Common Emitter Gain Control

[Dylfish]

Hi,

Is there a way of controlling / limiting gain when bypassing the emitter capacitor using a common emitter configuration. Putting a pot in series with the cap would form a Rc filter which i wouldn't want.



This circuit is adjusted so the Vcq is around 4.5 v (1/2 supply voltage) with 1ma through Ic & Ie. Without the bypass cap id only get a 4.5 gain factor, but if i place a volume / gain on the end to control the gain when using the Bypass to control the output it will clip hard to the point where the signal will distort when i'm after a clean boost.

Are their any trick I can use so that I can obtain a higher gain factor (say 10), have the circuit so it doesn't clip and also keep a somewhat suitable Q point for Class A?

[R.G.]

Use a resistor or pot in series with the emitter cap, but make the cap BIG. This pushes the RC filter frequency so low that it's below audio. In this case, you might double the gain to ten-ish by using a 1K resistor and a cap bigger than 10uF to get down to 15Hz-ish. Be generous - use a 22uF or a 47uF. This ought to get the low frequency rolloff effects well below the frequencies the human ear can detect.

As a side note, one reason opamps have eaten a lot of the analog design world is that they're simpler to design with, and only an eyelash more complex on the board. In your case, you're using a three-leg transistor, four resistors and three caps. An opamp circuit could use two resistors for biasing ( just like the transistor) one resistor for a feedback resistor (like the collector resistor) and one resistor for setting gain (like the emitter resistor). Both need a capacitor in and out, so they're equal there. The emitter cap can be replaced as the cap to block DC from the gain setting resistor. Now you get any gain you want, with high input impedance and high stability of operating point.

The "penalty" you pay for this is that the opamp has two more pins to connect: power and ground to the chip; and is in a bigger package, a DIP-8. Notice that you can get a second opamp for free in that same package, so that's some payback for using the bigger package.

[PRR]

> a higher gain factor (say 10), have the circuit so it doesn't clip

If you only listen to a 100mV signal generator..... 100mV times 10 is 1000mV or 1V peak. You have 9V of battery and should be able to swing most of that. Say 6V peak-peak or 3V peak. No clipping.

Many musicians use guitars. Depending on pickup and pluck-factor, output may be under 20mV or over 500mV. A 500mV peak, times 10, is 5V peak or 10V peak-peak, which will not fit under a 9V rail no way no how (except dented).

> filter which i wouldn't want

*Everything* has filters in it. Nothing is flat to infinite frequency, and we really don't want flat-to-DC in audio work. What R.G. says-- pick your filter-corner to be WAY out of your way. Below 20Hz is low-enuff for most small simple audio. (Large systems with many dozen stages must aim much lower.) Guitar may not need to go so deep: you usually *don't* want to boost your heavy 82Hz open-string as much as the soprano tones. It just flubs the speaker and annoys your bass player.

> somewhat suitable Q point

Don't see a problem with your op point. 1mA?? Collector is mighty close to 4.3V. Unloaded it would swing up 4.7V, however 10K external load on 5K Rc gives about 2/3rd or 3V peak positive. Going down, Re has about 1V and if fully bypassed then Rc could pull down 4.3V-1V= about 3V. If unbypassed then Re will swing up another V for nearer 2V peak negative swing. You might fiddle the collector idle voltage up a hair, but it's already so close to optimum that it won't be any WOW difference.

A question you did not ask: what is the input impedance and what does that do to guitar (which is sensitive to loading)?

You have a load on the output, bravo!! But that pot can go to zero ohms. Is that likely in real life? What is the usual range of loads in a guitar chain? In pencil-checking (or sim) it may be good enough to check the highest load and the lowest load, assume the in-betweens will come out in-between. (Sometimes they don't but here they will.)

When you work-out all the interactions, you *may* see wisdom in R.G.'s tip to use a 13-transistor amplifier (TL07x) when a 1-transistor design turns out marginal in several ways. However I fully approve of spinning the variations and learning the performance limits of 1- and 2-transistor amplifiers.