Input resistors

[lessthanmatt88]

I see on some circuits that there is a pull down resistor in front of a series resistor to the gate of a JFET or base of a transistor, sometimes its the opposite way around.  What effect does this have on the circuit and how does it affect the input impedance.

One more thing... where does the idea that you can just add the resistors (weather they are parallel or not) to get the input impedance come from?  It makes sense some times but it seems it doesnt mathematically make sense.

[bioroids]

Hi

The pull-down resistor is not really part of the input stage, as it goes before the input capacitor. The resistor at the gate (or base) of the transistor is part of the biasing scheme.

Both are resistances to ground so you have to take them into account for calculating the input impedance of the whole stage (besides possibly other resistors and the transistors too). If the resistaces are in parallel the total resistance is not the sum, but rather goes with the 1/(1/R1 + 1/R2) formula.

Luck!

Miguel

[lessthanmatt88]

Well to be more specific, the differences between the input stages of the Eighteen, the Fetzer Valve, and the English Channel from Runoffgroove.com.  The Eighteen has an input with a 1 Meg pulldown resistor, the F-Valve has a 68K then a 1 M in parallel, and the English Channel has a 1 Meg pulldown before the 33k, the opposite orientation from the Fetzer Valve.  I was wondering what importance the order of those resistors have and how they affect the input impedance and also what effect there is when you dont have a resistor in series with the front end as per the Eighteen.

[R.G.]

Here's some good reading:
"What are all those parts for?" - http://geofex.com/circuits/what_are_all_those_parts_for.htm
"Why does my pedal pop?" - http://geofex.com/Article_Folders/box_pop.htm
"Impedance" - http://www.geofex.com/Article_Folders/impednc.htm

In the case of the three things you mention, the reasons are
1. bias; the JFETs have to have their gates biased to ground. This is done by a 1M, possibly in series with other resistors. The exact value does not matter much to the FET. But it matters to the inputs.
2. The input impedance was chosen to emulate the tube circuits these were taken from. The eighteen has a 1M to ground because that's what the Marshall 18W has.
3. The Fvalve has two 68K's in parallel, and the circuit reduces to the 1M to ground with a 34K in series from there to the gate, same as the EC.
4. The EC has a 1M to ground, then a 33K in series to the gate of the FET. This is the same as the Fvalve reduces to.

The series resistance has no effect on the biasing. If it were a 68K or 34K parallel resistor to ground, that would load down the guitar a lot and make the treble muffled - the so-called tone sucking effect. But it's not that way, it's in series.

A resistor in series with a gate or a tube grid is confusing - the gate or grid is many, many megohms, so what does a puny resistor do? It can't make the gate resist more.

What it does is cause a treble rolloff with the parallel capacitance at the gate or grid. There is always some capacitance to ground at the gate, and the feedback miller capacitance from the drain/plate looks like a capacitor to ground. That capacitance and the series resistance cause a loss of signal at the frequency where F = 1/(2*pi*R *C). The Fetzer valve write up shows this effect - look at figure 4. Series resistors with gates and grids are often referred to as stopper resistors or grid stoppers, because they stop RF oscillations in many cases by lowering RF gain while leaving the audio.

[lessthanmatt88]

Thanks, I understand that a little more now.