tube manual.   If you're still having
trouble with oscillation when you make this check, kill it with a 1k
resistor across one or (if necessary) both grid windings.

Don't load the IFTs as this dramatically reduces their ability to back up
the selectivity of the mechanical filter.    As a last resort, increase the
values of the cathode resistors.

That's all I can think of.  Good luck with it!

Walt
KJ4KV