Post PI splitting of signal

[lowell]

Hey,
I want to split the post-PI signals again so that a pp pair of 6L6 and a pp pair of 6V6 can have different bias settings. Would 2 1k resistors after the PI coupling cap be sufficient to split them? I'm then thinking that I'll need 1 more coupling cap on one set of tubes to keep the grid bias arrangement from interacting. I'm aware that 2 coupling caps instead of 2 1k would work, however I want the AC signal AND DC to not interact if I were to turn off signal to the pp pair of 6L6. I hope yall can understand what I'm talking about. Sorry for the horrible explanation.

[octal]

The problem with this scheme is that the 6v6s will go into grid clamping at a lower drive voltage than the 6l6s and it will take quite large isolation resistors to prevent this from happening. You'll need either dual PIs or a cathode follower to fix it. IMHO.

[cbarrow7625]

Doesn't Mesa "fix" that problem by using a voltage divider in front of the lower biased tubes (i.e., the 6V6 in this case)? That way one PI can drive both sets of tubes into OD at the same, or similar, times.

I don't think that two 1k resistors is going to give you enough isolation. The two bias controls will interact. You might luck out and get them both set correctly but I doubt it.

as you alluded, the best way is to use two coupling caps per PI output (4 total) to completely isolate the DC bias to each pair of tubes. If you want to "shut off" the 6L6 then use a 4PDT relay (or switch, or pair of DPDT relays) to switch in large series resistors immediately following the coupling caps and then shunt the ends closest to the power tubes to ground using large capacitors (like 10-25uF or more). The large resistors will isolate the AC from the 6V6 side & the large caps at the 6L6 grids will ground out the AC going to the tubes (but not the DC bias). This has the advantage of not messing up the output impedance too much (not nearly as much as lifting the ground on the 6L6 cathodes to turn them off).

Think about it, it might make sense.

[lowell]

Ok yes that is a good option too. I didn't think of that way of doing it, and yes I'd like to avoid the impedance changing when turning of the pair of 6L6. I've done the "lifting the cathode" way before but this way makes more sense. We'll see how it sounds!

[lowell]

What if I put a 10uf cap b/t both 6L6 grids after the large resistors? I'd think this'd work and save on an extra cap because the out-of-phase signals'd cancel eachother. If so, is an electrolytic gonna work? Does cap polarization matter here?

[cbarrow7625]

Yes, polarization matters. You would need to get a non-polarized electrolytic. They are available from any electronics house, they shouldn't be hard to find.

However, if you want the signal to but shut off completely that will not work well. THe drive signals form each half of the PI are not 100% identical so they will not cancel completely. They will reduce in level a great deal though. It might be close enough to get the effect you desire.

[lowell]

right on... thanks!