[sdiy] tube preamp and fx chain

[René Schmitz]

Hi Antti,

Antti Huovilainen wrote:
> At DC yes (IIRC, the grid current was pretty much eliminated at approx 
> -0.5V bias). With "high" input signal, the grid current still has a 
> dynamic effect.

Yes, but this depends largely on how much current the pickup can supply 
without going into the knees. I guess that it can well source a couple 
of microamps without a significant drop in output voltage. The question 
here is just if the source (pickup) is stiff against the load.
The tube itself is fairly tolerant to this, as long as you supply the 
grid current, it'll continue to operate linearly. Infact there are a lot 
of circuits where they purposely push the grid to positive voltages, but 
this is mostly RF stuff. But you'll see this also with high-enders 
(yuck) misusing transmitting triodes. Often driven from a small poweramp 
via an interstage transformer.
(It'll work more like a bipolar transistor in this case. I saw nice 
graphs for one of the russian subminiature tubes I have used, which 
allows pushing the grid to 100V+, for short duty-cycle switching 
applications...)

> Do you perhaps mean the same designers who used highly scientific 
> methods for tuning guitar speaker cabinets, such as Mr. Marshall? ;) 
> (for those not in the know, Marshall 4x12" cabinet is pretty much the 
> smallest that can physically fit the four speakers).

Yes, those are the same guys who use grid leakage to get a special 
sound. :-P

> Assuming ~250V B+ and 100k plate resistor for preamp (seems pretty 
> common) and biasing the stage near maximum output swing (150V plate DC 
> voltage or so), bias voltage is roughly -1.1V. That's just 0.6 volts 
> from the point where grid current starts to flow in the microampere 
> range. You can't really bias 12AX7 at -2V with resistive plate load, get 
> max output swing and keep it within the allowed voltage limits.

Grid current is an exponential function, I(Ugk) = Io exp(Ugk/0.1V), so 
it would drop to almost insigificance for the steady state.
You also don't need the maximum swing, only so much as your gain 
permits. (And if they really were after distortion, they wouldn't have 
bothered with pure class A operation, now would they? :)

Cheers,
  René

-- 
uzs159 at uni-bonn.de
http://www.uni-bonn.de/~uzs159