> the 6386 Rp value is listed at 4250
Sure, at the test condition. 6386 is fairly low-Mu. For a given cathode size and grid technology, maximum Gm is fixed, Rp is Mu/Gm. Or Gm = Mu/Rp. 17/4250 is 4,000 uMhos, which is what the book sez.
FWIW, the Fairchild idles at much higher voltage and I think somewhat lower current than the 6386 book spec. Rp is probably like 8K rising to 50K, Mu 10 going down to 4, Gm 2,800 falling to ~300.
> which is lower than a lot of those remotes.
If comparing to the Pentodes: hell yeah! Pentode plate resistance is meaningless in audio. Sure Larrchild heard a difference between 5749 and 6BA6, but it wasn't the Rp. That's swamped by other circuit impedances. Probably grid-pitch or bracing or other "minor" change caused the audible difference. Pentode Rp is meaningless, but it has a lot to do with the screen grid pitch and spacing, which has other effects on the sound.
> I do knot know what this lower Rp would do to the circuit...
Not sure which circuit we have.
In Pentode operation, we can neglect Rp per se. In Triode operation, it is a Big Problem. Since Mu really does not change much (only about 2:1 in the 6386), we really have variable-Gm limiters. But Rp changes too! And the gain of a triode is Gm*(Rp||Rl). If Rl could be made infinite, and if Mu were constant (for most tubes it is nearly constant, even 6386 can't be varied more than about 2:1 or 3:1), then as Gm falls Rp rises and Gm*Rp is a constant equal to Mu! Even on 6386 we could only get 6dB or 10dB GR.
That's where a Pentode is an advantage. Using tubes, resistors, and iron coils, Rl is never as high as Rp. Now gain is Gm*Rl. Gm varies, Rl is fixed, gain changes.
In most triode limiters, the idle Rp is less than Rl, gain is Mu, and does not change a lot when you start to starve the tube. Finally you starve it enough for Rp to get higher than Rl, and now gain falls as fast as Gm.
I went the other way on my 12AU7 limiter. Using high Rl is good for gain, reduced tube-count, but we have chips and gain is cheap. I loaded the Rp=>6K 12AU7 with 1K. Since the gain equation is nearly Gm*Rl, as soon as current and Gm fall, gain falls. Gain is low (nearly unity) but we can fix that for a buck: 5532 has gobs of gain with noise as low or lower than 12AU7 so we can afford unity gain or even a loss through the tube. And 1K-2K impedances work with cheap transformers, unlike the ultra-high impedances used in most tube-era limiters that used transformer coupling out of the vary-gain stage.
