[Milsurplus] [regenrx] Re: pentagrid detector sensitivity

[Michael A. Bittner]

FWIW, the 1R5 (battery equivalent to the 6BE6) is used as a grounded grid UHF amplifier with grid-1 grounded and grids 2, 3 & 4 tied to the plate in the AN/PPN-2 IFF set.  BTW there's a PPN-2 on eBay right now with a BIN price of  $2,700.00.  Also, the Hallicrafters S-72 uses the 1R5 as its local oscillator in a conventional triode configuration with grids 2, 3 & 4 tied to the plate.  Why??  What's so great about the 1R5 that it's used as a GG amp or an LO??  Mike, W6MAB


  ----- Original Message ----- 
  From: davidpnewkirk 
  To: regenrx at yahoogroups.com 
  Sent: Thursday, January 05, 2012 5:12 PM
  Subject: [regenrx] Re: pentagrid detector sensitivity
  --- In regenrx at yahoogroups.com, "kyoritsu" <rikkyograsing at ...> wrote:
  >
  > Forgot to mention a couple of experiments I did with the 6BE6 as a regenerative detector.
  > 
  > I tried injecting the signal on grid 3 (keeping the oscillator tank and grid leak on grid 1). Lyle Williams (The New Radio Receiver Handbook) wrote you could inject the signal into the screen grid of a pentode detector, and I wanted to see how this would work on the other control grid of a pentagrid. Well, it doesn't work. Of course, it can still oscillate but the only station I heard was one local AM broadcaster, regardless of where it was tuned or what band. I was hoping for antenna isolation.
  > 
  > Also tried tying together the two control grids, 1 and 3. Sort of a belt and suspenders approach to regeneration, since the two screen grids are internally tied together in the 6BE6. This does work, but there's a clear reduction in gain.
  > 
  > So far, the 6BE6 works best with grid 3 grounded. But since it is a converter tube, you could add a separate heterodyning oscillator circuit in addition to the usual regenerative detector, something that was suggested back in the 1930s and still comes up from time to time now.
  > 
  > The 6BE6 goes into oscillation at a lower voltage than any of the RF pentodes I tried. It might be a candidate for a low B+ regen. Also, perhaps some of those extra grids could be put to work in space charge mode.

  Some comments on these way-cool experiments. What's unusual about your using the 6BE6 as a replacement for the 6BZ6 is that grid 1 of the 6BE6--which, per the 6BZ6 basing, its its control grid--is its _oscillator grid_. Grid 3 on the 6BZ6 is the tube's signal control grid; the 6BE6 specs in the RCA Receiving Tube Manual say that with the tube's grid 1 grounded and signal fed in at grid 3, the BE6's transconductance (7.25 millisiemens in this connection) is almost that of the 6BZ6 (8 mS) under more or less the same operating conditions (and with its grid 3, its suppressor grid, grounded). I'd be interested in learning how the 6BE6 operates as a regenerative detector with its grid 1 grounded and its grid 3 serving as its only control grid. Perhaps we can find manufacturer specs for the 6BE6 operating as an amplifier, with grid 1 as control and grid 3 grounded; but the tube was designed for grid 3 to be the signal gozinta.

  Yes, the more grids a tube has, the generally more noisy it is--in small-signal operation. (Small-signal operation is when an active device is operated in such a way that the incoming signal is small enough not to shift its dc operating point--that is, when the device is amplifying linearly.) When a tube operates as an amplitude-self-limiting oscillator it's in large-signal mode because its dc operating point--its dc bias--has shifted as a result of the presence of the signal it's generating and handling. A regenerative detector operates in large-signal mode both non-oscillating (it must be operating nonlinearly to "detect" AM) and oscillating (in which state it amplitude-self-limits, commonly by grid-cathode conduction).

  So although it seems facile to consider that a pentagrid regenerative detector might be or should be noisier than a detector with fewer grids, we can't directly that from what we know about the relative noisiness of tubes with differing numbers of grids operating *small*-signal.

  Another thing to keep in mind when evaluating different tube types that can be plugged into a given basing hookup is that in most screen-grid detectors, only the embedded triode consisting of the tube's cathode, grid 1, and grid 2 does the RF amplification/oscillation. The specification conveying the most significance to builders of such detectors, the grid-1-to-grid-2 amplification factor (mu, pronounced mew), is only rarely included in tube specifications. (If "triode connnected" data is available for a given pentode or tetrode--"triode connected" in this case usually meaning "by connecting the plate and grid 2 [screen] together"--the mu reported can, per Lankford-Smith in the Radiotron Designer's Manual, be taken as closely equivalent to the grid-1-to-screen mu.)

  The variation from type to type of grid-1-to-screen mu is the main reason you've encountered significant regen-control-setting variation for critical regeneration across the various types you've tried. The RCA 6AK6 data doesn't include its triode-connected mu; as an example, though the 6F6 pentode has a mu of 7 when triode connected. In contrast to this, the 6AU6, a sharp-cutoff pentode with transconductance about half that of a 6BZ6, has a mu of 40 when triode-connected. So, yes, I'd expect that you'd have to turn up your regeneration "considerable" with a 6AK6 relative to a 6BZ6 or a 6BZ6. 

  That you had to turn up regeneration with the 6AK6 does not necessarily mean that this tube is "worse" as a detector than the BZ6 or BE6; it merely means it's different--for as I've reported earlier in this forum, I find that tubes with lower grid-1-to-screen mus are generally more frequency-pulling-resistant in the presence of strong signals than tubes with higher mus when they're all compared as oscillating detectors.

  Best regards,

  Dave
  amateur radio W9VES