Coupling capacitor values/output impedance with different plate loads

Dr. Toobz · 3325

0 Members and 1 Guest are viewing this topic.

Offline Dr. Toobz

  • Sr. Member
  • ****
    • Posts: 432
Looks like my summer 2013 project will be a Quickie 1.1, my second Quickie, in fact (the first, however, exists only as a scarred plate and a bird's nest of wires and parts, thanks to constant experimenting).

Here's a discussion that we might want to pin as a sticky: how changing the plate load on the 3S4 tubes from resistor to choke to CCS may or may not change the effective output impedance of the preamp, and if so, what that may mean in terms of driving longer runs of cable. Additionally, I'm curious as to whether the coupling capacitors would change in tandem.

The stock configuration appears to yield an effective Z-out of 2000 ohms, as a trioded 3S4 is supposed to be around 4k for plate impedance in this configuration, and the stock plate resistor is also around 4k. The stock coupling capacitor value is 2.2uF, if I remember correctly. Here's where I find things to get murky: the PJCCS supposedly doubles the output impedance (to 4k), from discussions on previous threads. Yet, the coupling capacitor presumably stays the same (e.g., 2.2uF). So, we've effectively doubled the output impedance - does this mean that the capacitor value should also double to keep the same corner frequency in the bass region and/or drive the same length of cable without any issues? The 150H Hammond chokes also present a different puzzle - DCR is 3.7k, so pretty close to the stock value. On the surface, I'd assume this means the same output impedance, and thus, no changes to the capacitor value. But, a choke is not just a resistor! Given the large inductance value, would the coupling capacitor need to change value? For example, on the older SEX amp, going to the MQ iron and choke, which have better inductance than the stock Speco and 30H plate choke combo, bumps the output capacitor value from 1uF to 2.

For what it's worth, I'm probably going to end up with the chokes on this Quickie, to purposely "fatten up" the sound. My Heresies sound thin and pathetic on wood floors with the Stereomour....sometimes accuracy is not more musical!



Offline Paul Joppa

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 5834
With a choke load, the midrange impedance is very high, so the output impedance is around 4K. It is naturally more complicated in the deep bass; 150 henries at 5Hz is about 4K and thus not high enough to ignore unless the corner frequency is well above 5Hz. When a choke load is used, it becomes difficult to model the response because of the eddy-current losses in the choke core, which appear as a frequency-dependent parallel resistance which is specific to the particular choke being used. Experimentation is the most practical answer at that point.

There are many who argue that a corner frequency (relative to the minimum load impedance) well below 20Hz is better. Things get pretty murky at this point, with 2Hz and 5Hz being frequently quoted as desirable targets. The connection of these arguments to actual audible experience is - shall we say? - tenuous at best. We can perceive sound well below 20Hz, but few of us can reproduce it in the home that low, or find source material with such content ...

A separate argument can be made for a minimum capacitance relative to the output impedance, which aims to minimize pickup of 60-Hz electrical fields. If the minimum load impedance is 10 times the output impedance (a commonly-used standard) then that would give a 6-Hz corner into the load. This is what I use when there are no other significant considerations. It applies to the coupling between driver and power tube, as well as any preamp output.

There is no good argument for a maximum capacitance, or an optimal capacitance, other than cost and bulk.

Paul Joppa


4krow

  • Guest
   Whew! Since I am using the choke in the Q, I was worried that some electrons have been bent to a new direction, resulting in a chalk board of calculations and theory. As always, I am fascinated by these discussions, even when I don't have adequate knowledge to respond intelligently. You rarely can change one thing w/o the possibility of a domino effect.



Offline earwaxxer

  • Hero Member
  • *****
    • Posts: 1336
I love this kind of discussion! Way the "f" over my head but love it all the same. Over time, I pick things up. So, its all good. I know so much more now than I did before I built the Quickie. Thats a good thing.

Eric
Emotiva XPA-2, Magnepan MMG (mod), Quickie (mod), JRiver, Wyrd4sound uLink, Schiit Gungnir, JPS Digital power cord, MIT power cord, JPS Labs ultraconductor wire throughout, HSU sub. powered by Crown.