how to drive a 0.4 ohms load?

[dr.power]

Don't understand why you think they would not be in phase. Each identical amp fed by the same input would have negligible output phase shift difference between them.

Actually I am sure that there would be an out of phase at the final outputs by using independent amplifiers, though I do not know if it is negligible or not.
Anyway If I do not find another solution I will use the independent amplifier method.
Is the output of 2 independent amplifiers connected to 2 independent speakers side by side just 2x power of each? For instance what about using 2 LM3886 each delivering 68W?

 

[dr.power]

Exactly - simply wire the speakers in series/parallel - it's standard practice (fears of damping factor problems are just that, totally ungrounded fears).

As for worrying about phase problems with different amplifiers, that's nonsense as well - PA systems commonly use dozens of different amplifiers with no such problems.

Well, I guess my speakers are normally used and happy with 56Vp-p (20Vrms). I want to use LM3886 which is a 68W power amplifier and the MAX of its voltage supply is 84V, So how to make the speakers in series for this amplifier?
I am getting confused, you know, My LM3886 is currently HAPPAY with EACH speaker, so what are you saying by using a transformer or make the speakers in serries, Plaese somebody enlightens me, I am not able to see what you are talking!!

 

[crutschow]

Actually I am sure that there would be an out of phase at the final outputs by using independent amplifiers, though I do not know if it is negligible or not.
Anyway If I do not find another solution I will use the independent amplifier method.
Is the output of 2 independent amplifiers connected to 2 independent speakers side by side just 2x power of each? For instance what about using 2 LM3886 each delivering 68W?

Still don't understand why you are "sure" there's a phase shift between amps. If it's intuition you are using, that's generally not reliable when determining the operation of electronic amplifiers.

Each amp will have a near identical phase shift (which at the low to mid frequencies is likely very low), thus there will be negligible phase shift between their outputs.

The power of two amps is simply twice that of one.

 

[dr.power]

If you have 10 speakers and play 50W from each one then obviously the amplifiers must provide a total of 500W.

Thanks.
Yea, That's right, that's why a transformer is not able to help, unless I use a 500W amplifier essentially designed for a 4 ohms load, and then use a matching transformer so that not to overdrive the amplifier by 0.4 ohms speakers, right?? Of course I am not able to use a single 68W amplifier for my job and just add a transformer, are you agreed?

An amplifier rated for an 8 ohm load might blow up if the load is 4 ohms or less because it is almost a dead short and the current will be much too high. You could use 5 amplifiers spec'd for 100W into 4 ohms and connect two 8 ohm speakers in parallel to each amplifier.

I just got 4 ohms speakers, not 8 ohms.
I thought of using 10 of LM3886 amplifiers independently suited for each speaker if there is no other better way.

I am wondering why nobody wanted me to use PARALLEL amplifiers (just make several LM3886's in parallel so that drive the all speakers in parallel?)??! I think by doing so I will not have the problem of out of phase for the speakers, right?
I myself guess that by making a parallel pack out of several LM3886's I will have a a net amplifier having much lower output impedance to drive my 0.4 ohms load, Is it correct?

 

[audioguru]

The LM3886 has an output of 68W into 4 ohms when its has a 56V total supply. If you connect two 8 ohm speakers in parallel then each gets only 34W. With a 70V supply the output into 4 ohms is 130W which will destroy the IC.
None of the graphs show a supply more than 56V when the load is 4 ohms and the output is 68W.

If you connect two speakers in series then each one gets only 1/4 of the output power.

 

[dr.power]

The LM3886 has an output of 68W into 4 ohms when its has a 56V total supply. If you connect two 8 ohm speakers in parallel then each gets only 34W.

yea, congratulations, That's why one LM3886 is not sufficient for me, making the speakers in series/parallel or using just a transformer is not useful for a single LM3886 as well.
Please can you tell me how much of voltage a single 4 ohms speaker sees by 56V total supply while using LM3886?

With a 70V supply the output into 4 ohms is 130W which will destroy the IC.
None of the graphs show a supply more than 56V when the load is 4 ohms and the output is 68W.

So how much of load preferably I should have to drive a speaker at the maximum of speaker voltage (i.e 56V)? please remember, I told that my speakers are happy to work at 56Vp-p.

If you connect two speakers in series then each one gets only 1/4 of the output power.

Which of course is a very bad situation, because I will lose the max of output for the speakers. If i wanted to do so, Then I could use the higher impedance speakers. So it is not the right solution.

 

[carbonzit]

I have desired to design and make a line or planar sound source using 10 or more (if needed) common 4 ohms speaker (I have more than 20 of them).

Either I really don't understand what ypu're asking here, or you're just plain being ridiculous.

As I understand it, you have:

• At least 20 4 ohm speakers, other characteristics unknown, and
• At least 12 LM3886 power amplifiers.

You want to drive those speakers (10 or more, according to your original post) with at least one of those amplifiers.

What's the problem? Just wire 16 of the speakers in series-parallel. Specifically,

• wire 4 sets of 4 speakers in series
• wire the 4 sets in parallel

Voila! You have a total impedance of 4 ohms, perfect for your 3886s.

You haven't told us:

• How much total power you require
• What the power-handling capacity of each speaker is
• What you're trying to accomplish here

If you need more power than one 3886 will provide them, then run two or more in bridged mode. They're made so you can do that. (Check the datasheet.)

I don't see what the problem is.

 

[dr.power]

Either I really don't understand what ypu're asking here, or you're just plain being ridiculous.

As I understand it, you have:

• At least 20 4 ohm speakers, other characteristics unknown, and
• At least 12 LM3886 power amplifiers.

You want to drive those speakers (10 or more, according to your original post) with at least one of those amplifiers.

What's the problem? Just wire 16 of the speakers in series-parallel. Specifically,

• wire 4 sets of 4 speakers in series
• wire the 4 sets in parallel

Voila! You have a total impedance of 4 ohms, perfect for your 3886s.

You haven't told us:

• How much total power you require
• What the power-handling capacity of each speaker is
• What you're trying to accomplish here

If you need more power than one 3886 will provide them, then run two or more in bridged mode. They're made so you can do that. (Check the datasheet.)

I don't see what the problem is.

I think that I respond almost all of your questions in my previous posts.

How do you want me to put the speakers in series/parallel and use a single LM3886 yet get the MAX power out of the speakers???!!!

Suppose I put 2 of the speakers in series, Now I'll get 8 ohms of load, I have to increase the voltage of the LM3886 by a factor of 2 (i.e 112V) to get the same power as using a single speaker. It is not possible. Hope that makes sense.

 

[carbonzit]

I think that I respond almost all of your questions in my previous posts.

How do you want me to put the speakers in series/parallel and use a single LM3886 yet get the MAX power out of the speakers???!!!

Suppose I put 2 of the speakers in series, Now I'll get 8 ohms of load, I have to increase the voltage of the LM3886 by a factor of 2 (i.e 112V) to get the same power as using a single speaker. It is not possible. Hope that makes sense.

No, it does not make sense.

It sounds like you're saying you won't be satisfied until you can get the "maximum power" (i.e., the 68 watts that the 3886 is capable of providing) out of each speaker. Meaning that if you have 10 speakers, you are demanding 680 watts of power?!?!?

Which is why I asked what you're trying to accomplish here. It's possible you have some strange application where you simply must get the maximum possible power out of each speaker.

Do you even realize how much power 68 watts is? And are you going to be running this whatever-it-is at maximum power, wide open?

Generally speaking, audio amplifiers (with some exceptions) are not expected to be run at their full rated power.

The solution I offered you--wiring your speakers in series-parallel--will work with what you have. It's up to you to explain to us why that is not an acceptable solution. I still don't get it.

 

[crutschow]

Suppose I put 2 of the speakers in series, Now I'll get 8 ohms of load, I have to increase the voltage of the LM3886 by a factor of 2 (i.e 112V) to get the same power as using a single speaker. It is not possible. Hope that makes sense.

As stated by others, you put the speakers in series-parallel. Thus you have two sets of two speakers in series and then parallel the two sets. That puts you back at the original impedance of 4 ohms. Now you will get the same total power at the same amp voltage as you would with one speaker. Of course, the power is now shared between 4 speakers but that's to be expected.

Perhaps you need to review ohms law of parallel and series circuits.

 

[audioguru]

Please can you tell me how much of voltage a single 4 ohms speaker sees by 56V total supply while using LM3886?

I don't know where you got a signal that is 56V p-p into 4 ohms because it is 100W but an LM3886 has a max output of 68W when its supply is 56VDC. An applications note for the LM3886 IC explains that its SPIKE protection circuit turns on when the output is 68W into 4 ohms which is is 16.5V RMS which is 46.6V p-p.

Simply use one LM3886 for each speaker and use a 56VDC supply or plus 28VDC and minus 28VDC. The phase of the sound from a speaker will be identical to the others (unless a speaker is connected backwards).

 

[carbonzit]

As stated by others, you put the speakers in series-parallel. Thus you have two sets of two speakers in series and then parallel the two sets. That puts you back at the original impedance of 4 ohms. Now you will get the same total power at the same amp voltage as you would with one speaker. Of course, the power is now shared between 4 speakers but that's to be expected.

Carl, I'm certainly not a mind-reader, but I get the feeling the O.P. somehow wants/needs/requires/fantasizes about getting the "maximum power" (the 68W of the LM3886) out of each speaker. Of course, since they haven't told us what they're trying to do with this project, there's really no way of knowing.

 

[audioguru]

Actually, the OP wants 56V p-p to each 4 ohm speaker which is 100W that is too much for one LM3886 amp. The applications note shows two amps in parallel (with low value series resistors at their outputs) then the power into 4 ohms is about 110W.

 

[carbonzit]

But they haven't told us why they want 56V P-P to each speaker. Why would that be a requirement? Are they trying to drive an entire room full of people deaf?

 

[dr.power]

No, it does not make sense.

It sounds like you're saying you won't be satisfied until you can get the "maximum power" (i.e., the 68 watts that the 3886 is capable of providing) out of each speaker. Meaning that if you have 10 speakers, you are demanding 680 watts of power?!?!?

My speakers ORIGINALLY were connected to a LM3886 when I bought them, I guess that it means that each speaker works at its maximum of power while connected just to one LM3886.
Yes I mean That If I want to get the maximum of power out of 10 speakers all connected in parallel then I need 680W, It means I neither can e LM3886 and then make the speakers in series then in parallel to get almost 4 ohms out of the total speakers yet get the maximum of power out of speakers, Nor can use a matching transformer and connect all speakers in parallel to its output to get the maximum of power. Hope it makes sense now. It is stright forward I guess, So I can not understand why you suggest me to use the speakers in series/parallel and use just one chip!!!?

Which is why I asked what you're trying to accomplish here. It's possible you have some strange application where you simply must get the maximum possible power out of each speaker.

The application is not strange. I told here in this thread that I have decided to design a line or maybe a planar speaker using the said speakers. I think that I HAVE TO have the maximum of power out of the said speakers because I need to experience the directivity of the speakers in FAR-FIELD. Is this makes sense too?

Do you even realize how much power 68 watts is? And are you going to be running this whatever-it-is at maximum power, wide open?

Yea I think I know how much it is. P=V^2/R.
Well, The input voltage going to the amplifier naturally changes, it means that the output voltage and hence the output power changes, It might reach to more than 60W if needed. right?

Generally speaking, audio amplifiers (with some exceptions) are not expected to be run at their full rated power.

In a test I just got 56V across of my 4 ohms speaker while connected to a LM3886 but the amplifier went to saturation because the Power supply was just a bit less than 60V.

The solution I offered you--wiring your speakers in series-parallel--will work with what you have. It's up to you to explain to us why that is not an acceptable solution. I still don't get it.

If you yet can not get it yet please reread this post.
guidance, the point is to get the MAXIMUM of OUTPUT out of speakers

 

[dr.power]

As stated by others, you put the speakers in series-parallel. Thus you have two sets of two speakers in series and then parallel the two sets. That puts you back at the original impedance of 4 ohms. Now you will get the same total power at the same amp voltage as you would with one speaker. Of course, the power is now shared between 4 speakers but that's to be expected.

Perhaps you need to review ohms law of parallel and series circuits.

Well, I do not need to review ohms law.
It's strange!!! That's what I tried several times to say to those guys who wanted me to use the speakers in series-parallel, It's strange that now you want me to review ohms law!
WE whole now that by using the speakers in series/parallel Each speaker does not see the max power needed for it because it does not see the voltage of while used lonely.
Sorry, but apart from putting the speakers in parallel (of course demand more current from amplifier ) how do you guys want me to put the speakers in series yet get the maximum of power out of each while the output voltage of the chip is shared between 2 speakers?

 

[crutschow]

I believe we are having a problem in semantics. Exactly what you mean by getting the "maximum power out of each"?

If the amp is putting out it's maximum power than that is all it can do. With proper series-parallel connection of the speakers, they will have the combined equivalent impedance for which the amplifier delivers its maximum power. But of course that power will divided by the number of speakers. What more could you expect? You can't put more total power into the speakers than an amp can deliver. For more power you need to go to more amplifiers.

 

[carbonzit]

WE whole now that by using the speakers in series/parallel Each speaker does not see the max power needed for it because it does not see the voltage of while used lonely.

Well, parsing your questionable English, I can only tell you this:

OK, let's say you wanted to get "maximum possible power" out of each and every speaker, namely, the 68 watts you previously drove a single speaker with. (We'll leave aside that 1) this figure is totally arbitrary, based on the parts you've selected to use, and 2) that it's unusual to expect a speaker to be driven continuously at an amplifier's maximum rated power.)

You can do it by wiring the speakers in series/parallel.

Just have enough LM3886s connected in parallel (bridged} to provide the "maximum power" to n speakers.

If you have 16 speakers (chosen to give a total impedance of 4 ohms), use 16 LM3886s.

Now the total "maximum power" of the bank of amplifiers (1,088 watts) will be evenly distributed among the 16 speakers, resulting in your "maximum power" (68 watts) at each speaker.

Don't believe me? Draw the circuit and analyze it; each speaker will receive 1/nth of the total power.

 

[dr.power]

I don't know where you got a signal that is 56V p-p into 4 ohms because it is 100W but an LM3886 has a max output of 68W when its supply is 56VDC. An applications note for the LM3886 IC explains that its SPIKE protection circuit turns on when the output is 68W into 4 ohms which is is 16.5V RMS which is 46.6V p-p.

Well, Actually the Amplifier did not work correctly by doing so, I guess that it went to saturation by doing that job.

Simply use one LM3886 for each speaker and use a 56VDC supply or plus 28VDC and minus 28VDC. The phase of the sound from a speaker will be identical to the others (unless a speaker is connected backwards).

Honestly,as you know what I am designing is a liner speaker (directional) and uses the wavefronts of the speakers to generate a directional output sound, thats why I had a concern regarding phase difference of the amplifiers while used separately for each speaker.

.I myself prefer using whole LM3886 chips in parallel to get the lower output impedance. Please somebody explain me the pros and cons of using all amplifiers in parallel or using separate amplifiers for each speaker.

Thanks

 

[dr.power]

Carl, I'm certainly not a mind-reader, but I get the feeling the O.P. somehow wants/needs/requires/fantasizes about getting the "maximum power" (the 68W of the LM3886) out of each speaker. Of course, since they haven't told us what they're trying to do with this project, there's really no way of knowing.

By getting the maximum of output power from the speakers (and thus from the amplifier) each speaker will give the maximum of output sound which is needed for me in far-fields (please remind that my design is a liner speaker..)