lm3886 super amp

[xplod1236]

Somebody tell me if there is anything wrong with this idea other than that it is dangerous because of the high voltage.
I have a +/- 75v ps from an amp, and I was thinking about using lm3886s with it. I would parallel 5 chips and set them up to run off the positive supply and gnd. The other 5 would run off the negative supply and gnd. The chips connected to the positive rail would only reproduce the positive half of the wave, while the chips connected to the negative rail would reproduce only the negative half. I know that I would have to set up virtual grounds at +/- 37.5v. I checked this setup in a simulator, and it should work.
Can these chips handle voltage that high? In the datasheet it says that |v+| + |v-| has to be less than 84v, and in this case, it would be 75v.

[joensd]

The circuit you have posted won´t work.
Using single supply you must set a voltage reference for each opamp at half the single supply in order to make it able to swing let´s say from 0-75V.

Have a look at the LM3886 datasheet how to connect for proper single supply operation.

If your transformer has got two primary windings (each 120V) I´d wire them in series so you effectively get half the secondary voltage.
IMO a better solution then the single supply approach.

Did you try the +-75V supply before or why "xplod" ?

regards
Jens

[xplod1236]

Originally posted by joensd
Using single supply you must set a voltage reference for each opamp at half the single supply in order to make it able to swing let´s say from 0-75V.

That is correct if the chip is used by itself, so you get a push-pull amplifier. I only need each set of chips to do one thing--one set will push (+v), while the other set will pull (-v).

Did you try the +-75V supply before or why "xplod" ?

No, I did not try this before. And the xplod is from sony xplod subs that i used to have.

[xplod1236]

forgot to attach another picture

[djk]

If you want something that will work, just build a bridge amp.

The transformer can be re-wired as a full-wave center tap with two diodes to give 75V, the bridging allows for a 150V P-P into the speaker.

If you're bored, build the bridge with the patented 'X' feedback scheme as suggested by Nelson Pass.

I could show you how to make your scheme actually work, but it will exceed the Vceo of the output devices in the chips.

[carlosfm]

Originally posted by joensd
The circuit you have posted won´t work.

You took the words out of my mouth...
How can a circuit that doesn't work be simulated?

[janneman]

Originally posted by carlosfm


You took the words out of my mouth...
How can a circuit that doesn't work be simulated?

The simulator doesn't know that it has to blow up the chips when it goes over the max voltage..
The map is not the world!

Jan Didden

[djk]

What we need is a simulator with a sound and smell option (BANG! and Magic Smoke!),

or,

A hand coming out of the monitor to slap us with a fish (Monty Python sketch).

[sam9]

"LM3886" and "super amplifier" are terms I'm not sure go together too well. However, if you want to get some use out of your 75-0-75Vdc PS (which I presume means about a 52-0-52 Vac trransformer) here's an option that is not "super" could be pretty damn impressive:

Connect the two secondaries in parralel to get a 26-26 transformer. There will be no centertap so you need to look as Rod Elliot's project #43, fugure 2. Build a PS like this wit suitable mods for the voltage and current involved suvh as providing +&- power to the opamps by using a couple of zenners. The higher rails have no impact here because the opamps are just dealing with the DC offset. This gets you about a 35-0-35 PS (with just about the most perfect 0V dc offset you'll ever see.

Now, on the same website check out projects #9 or #81 and also read the artical on biamping. Choose one of the projects and buy the PCBs (Rod deserves the trade!)

Depending on the current available from you reconfigured transformer and PS, you should be able to drive 4 or 6 LM3886's very well. Configure these as either a bi- or tri-amp. You should be able to get everthing into one fair sized box.

Now you just need to mod your existing speakers if they don't already have bi/tri amping provisions. This may not ne a super amp with regard to power level, but if you select the XO points right and work carefully you stand a fair chance of getting a "super audio system" emphasis on system.

This is just a suggestion, of course. It's the best way I can think at the moment to get something really remarkable from the combination of LM3886's and the trandformer you describe.

[djk]

"Connect the two secondaries in parralel to get a 26-26 transformer"

Gee, I think someone already said that. If its salvage from a commercial amp it likely won't have 52-0, 52-0 secondaries. Most likely it will have 52-0-52

"The transformer can be re-wired as a full-wave center tap with two diodes to give 75V, the bridging allows for a 150V P-P into the speaker."

"Depending on the current available from you reconfigured transformer and PS, you should be able to drive 4 or 6 LM3886's very well. Configure these as either a bi- or tri-amp. You should be able to get everthing into one fair sized box."

Unless you bridge, you will need to use an output coupling cap for each LM3886 since you only have a single-ended supply. You need to re-read the ESP article.