What is bootstrapping (capacitor)?

[swinginguitar]

Been seeing it used in audio gear design books - using a cap to link a later stage to an earlier stage for improved linearity etc.

what's the theory on it? how is it different from feedback? being capacitance, is DC factored out of the picture?

[Rob Strand]

The idea is to increase the effective input impedance seen at the input to the stage.    it is actually a generalized trick - not just limited to audio or impedance.

In the typical case for audio amplifiers boot strapping increases the effective resistance of a input bias resistor.  The boot strapping is done by taking an output which is in-phase with the input and nearly the same level as the input then applying to bias resistor at the "boot strap point".

Take an input stage without bootstrapping where the effective input resistance of the bias stage is R.  This resistance R appears from the input line to ground.  Here Iin = Vin/R;   Rin = Vin/Iin = R

Now imagine lifting the ground point of that input resistor up and connecting it to a voltage source Vbs = k * Vin.   the input current in this case is iin = (Vin-Vbs)/R = (Vin - k*Vin)/R. Then, Rin = Vin /Iin = Vin*R/(Vin-k*Vin) = R/(1-k). If k is chosen to be close to but less than 1, say 0.9, then Rin = R/(1-0.9) = 10 * R.   ie. a much higher input impedance than the actual resistor value R.

Often boot-strapping is placed around stages which have a gain of just under 1, like a CC amplifer.  In most implementations you will see two resistors used for the bias stage.  Only the resistor connected to the input point affects the input impedance, the other resistor is for DC biasing purposes.

The reason you use it is you would like a high input impedance but the DC biasing conditions generally require you to use lower value resistors.  Boot-strapoing gives you both.

As for improving linearity, it doesn't actually "improve linearity".  It increases the input impedance of the stage and this may allow the previous stage to have a higher gain,  higher gain = more feedback = better linearity.   Another way to get high impedance is to use a transistor current source, this may introduce non-linearity, so the boot-strapping method is a "more linear" way of achieving this.

Boot-strapping is a form of feedback but it the intent is to affect impedance.  You can use bootstrapping without adding or affecting feedback.

[merlinb]

Bootstraing is another name for impedance multiplication. It is a method of making an impedance appear larger or smaller to AC signals, using a simple electronic trick.

The trick is to apply your signal to one end of the impedance, and a similar voltage to the other end. This is normally achieved by buffering the original signal and coupling it to the opposite end of the impedance via a capacitor. By controlling the voltage across the impedance, you control the current through it, and therefore how big it 'appears' to be as far as the original signal is concerned.

When most people talk about bootstrapping they are usually taking about increasing the input impedance of a circuit (I think the IEEE even defines it as that, which is a terrible definition). But there other examples of it too, such as the Miller effect; the inverted signal voltage on the 'other end' of the input capacitance of an inverting amplifier makes that capacitance appear bigger i.e., its impedance appears smaller.

[Gus]

http://www.diystompboxes.com/smfforum/index.php?board=6.0

look at the univibe schematic

Bootstrapping can be used in power amp output driver stage

It can be used to max gain in a circuit
http://folkurban.com/Site/BootstrappingforGain-692.html

do a web search you should find some interesting things

[tca]

Here is an example for a 1W audio amplifier from Mullard's book "Transistor Audio and Radio circuits" (1st Ed.). On the schematic C6 blocks the DC from the speaker and acts also as a bootstrap by connecting to the current source (the R7=510Ohm resistor).

Cheers.

[Kipper4]

Not hyjacking here but i'm reasonably certain that last mullard amp is similar to the Deacy amp i made.

[tca]

Not hyjacking here but i'm reasonably certain that last mullard amp is similar to the Deacy amp i made.

No. The Deacy and earlier SS amps from Mullard's that you can find are transformer coupled (input and output) (see Mullard Reference Manual of Transistor Circuits). The one I mention is capacitor coupled, completely different amps.

[Kipper4]

your right , my bad i forgot about the transformers.

[Gus]

Here is an example for a 1W audio amplifier from Mullard's book "Transistor Audio and Radio circuits" (1st Ed.). On the schematic C6 blocks the DC from the speaker and acts also as a bootstrap by connecting to the current source (the R7=510Ohm resistor).

Cheers.

Note the bootstrapping has DC current in the speaker. There are other ways of bootstrapping the drive for the output stage

You could split R7 into 2 resistors that sum to 510k and connect a new cap from the emitters to the center node and connect the top resistor to -9VDC,
disconnect the speaker from the "top" of the 510K

Maybe they bootstrapped like that to save two parts an added cap and resistor

Another link for a power amp http://www.passdiy.com/pdf/citation.pdf

[tca]

Here is an example for a 1W audio amplifier from Mullard's book "Transistor Audio and Radio circuits" (1st Ed.). On the schematic C6 blocks the DC from the speaker and acts also as a bootstrap by connecting to the current source (the R7=510Ohm resistor).

Cheers.

Note the bootstrapping has DC current in the speaker. There are other ways of bootstrapping the drive for the output stage

You could split R7 into 2 resistors that sum to 510k and connect a new cap from the emitters to the center node and connect the top resistor to -9VDC,
disconnect the speaker from the "top" of the 510K

Maybe they bootstrapped like that to save two parts an added cap and resistor

Another link for a power amp http://www.passdiy.com/pdf/citation.pdf

DC current in the speaker is quite small... not enough to damage it.

[Gus]

tca

Yes I understand the current is limited in the speaker.  I was posting to show another way to do the bootstrapping.

[tca]

Yes I understand the current is limited in the speaker.  I was posting to show another way to do the bootstrapping.

ah, ok. I was reading Teemu's book about bootstraping and he notes that this kind of bootstrap does not appear in hi-fi or professional equipment, only in small radios that were made at the time (60's?). I guess just to save some $ in an extra capacitor and resistor.

Cheers.

[swinginguitar]

Bootstraing is another name for impedance multiplication. It is a method of making an impedance appear larger or smaller to AC signals, using a simple electronic trick.

The trick is to apply your signal to one end of the impedance, and a similar voltage to the other end. This is normally achieved by buffering the original signal and coupling it to the opposite end of the impedance via a capacitor. By controlling the voltage across the impedance, you control the current through it, and therefore how big it 'appears' to be as far as the original signal is concerned.

When most people talk about bootstrapping they are usually taking about increasing the input impedance of a circuit (I think the IEEE even defines it as that, which is a terrible definition). But there other examples of it too, such as the Miller effect; the inverted signal voltage on the 'other end' of the input capacitance of an inverting amplifier makes that capacitance appear bigger i.e., its impedance appears smaller.

so rather than the "bottom" end of the impedance going to ground, it is now connected to a voltage moving with the input signal ie swinging up and down in sympathy to the input?

is a capacitor used to block DC from the following stage getting back? and/or control what frequencies get  sent back?

[merlinb]

so rather than the "bottom" end of the impedance going to ground, it is now connected to a voltage moving with the input signal ie swinging up and down in sympathy to the input?

is a capacitor used to block DC from the following stage getting back? and/or control what frequencies get  sent back?

Yes, the capacitor is needed to preserve the DC conditions in the different parts of the circuit. We are, after all, only interested in making the impedance appear larger (or smaller) to AC signals. (If we wanted a larger impedance at DC then presumably we would just use a bigger impedance!)

[rring]

Lot of good stuff has been said already but I would point out that bootstrapping is by definition positive feedback(in phase) and while it can create effective AC impedance increase - it will increase noise and decrease linearity . Series negative feedback will increase impedance and improve linearity but at the expense of gain. For example a source or emitter resistor (un-bypassed).

[Paul Marossy]

Series negative feedback will increase impedance and improve linearity but at the expense of gain.

I don't know with 100% certainty if negative feedback in tube amps is the same concept, but I have noticed that it reduces gain. In the one case I'm thinking of right now, removing the negative feedback sure made it sound a lot worse.  

[merlinb]

Lot of good stuff has been said already but I would point out that bootstrapping is by definition positive feedback(in phase) .

Well, that's half correct! Bootstrapping *may* use positive feedback (to increase an impedance), but it may also be negative feedback (to decrease an impedance) e.g. Miller effect. It certainly isn't positive feedback 'by definition' (unless you only adhere to the bogus IEEE definition of the word).

[wavley]

Series negative feedback will increase impedance and improve linearity but at the expense of gain.

I don't know with 100% certainty if negative feedback in tube amps is the same concept, but I have noticed that it reduces gain. In the one case I'm thinking of right now, removing the negative feedback sure made it sound a lot worse.  

If the amp is designed to not have negative feedback then it can sound really great, like the Rivera designed Fenders.  It reduces gain, but increases headroom (because you aren't running flat out all the time), and extends frequency response.  It also lowers the output impedance.  Aiken has a great article http://www.aikenamps.com/GlobalNegativeFeedback.htm

[Paul Marossy]

Series negative feedback will increase impedance and improve linearity but at the expense of gain.

I don't know with 100% certainty if negative feedback in tube amps is the same concept, but I have noticed that it reduces gain. In the one case I'm thinking of right now, removing the negative feedback sure made it sound a lot worse.  

If the amp is designed to not have negative feedback then it can sound really great, like the Rivera designed Fenders.  It reduces gain, but increases headroom (because you aren't running flat out all the time), and extends frequency response.  It also lowers the output impedance.  Aiken has a great article http://www.aikenamps.com/GlobalNegativeFeedback.htm

Yeah, of course if it's designed to not have it that's one thing. But in at least one case, on an old amp I was messing around with, the negative feedback definitely had to stay as it sounds kinda bad without it.

[wavley]

Series negative feedback will increase impedance and improve linearity but at the expense of gain.

I don't know with 100% certainty if negative feedback in tube amps is the same concept, but I have noticed that it reduces gain. In the one case I'm thinking of right now, removing the negative feedback sure made it sound a lot worse.  

If the amp is designed to not have negative feedback then it can sound really great, like the Rivera designed Fenders.  It reduces gain, but increases headroom (because you aren't running flat out all the time), and extends frequency response.  It also lowers the output impedance.  Aiken has a great article http://www.aikenamps.com/GlobalNegativeFeedback.htm

Yeah, of course if it's designed to not have it that's one thing. But in at least one case, on an old amp I was messing around with, the negative feedback definitely had to stay as it sounds kinda bad without it.

I forgot to add that a lot the amps I've worked on that didn't have negative feedback, although they might have sounded great, were on the hairy edge of bursting into uncontrollable oscillation, I remember going round and round with one particular Tonemaster.  I meant to have a caveat in there.

I've been playing around with the negative feedback in one of my amps and put a pot in to control it.  I also tried it without and got the same results you did.  One of the things I find that I don't like in negative feedback circuits is the cap that is like a presence control wired on 10, I've ditched that cap in every amp I own that has one, I keep saying that I'm going to put a proper presence control in, but I never do.