Simple bridged amplifier design question

If both outputs are AC coupled to the speaker, there is no common-mode voltage across the speaker. :-)

But I take your point, certainly -- without common mode feedback or something similar, the output DC bias points are surely going to be a bit different. Thanks for the link...

---Joel

But then there will be an electrolytic that dries up or oozes out after a few years :-)

[...]

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Hmm... good point... presumably you'd have this same problem if you did take two standlone commercial "mono blocks" and bridged them though, right?

My gut feeling is, "sure!", but I'm only a tech - a real engineer would worry about phase shifts and compensation and all that schtuff.

Why not just slap one together and see if it flies?

Good Luck! Rich

If by mono-block you mean half-bridge amps it can be ok sans cap. They are supposed to keep DC close enough to zero. So if you make a full bridge out of two of those you'd get twice the potential offset. But you also need a speaker with four times the power rating which is then likely able to stomach that increased offset.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

I thought the idea was that the engineer designs the thing in SPICE and then when, on the bench, it just sits there and oscillates, the tech tells him what to do to actually make it work? :-)

I plan on it... although I do want to SPICE it today or tomorrow first.

---Joel

Actually, you should only need one cap for DC blocking. If ther's no DC current, there's no DC current. :-)

Have Fun! Rich

If you take two single amps and feed two speakers, for the most part equals doubling the voltage or quadrupling the power to one speaker.

greg

in both cases, you are generating twice the air drive.

Assuming phase coherance.

greg

Actually by "mono-block" I just meant some random single-channel amplifier with a capacitively coupled output that's meant to drive a speaker connected to ground -- if you use two of them (one on either side of the speaker terminals, with one amp driven with an inverted input), you're back to those output caps having no DC bias on them.

Here's a nice mono-block amplifier for you:

:-) Says it's not bridgeable, though, so apparently it doesn't leave one side of the speaker at ground.

---Joel

I doubt this class of amp has output caps. Either they are fed symmetrical DC or they are already a bridge. Some half-bridge amps can turn unstable if bridge-configured with a 2nd external amp or the mfg just doesn't want early "warranty" returns.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

A mono block could be a bridged amp. Its just meant to play MONO.

Speaking of mono blocks, there is an old Marantz 2 mono blocks for sale. They fit inside or in back of a model 15. I think this was before Marantz went cheap to Superscope.

greg

I envision a differential input, like a long-tail pair, albeit I've got doubts of doing that at 5V; a gain stage or two, and a pair of complementary-symmetry outputs - 11 transistors and a handful or R's and maybe a couple of C's total; but like I said, I'd have qualms about tackling it with only 5V to work with.

Good Luck! Rich

Sure there is! You should put a leakage resistor to GND to define that voltage, too. :)

The point of my circuit is, common mode voltage is *undefined*, so let's put feedback around it to define it.

Consider that, in a half of that circuit, which forms the typical diffamp and VAS of a power amplifier, any variation in diffamp bias goes through one side that's simply shunted to +V, and through the other side which is safely part of the feedback loop. So it doesn't matter very much what bias that circuit has.

But taking plus and minus outputs from a diffamp, the other side now matters as well. Feedback is differential only (the diffamp still cancels common mode input), so it is impossible to define common mode output. This results in clipping or crossover distortion as one or the other VAS or output follower slaps the rails. Differential feedback will keep the differential output looking "okay", but it'll be coasting on a bed of snot, careening into the rails in an ill-defined manner.

There are other possibilities with common mode control, too. You could decide to simply keep both outputs as close to 0V as possible, all the time. Then, when one half of the signal swings, one side goes up, the other staying put until the other half of the signal comes. This does invite big crossover distortion, but because it's well defined, it may be managable.

This drive approach resembles H-bridge driving a stepper motor, where the outputs are always in well-defined states (always active high or active low, never off) and 0V output is when both are in the same state (low-low or high-high).

Tim

--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms

The airport gestapo^H^H^H^H^H^H^H police in Frankfurt rousted me for stealing electricity when I plugged my laptop into the wall.

Give some idiot a SMG, he thinks he's somebody.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058
hobbs at electrooptical dot net
http://electrooptical.net

You finally replaced the Wester Electric hand crank powered set? ;-)

--
You can\'t have a sense of humor, if you have no sense!

Jim Thompson a écrit :

educational.

is

solution.

don't

At say 100R RDSon and a targeted 0.1R, that's just only 1K gates, or 167 so14, each arm, so 330 IC.

Speak of light and small...

--
Thanks,
Fred.

Now I don't believe that. I've measured CD4xxx at 250 ohms typical, Vdd =

12V. HC is more than 2.5 times better, and that's at less than half the supply voltage. Hell, I get 70 ohms typical from a 38HE7, and that thing was built in 1960!

Driving an 8 ohm speaker, 1 ohm would be more reasonable... no it won't have extraordinary efficiency, so what. Or even higher, if power density is really important. Even so, that's still an awful lot of gates, even bus drivers (HC240ish?).

Tim

--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms

educational.

is

solution.

don't

Except I doubt the RDSon number you used.

...Jim Thompson

--
| James E.Thompson, P.E.                           |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
 I love to cook with wine     Sometimes I even put it in the food