Signal Splitter

[catcat]

Does anyone have an idea how to make a hi-fi audio signal splitter (It accepts one input and gives a inverted and noninverted output). I am going to use it for my audio amp project and I would like it to be simple and with discrete components. A MOSFET version would be good, too.

 

[Nigel Goodwin]

It's called a 'phase splitter' - simple way is a single transistor amplifier, with equal collector and emitter resistors - the collector is inverted, and the emitter is in phase.

Why actually do you want one?.

 

[catcat]

For a birdged audio amplifier.

 

[Nigel Goodwin]

For a birdged audio amplifier.

Use an opamp to invert one channel, it's more accurate!.

For an extra question, what amps are you trying to bridge?, and why? - a LOT of people seem confused by what bridging actually achieves!

 

[Sceadwian]

The only reason I can think of the OP would want an external signal inverter as described would be to manually bridge two seperate mono amplifiers, leting one drive the + lead of a speaker and the other drive the - lead. Most two channel amplifiers have a built in ability to do this to increase output power for a single speaker application. The original poster however does need to clarify what he has and his goals as the signal input impedance and output impedances weren't even hinted at.

 

[audioguru]

Catcat,
Post the audio amplifier schematic you want to bridge, the speaker's impedance and the power supply voltage and current rating. Then we can help you with it.

 

[catcat]

It's Internal.

Bridged because the output power will double, and with 2 identical output stages, the output bias voltage is the same, with the voltage the same, there is no dc through the speaker. And because of that, I do not need output capacitors that act as high pass filters, witc means more bass.

 

[Hero999]

It's Internal.

Bridged because the output power will quadruple, and with 2 identical output stages, the output bias voltage is the same, with the voltage the same, there is no dc through the speaker. And because of that, I do not need output capacitors that act as high pass filters, witc means more bass.

Fixed.

Remember ohm's law? Doubling the voltage quadruples the power? [latex]P=\frac{V^2}{R}[/latex]

Also, there often is a small DC curent through the speaker but it's very small.

 

[Nigel Goodwin]

It's Internal.

Bridged because the output power will double, and with 2 identical output stages, the output bias voltage is the same, with the voltage the same, there is no dc through the speaker.

The output will be the same as the two individual amplifiers, BUT in one speaker rather than two speakers. It's a question of the number of speakers and what impedances they are.

If your two amplifiers output 100W each into 4 ohm speakers (200W total), then bridging them will provide the same 200W into one 8 ohm speaker.

This is why we asked for more details, like the amplifiers amd speakers you're using.

And because of that, I do not need output capacitors that act as high pass filters, witc means more bass.

Makes no difference, the output coupling capacitor (assuming it's such an old amplifier it has one) will be specified to give the required bass response. But anything remotely modern is likely not to have one anyway.

 

[Nigel Goodwin]

Remember ohm's law? Doubling the voltage quadruples the power? [latex]P=\frac{V^2}{R}[/latex]

A popular misconception! - the amplifers can't provide ANY more power, attempting to feed too low a load would 'try' to quadruple the power, but would simply kill both amps!.

 

[audioguru]

A bridged amplifier doubles the voltage across the load and also doubles the current in the load. Therefore the output power is nearly 4 times as much as a single-ended amp. Each amplifier in the bridge must be able to handle the high current and high amount of heating.

You won't get more bass from a bridged amplifier compared to a single-ended amplifier if its output capacitor is large enough.
A 2200uF capacitor feeding an 8 ohm speaker cuts 9Hz and lower. The response at 45Hz and above is flat. No music goes down to 9Hz and no speaker goes that low. Your hearing doesn't go that low.
If the output capacitor is only 22uF then it won't have bass.

A TDA1554Q bridged car radio amplifier IC doesn't have a phase splitter. One of its amplifiers is inverting instead.

 

[audioguru]

Car radio amplifier ICs don't have enough supply voltage for much power, so they use a high amount of current into nearly a dead short to make a lot of power.

 

[catcat]

It is not an ic amplifier, it uses transistors. But after your comments I agreethat the power will quadruple.

and the biggest NP electrolytic cap I can find is 220:mu: F.

I also read that electrolytic caps sound crappy.

 

[Nigel Goodwin]

It is not an ic amplifier, it uses transistors. But after your comments I agreethat the power will quadruple.

I would suggest you try reading my posts! - or you're going to have two dead amplifiers!.

You still haven't given any details on your speakers, amps, or power supply - we can all be more helpful knowing that.

and the biggest NP electrolytic cap I can find is 220:mu: F.

You don't want a NP electrolytic - no reason whatsoever for one!, in fact it would be a BAD thing for a single amp - and for bridging, you remove them.

Are you building an amp from scratch?, if not the amps will already include the required capacitors (if it needs them).

I also read that electrolytic caps sound crappy.

Every CD (or vinyl record) you own has probably already been through hundreds of them in the recording process - nothing wrong with electrolytics in the right place and in a decent design.

 

[catcat]

The amplifier is a class-a complementary wich I will build from scratch, the speaker is a 8ohm 5.25'' woofer with a 8ohm 1'' dome tweeter that has a series 4.7uf polyester film capacitor. Also I only have a single-supply up to 20V.

 

[Nigel Goodwin]

The amplifier is a class-a complementary wich I will build from scratch, the speaker is a 8ohm 5.25'' woofer with a 8ohm 1'' dome tweeter that has a series 4.7uf polyester film capacitor. Also I only have a single-supply up to 20V.

The class A part sounds a waste of time?, but your low supply voltage and 8 ohm speaker really does call for bridging.

Your supply will enable a single amp to provide about 10W in to a 4 ohm speaker, bridging two of them will provide 20W in to an 8 ohm load. Obviously you need to be careful with the amp design, you need as much voltage swing as possible from the amp (my wattage figures are based on 18V p-p output). This pretty well means you can't use FET's as they lose too much voltage.

Is there any reason you can't use a higher supply?, 20V seems a bit obscure?.

 

[catcat]

Here's a diagram.

 

[audioguru]

A class-A amplifier makes too much heat. Amplifiers with complimentary output transistors usually operate in class-AB to remain fairly cool.

A bridged amp using a 20V supply can deliver only about 16W or 18W to an 8 ohm speaker.

A 5.25 inch speaker is too small to be a woofer. Also it is too small to be sensitive. The tweeter will need to be attenuated to match its low output.

 

[Hero999]

A popular misconception! - the amplifers can't provide ANY more power, attempting to feed too low a load would 'try' to quadruple the power, but would simply kill both amps!.

Oh, I know that, of course you need to uprate the design accordingly.

 

[Nigel Goodwin]

Oh, I know that, of course you need to uprate the design accordingly.

So you uprate your 100W per channel 8 ohm amp to provide 200W per channel into 4 ohms (total 400W) - then bridge it to give 400W into 8 ohms. See what I mean, the power isn't any different, it's only the speaker impedance which changes (and goes from stereo to mono). In fact if you take your two 4 ohm speakers, and put them in series (to give 8 ohms), you can even use the exact same speakers!.

The only places bridging is useful is you have too high a speaker impedance (and can't rewire it as lower), or (the biggest reason) to give more power off too low an HT rail - like car amps!.