Bridge Amplifier

Good idea

Hi Daniel, it is definitely possible.

Here is a very good example - Accuphase A680 (see attached schematic).

Symmetric input stage, then mirrored common base VAS. Each side of it works on separate drivers and OPs.

Two DC offset integrators, 2 global NFB loops, balanced input.

Extremely elegant...

Cheers,
Valery

DC servos are optional - you can always use capacitors in NFBs, same as for the normal amp.

This is just a good industry example - you can build a simpler desing based on the same principles... I will think if I can throw-in some ideas.

Kind regards,
Valery

The Hafler website has schematics for the P and Pro series amps with bridging schematics IIRC the Crown DC300a also had a similar bridging circuit GOOGLE is your friend...


Regards, Elwood

Tricky servo

Here we go

Sorry - could not avoid DC servo here.
The problem is that in such a bridged design it is important to stabilize the DC parameters and the best way to do it I see at the moment is to stabilize the LT current via DC servo with open collector, controlling long tail. When it is stabilized this way, all queiscent currents are set and the outputs are balanced between the rails. It is possible to do it manually with a trimmer, but because of temperature changes, elements tolerance, etc. it will not be stable. Integrator provides auto-balancing.

I have also played with some values - not important.

Result - a decent amp, self-balanced, stable, pretty efficient for such a simple design.

I had to artificially ground one of the outputs - otherwise multisim's THD meter does not work (rather inconvenient) - in real prototype the ground would be between the power sources.

By the way, now it is easy to use a single power source if you like - just set the point marked GROUND between the rails (couple of resistors with filtering caps) - DC servo will balance the scheme automatically.

Appeared to be an interesting project in the end
Let me know if you need more explanations.

Cheers,
Valery

P.S. You can try differential VAS - it will work even better. But DC servo will still be required if you want auto-balancing feature.

danny92 said:

I want to know if it's possible to make amplifier with a bridge
topology for single supplies without using two distinct amplifiers.

Daniel

Click to expand...

Hi,

Possible, but hopelessly inaccurate AFAIK. The implication is
only one feedback loop, which cannot set the dc conditions.

If it could be done, it would be common, it isn't, it can't.
Car stuff has exploited every single supply option available.

rgds, sreten.

Single PS option. Same performance.

Just for fun

sreten said:

The implication is
only one feedback loop

Click to expand...

Hi sreten, you can always use two feedback loops in bridged designs.
Then you have to solve a DC stability issue, but here an integrator is your friend (as one of possible solutions).

Regards,
Valery

could also read up on fully diferential I/O op amps - which do have a common mode control loop built in

the common mode loop can be very nearly orthognal to the differential gain - very little interaction between the two


Pass SuSy may also illustrate an option for "integrated" bridging

I just roughly chose the queiscent currents for each stage - 1mA per LTP output (2mA for the long tail), 10mA for VAS and 100mA for the OP. Then I reduced R1, R2 and R7, R8 respectively to increase VAS gain, allowing lower distortion via deeper global NFB. Gain increase required more comprehensive compensation.

C5 between the LTP outputs is a so called lag compensation cap. Actually, this design will work without it, but you will find some signs of oscillation (not criminal, but still) after a fast front, at square wave for example. With C5 in place square wave is nice and clean.

Yes, DC offset at both outputs very much depends on IPS current, so DC servo compares it to the point between the rails and regulates, keeping the output offsets equal to that point. It is a very accurate self-balancing "mechanism". LF411 is very good for this purpose due to low input currents and low-drift design in general. I invented the way of driving the IPS current with "open collector" just yesterday, thinking about your design.

Thank you for inspiration

I want to know if it's possible to make amplifier with a bridge topology for single supplies

Click to expand...

Pass shows at least one amplifier that runs off single polarity supply and yet has balanced output for the speaker load.

Daniel, if you move ahead with live prototype - pls keep me in the loop. Interesting project.