Bootstrapped Op Amp Articles

[mediatechnology]

I moved discussion of the following "Voltage-bootstrapped current-boosted NE5534 swinging op amp for audio line driver providing +27 dBu into 600Ω" circuit to here: https://www.proaudiodesignforum.com/for ... ?f=6&t=927


Voltage-bootstrapped current-boosted NE5534 swinging op amp for audio line driver providing +27 dBu into 600Ω

[mediatechnology]

I happened to be looking at the dBx 160 output stage today and happened to notice that its op amp supply rails appear to be voltage-bootstrapped by a tap on the output stage bias divider string.

I saved this schematic years ago as an example of a ground-compensating output.

Never noticed the mild bootstrapping until today. It looks like at 0V output the LM301 is seeing about +/-19V.


dBx 160 Ground-compensating voltage-bootstrapped line driver output with turn-on mute.

[KMN]

I think I may have an excuse for trying a bootstrapped opamp circuit. In my case to minimize distortion.

I was looking for an audio grade op amp for doing filters at +/-36V rails, where the non inverting configured inputs may see up to full rail voltage. Power output is very low but precision at large and small signal swings is a must.

LT and Analog have some higher voltage rail capable FET input and chopper opamps. Nothing specifically for audio. The highest Vrail high quality audio grade op amps I've found were the opa604's with +/-25 v rails absolute max. Edit: I should add TI does have a bunch of power amps for audio. I was looking for something in an SO8 package or so, not the giant packages in the TI power amps which are way overkill in my circuit. The LTC6090 is a bit of a compromise. SOIC-8 with power pad is workable, I guess. It definitely wont be getting hot with the amount of light loading in my circuit./edit

In ltspice none of those "high voltage" AD or LT parts seem to do very do well in the upper audio band. They all start distorting pretty heavily, even lightly loaded.

Lets just say that if my simulations are to be believed, the LTC6090 bootstrapped about 18V away from the output up to +/-36V rail...wow that thing is clean now per my sims.

I ended up using something like recommended in this article to get the sims looking real sweet. I started with a circuit similar to the one in the EDN article but ended up at something similar to this. Toward the bottom of the page in the link. Sorry idf this one was already linked and I missed it. https://www.embedded.com/print/4217177:

[mediatechnology]

Self did a lot of work (also published in Small Signal Design) with bootstrapping to reduce distortion.

Take a look at Dimitri Danyuk's work here: viewtopic.php?f=12&t=928

[mediatechnology]

"Class AB Amp Uses Two Floating Cells," Joseph Ting, EDN, March 19, 2009.

https://proaudiodesignforum.com/images/ ... h_2009.pdf


Class AB Amp Uses Two Floating Cells Ting EDN March 2009 Figure 1.

[flyboy71]

I just stumbled onto this thread when looking at bootstrapping. I have an idea to do this on an application that is based on audio but would require a single supply and a DC output at 1/2 the HV positive rail. I haven't come across bootstrapping with single supply applications. This particular application involves modulating a tube in an RF transmitter where the screen grid needs a resting positive DC bias and audio would ride on that level. Its actually a bit more complicated but this is the basis to get it working.

[mediatechnology]

I can't think of any additional limitations single-supply operation would impose.

It seems worth trying in your application.

[Tubetec]

Bootstrapping is a term I associate with transistors ,
not so much with valves ,
Is the principle involved ,applying NFB in parallel with the input ?
as opposed to series applied , like a tapped cathode resistor arrangement , which doesnt effect input impedence ,

Could we say a self biased cathode follower is bootstrapped ?
in that case the input impedence appears much greater than the grid resistance value ,

[mediatechnology]

Bootstrapping is a term I associate with transistors ,
not so much with valves ,
Is the principle involved ,applying NFB in parallel with the input ?
as opposed to series applied , like a tapped cathode resistor arrangement , which doesnt effect input impedence ,

Could we say a self biased cathode follower is bootstrapped ?
in that case the input impedence appears much greater than the grid resistance value ,

This thread refers to voltage bootstrapping of the supply rails to extend device signal swing capability versus bootstrapping to raise impedance.

[z11111]

Attached some basic examples of how Pacific Recorders, Auditronics and dbx bootstrapped their opamps for outputs through the 70s into the 90s.

The PR&E looks to be a pretty well thought out approach, like many other Pacific Recorders designs. Originally they were using Jensen 918 discrete opamps and added this scheme when they switched over to DIP8 ICs

The dbx scheme is found in many of their 900 series units, all of which had transformer less outputs, and they kept using the same scheme for the entire 20~ year run of the 900 series.

Auditronics used this same scheme with 709s, 301s, 5534s, 5532s, etc from the early 70s onward. Doesn’t appear to be as robust of an approach and I have repaired some of these configs where the transistors were dead. I’ve replaced their 2N3566 and 2N3638 pairs with more common (and better spec) BC237 and BC337, just gotta flip them 180 degrees because different pinout. They used a 2N2219A and 2N2905 pair on their "A-30" output cards (bus/program outputs -- think API 325) before hitting a Jensen 123 transformer.

Early/mid 70s version (LM709):

Early 80s version (5532):

..and an image of Auditronics whole A-30 output card: