Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Whas better dual 30w speakers or 1 120w speaker

Status
Not open for further replies.

spitso

Member
(Big mistake in tittle i meant dual 60w speakers not 30w)

Hi all

Im making an amplifier based on the lm4780 ic which is a stereo 60w or mono 120w.
Just wondering what would be best as in loudness, it would be 120w correct?but quality wont be as good?

thankyou
 
60W x 2 = 120W

same amount of power

it will mostly come down to the speakers, look at the SPL (sound pressure level)
Shavano Music Online - Speaker SPL - How much power do you need? is one of the first google results for "spl speaker"

so youll have to figure out whether a dual 60W setup will have more measured SPL than a single 120W speaker. and then compare price.

The quality (i.e., measures like THD) will be the same whether the chip is doing 2x60W or 1x120W, as I believe it simply runs both amplifiers in parallel to make 1x120W.

However it will probably "sound" better with the stereo separation.
 
Last edited:
The quality (i.e., measures like THD) will be the same whether the chip is doing 2x60W or 1x120W, as I believe it simply runs both amplifiers in parallel to make 1x120W.
Probably uses a bridge configuration to get the extra power. You can't parallel amps.
 
Reminds me of the speaker project in one of the U.S. hobbyist magazines back in the 1960s, the "Sweet 16". A speaker system consisted of a 4 x 4 array of either 6-inch or 8-inch speakers. They were cheap speakers but the system had good bass response. The size of a speaker cabinet could hardly compare with a modern-day Bose!

Dean
 
It doesn't make any difference does it?

I'd rather go for two speakers and have the option of stereo.
 
I think The Sweet 16 speaker was made with small cheap speakers but all different.
The designer wanted all their differences to average into a smooth flat frequency response.
But they all had a fairly high resonance frequency that produced no deep bass and a boomy sound. Their high frequency response was all over the place. Their phasing caused the sound to be very directional.

The Bose speaker with 9 little speaker drivers had no bass and no treble without its equalizer. The EQ response boosted the lows and highs and its response curve looked like a Happy Face.:D
 
ok cheers :)
Im going to go with the stereo configuration.

Do you believe it would be possible to run a subwoofer aswell? if so how should it be connected?
 
you would probably want a separate amp for a subwoofer. You could however build speaker cabinets with 3 speakers each: 1 tweeter, 1 midrange, and 1 woofer (not "sub" woofer) and then use crossover filtering to drive each speaker in its desired frequency range. (i.e., all 3 driven from the same amp) If done properly, this would probably increase the sound quality.
 
Just looked it up in wikipedia, didnt realize that. Though the configurations are similar. Either way it's the same two amps sharing the load.

Because it's not true - you CAN parallel amps, if you do it properly, and it's relatively common - again it's down to speaker impedances. In all three configurations the output power is identical.
 
Because it's not true - you CAN parallel amps, if you do it properly, and it's relatively common - again it's down to speaker impedances. In all three configurations the output power is identical.

At least I learned about the bridged configuration, which makes more sense to me anyway. Seems like it would eliminate the problems with paralleling amps, such as matching. How do you get around that problem? because it seems like its bound to fail eventually if the only way to get around it is "ensure good matching"
 
At least I learned about the bridged configuration, which makes more sense to me anyway. Seems like it would eliminate the problems with paralleling amps, such as matching. How do you get around that problem? because it seems like its bound to fail eventually if the only way to get around it is "ensure good matching"

You use identical amps, and design it correctly, the gain-clone IC's are specifically designed to be used in that way - like I said, it's ALL to do with speaker impedances:

Stereo - 2 x 60W (120W) in to 2 x 4 ohm speakers.

Parallel - 1 x 120W in to 2 ohm speaker (or two 4 ohms in parallel).

Bridged - 1 x 120W in to 8 ohm speaker (or two 4 ohm in series).

Notice, in all cases the power is EXACTLY the same, and can even use EXACTLY the same two 4 ohm speakers. If you're using two speakers though, it makes most sense to run it as 2 x 60W - other configurations are useful if you only have a single speaker (and of the correct impedance).
 
Just looked it up in wikipedia, didnt realize that. Though the configurations are similar. Either way it's the same two amps sharing the load.
No it isn't. In parallel (if it was possible) both amp outputs would drive exactly the same voltage swing across the speaker load and thus deliver half the current. However, since the outputs are not exactly alike, they would fight with each other and be sourcing or sinking current from one output to the other amp's output. That's why you can't simply strap two amp's outputs in parallel to try to drive lower impedance speakers.

A bridge is not a sharing proposition, each output drives independently of the other in mirror image of it. This doubles the voltage range the output can swing. Each output drives the same voltage swing as it did in single mode, but one goes up as the other goes down, doubling the voltage of the signal to the speakers. The outputs can be pushed harder for current capability since they are driving the same speaker impedance to higher voltage swing levels requirring more current capability.
 
Last edited:
Because it's not true - you CAN parallel amps, if you do it properly, and it's relatively common -
That's news to me, but I've only been doing it for 40 years. If amps are designed to run in parallel, they would have to contain some internal circuitry specifically to make one amp the "master" and the other the "servant" so that they don't try to force slightly different voltages at the load. They don't have exactly the same gain, so they would be trying to establish differet voltages as set by their own feedback.

It's theoretically possible to build in 'sharing" control circuits, I haven't seen it in audio amps, just high current power supplies.
 
again it's down to speaker impedances. In all three configurations the output power is identical.
If you mean for a specific speaker impedance, that is not correct. Bridge amps are used specifically to increase output power for a given impedance by increasing the voltage swing at the load. A single stereo amp used in a car (14V rail) is biased to center line (7V) and swings about 5V up and down from 7V. That means the speaker sees a signal voltage swing of 10V (p-p) which is AC coupled to the speaker whose other terminal is grounded. In a bridge with the same 14V rail, the speaker terminals "float" and one side goes up and the other goes down. It can swing about 12V each way, so that makes 24V (p-p) possible voltage excursion creating much more power for the same speaker impedance which is P = V(sq)/R where R is the speaker impedance.
 
Last edited:
How do you get around that problem? because it seems like its bound to fail eventually if the only way to get around it is "ensure good matching"
You have to have one of the amps voltage control loop being the master and the other amp having it's voltage loop disabled so it will act like a current source whose current is dictated by the master amp.

I have done this many times in power amps and power supplies where you need to be able to strap outputs directly in parallel and have them share current. The one I designed for the dual output power supply sitting on my work bench does this and the servant output supplies current at any ratio from 0 to 100% of the other output adjustable. At 100%, the two outputs share current equally. It is done by using the current signal from the master side to program the servant's current control amplifier. The master runs in constant voltage mode (voltage control loop active) and the servant runs in constant current mode.


As for relying on "good matching".... I thinks that's insane, I sure would not do it. The amps would not only have to be dead on matched, you would have to zero out the DC offset between them. And as they (or their components) heat up, gain and offset are going to change. Remember an audio power amp has a super low output impedance (like a tenth of an Ohm or less) and it's voltage feedback loop is trying to force a voltage across the speaker load. They have high loop gain (like 80 dB) so even with error voltages of a few millivolts, they are going to try to force it to change. And with such a low driving impedance, they can really crank out current to try to force the other output up or down. I wouldn't trust it.
 
Last edited:
If you mean for a specific speaker impedance, that is not correct. Bridge amps are used specifically to increase output power for a given impedance by increasing the voltage swing at the load. A single stereo amp used in a car (14V rail) is biased to center line (7V) and swings about 5V up and down from 7V. That means the speaker sees a signal voltage swing of 10V (p-p) which is AC coupled to the speaker whose other terminal is grounded. In a bridge with the same 14V rail, the speaker terminals "float" and one side goes up and the other goes down. It can swing about 12V each way, so that makes 24V (p-p) possible voltage excursion creating much more power for the same speaker impedance which is P = V(sq)/R where R is the speaker impedance.

I'm going to run some simple numbers here...

assume your amplifier is designed for a 10Vpp swing:
Single Amp: 10Vpp into 8Ω = 1.25App --> 1.25A*10V = 12.5W = 20V^2/8Ω
Bridge Amp: 20Vpp into 8Ω = 2.50App --> 2.50A*20V = 50.0W = 20V^2/8Ω
Parallel Amp: (each amp "sees" half the load of an 8Ω speaker):10Vpp into 4Ω = 2.5App --> 2.5A*10V = 25W = 10^2/4Ω
(This assumes the amps are perfectly matched, just for sake of calculation)

So a parallel amp will double the power of a single amp, whereas bridge configuration power goes as V^2.
However if your amps are limited in current drive to, say, 1.25A, the only way is to parallel them...

That is something that is not intuitive... that although you have two amplifiers one configuration has more power than the other.
The power of the chip the OP posted about is doubled when run in mono mode, which speaks of paralleling to me!

The **broken link removed** shows application circuits for both bridge and parallel schemes; the parallel connection uses some load-balancing resistors to get around the matching problem. Simple enough solution, if not efficient. Should have thought of that a few posts ago.

My guess is that the output stage is limited in its current-drive abilities so if you go for the bridged configuration you will run into some nasty problems above 120W.
 
Last edited:
I'm going to run some simple numbers here...

assume your amplifier is designed for a 10Vpp swing:
Single Amp: 10Vpp into 8Ω = 1.25App --> 1.25A*10V = 12.5W = 20V^2/8Ω
Bridge Amp: 20Vpp into 8Ω = 2.50App --> 2.50A*20V = 50.0W = 20V^2/8Ω
Parallel Amp: (each amp "sees" half the load of an 8Ω speaker):10Vpp into 4Ω = 2.5App --> 2.5A*10V = 25W = 10^2/4Ω
(This assumes the amps are perfectly matched, just for sake of calculation)

So a parallel amp will double the power of a single amp, whereas bridge configuration power goes as V^2.
However if your amps are limited in current drive to, say, 1.25A, the only way is to parallel them...

That is something that is not intuitive... that although you have two amplifiers one configuration has more power than the other.
The power of the chip the OP posted about is doubled when run in mono mode, which speaks of paralleling to me!

This is wrong, actually backwards:

Parallel Amp: (each amp "sees" half the load of an 8Ω speaker):10Vpp into 4Ω = 2.5App --> 2.5A*10V = 25W = 10^2/4Ω
(This assumes the amps are perfectly matched, just for sake of calculation)

If you parallel the outputs of two amplifiers, each one supplies half the current and sees TWICE the effective load impedance. same voltage swing, half the current, effectively sees twice the impedance. The same voltage swing across the speakers, same power delivered to the speakers, but the current is divided between the two outputs. No power gain at all.

The reason to parallel outputs is if you want to drive lower impedance speakers (or parallel speakers which cuts their impedance in half) and the current required to drive that lower impedance is too much for one output.

The power of the chip the OP posted about is doubled when run in mono mode, which speaks of paralleling to me!
Unless you lower the speaker impedance somehow, paralleling outputs can NEVER increase power to the load. Impossible.

Power = V (sq) / R

Unless you increase the voltage swing or reduce the impedance, you can not increase power.
 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top