This is a difficult thing to get my head around. 
Imagine an opamp with it's zero volt power rail referenced to the speaker output connection of a unity gain power amp it is feeding. If we make the opamp a x1 buffer it works fine and it's output swing is not limited by it's supply rails because they move up and down together with the output signal. What I want to do though is to make this opamp stage have a x10 gain. This is a bit more tricky than it first appears. Can anybody help me here?
Imagine an opamp with it's zero volt power rail referenced to the speaker output connection of a unity gain power amp it is feeding. If we make the opamp a x1 buffer it works fine and it's output swing is not limited by it's supply rails because they move up and down together with the output signal. What I want to do though is to make this opamp stage have a x10 gain. This is a bit more tricky than it first appears. Can anybody help me here?
Graham: Assuming that I understand your question correctly 🙂, bootstrapped cascoding is usually accomplished through one of two process, and depending on the particular process, can be considered as a form of either feedforward or feedback.
For feedforward operation, the reference for the bootstrapped cascode is derived from the same signal used to drive the circuit being cascoded. In other words, you take the input signal, level-shift it, and feed it to the cascode. This method works best when the gain is 1 - otherwise you will need an additional gain circuit to make sure that the cascode accurately tracks the circuit being cascoded.
For feedback operation, the reference for the bootstrapped cascode is derived from the output signal of the circuit being cascoded. Since it is merely following the output signal and level-shifting it, the bootstrapped cascode can remain blissfully oblivious of whatever gain factor is used for the circuit being cascoded.
I use both methods regularly in my own circuits, and can confirm that they are great tools in practice - both dependable and invaluable.
Please let me know if this is what you wanted to know, as my answer is admittedly absurdly simple. Recalling some of the circuits that you have cooked up in the past 😀, I fear that you may have had something much more complex and sophisticated in mind when you initiated this thread.
regards, jonathan carr
For feedforward operation, the reference for the bootstrapped cascode is derived from the same signal used to drive the circuit being cascoded. In other words, you take the input signal, level-shift it, and feed it to the cascode. This method works best when the gain is 1 - otherwise you will need an additional gain circuit to make sure that the cascode accurately tracks the circuit being cascoded.
For feedback operation, the reference for the bootstrapped cascode is derived from the output signal of the circuit being cascoded. Since it is merely following the output signal and level-shifting it, the bootstrapped cascode can remain blissfully oblivious of whatever gain factor is used for the circuit being cascoded.
I use both methods regularly in my own circuits, and can confirm that they are great tools in practice - both dependable and invaluable.
Please let me know if this is what you wanted to know, as my answer is admittedly absurdly simple. Recalling some of the circuits that you have cooked up in the past 😀, I fear that you may have had something much more complex and sophisticated in mind when you initiated this thread.
regards, jonathan carr
Hi,
did You see TNT HDOA on www.tnt-audio.com? Maybe that's what You're looking for...
Keep well.
Jack
did You see TNT HDOA on www.tnt-audio.com? Maybe that's what You're looking for...
Keep well.
Jack
I was thinking that cascoding a JFET buffer (probably use it in a line amp or head amp) would actually approach the "absurdly simple" implementation. Since JFETs are depletion mode, their source is at a higher potential than the base (for N JFETS) so JFET cascodes can self bias.
I was thinking about the feedback/feedforward differences. They are more similar than not, but the feedback method looks like a cascode and the feedforward method looks like a tracking regulated power supply. Just an interesting observation.
-Won
I was thinking about the feedback/feedforward differences. They are more similar than not, but the feedback method looks like a cascode and the feedforward method looks like a tracking regulated power supply. Just an interesting observation.
-Won
Do a search at Linear Technology and Analog Devices on 'suspended supply' operation. The gain calcs are complicated and resistors will have to be selected to better than 1%
To me it makes more sense to use the Hafler 'Trans Nova' operation. The opamp supplies are referenced to ground and outputs drive ground. The center tap on the transformer becomes 'hot'.
QSC, Crown, Hafler, etc, all use this circuit.
To me it makes more sense to use the Hafler 'Trans Nova' operation. The opamp supplies are referenced to ground and outputs drive ground. The center tap on the transformer becomes 'hot'.
QSC, Crown, Hafler, etc, all use this circuit.
- Status
- Not open for further replies.