Noob question - High current amplifiers

[anandpkumar]

18 minutes ago, pwstereo said:

Possibly because it would limit sales for those with very low impedance dips, etc.

If you knew up front that the great sounding new speakers you heard at the shop were going to destroy your amp or at least render it not worth using, you might look at other speakers that present less difficulty to the amplification.

 

Yeah, I too guessed that. Same reasons that some amplifier manufacturers leave out the rated output at 4 ohm or 2 ohm 

 

Edited March 16, 2021 by anandpkumar

[Ittaku]

9 minutes ago, ghost4man said:

Con,

Where do you include watts per channel in this equation because typically that's what amplifiers are measured against when listing specs?

I get what you are saying about the intimate connection between voltage and current but it seems listening to Dave that current is everything. Apologies to Dave if I have misinterpreted the thrust of his statement.

Invariably very little if anything in the specs for a Sanders emphasises how much current it can deliver. Its exclusively geared towards the wpc into specific impedance levels.

BTW great thread guys. Have learned a lot!

 

Watts per channel is quoted as voltage squared divided by resistance. They never quote based on current, on the assumption that there's always "at least enough current" for the quoted impedance the amplifier will work with. In the case of Sanders, there's approximately 10x more than enough current on the assumption 4 ohm speakers are in places effectively far lower than 4 ohms (Quoted 900 watts at 4 ohms only requires 15A yet the white paper says they can deliver 135A as discussed earlier.)

Edited March 16, 2021 by Ittaku

[davewantsmoore]

2 hours ago, pwstereo said:

Don't we say the load (speaker) draws the current from the source (amplifier)?

Yes .... but it's a bit confusing to put it that way if you break down exactly what is happening and why.

[RoHo]

3 hours ago, davewantsmoore said:

It's complicated.   Single numbers, or maximums, etc.... don't tell you what you need to know.... and people will interpret it all wrong.

 

 

... but we can see (hear) from the fact that sometimes amplifiers sound different to each other when they shouldn't......that we are abusing them by tortuous speaker designs.     Misbehaving amplifiers really sound (at best) terrible .... so it's a sad thing.

We've talked previously about the need to quote the SPL used for a speaker's frequency response measurement in order to be able to meaningfully compare it to that of another speaker.  So now I realise it is the same for an impedance measurement!  So you really need a family of curves for both to properly describe a speakers behaviour - which I have never, ever seen published in the consumer arena  

So if the data isn't available is it actually possible to make an "informed choice"?  

[davewantsmoore]

2 hours ago, pwstereo said:

A potential difference exists between two terminals, a connected load draws a current from the terminals.

Think of the (voltage) potential difference as a force.... so the amp is doing the pushing, rather than the speaker doing the pulling.

[pwstereo]

5 minutes ago, davewantsmoore said:

Think of the (voltage) potential difference as a force.... so the amp is doing the pushing, rather than the speaker doing the pulling.

OK, so I'm being a bit silly now, but if my 240V power outlet is switched on with nothing plugged into it, the power station isn't pushing electricity out into the room. But if I plug in my fan heater it is drawing almost 10 amps on high.

[davewantsmoore]

1 hour ago, Steffen said:

However, the signal is a voltage level, not a current.

 

It can be a current... if you design an amplifier to do that.

 

Then you need to (re)design the speaker to have the desired frequency response, as they will not be at all the same under "current drive" vs "voltage drive"

 

1 hour ago, Steffen said:

 

We feed the speaker a voltage and measure the sound level it produces. We expect that relationship to be reasonably flat across the frequency range of the speaker. That’s what makes a quality speaker.

 

But as mentioned.... You can make a speaker which has any response you want under voltage or current drive.

 

In fact contrary to what you said.... if you take a speaker driver and hook it up to a "voltage source"amp and current source amp.....  you will often find the current source amp has a flatter response.     The inductive rise counting the HF rolloff  ..... and the impedance spike at the system resonance, countering the rolloff at LF.....  but that's really a moot point, as you can make a speaker have any response you want, if you design the system so.

 

If it's a commercial speaker dsignd for a voltage source..... then sure, you need to use that.

 

[davewantsmoore]

1 hour ago, ghost4man said:

But listening to @davewantsmoore during this conversation its really been impressed upon me how important current is in this. Dave I wonder why speaker manufacturers dont go to the trouble of identifying the sweet spot for the current required to get the best out of its speakers. I say this because its often the case that audiophiles will choose an amplifier on the basis of wpc capabilities at particular loads. 

 

Because it varies wildly with how loud you play.    Manufacturers don't want to scare away their customers by making things complicated or telling them they "need a lot".

 

... but ask them directly and all good ones will help.

[davewantsmoore]

1 hour ago, Ittaku said:

You can't make a current amplifier without it being a voltage amplifier as well. The voltage determines how loud the driver will play provided enough current is being delivered for the impedance. Likewise current determines how loud the driver will play provided enough voltage is being delivered for the impedance.

Yes, of course.... but I'm not sure what your point is.

 

1 hour ago, Ittaku said:

Saying one is more important than the other is missing the point

 

... but one IS more important than the other.    In the speaker, it is ultimately only the resulting current flow which matters.

 

By encoding the music as a voltage.... we leave this current flow to the mercy of distortion (and bad type of distortion at that).

 

[Ittaku]

1 minute ago, davewantsmoore said:

By encoding the music as a voltage.... we leave this current flow to the mercy of distortion (and bad type of distortion at that).

Tell me exactly how one is supposed to encode music as a current?

[Steffen]

9 minutes ago, davewantsmoore said:

In fact contrary to what you said.... if you take a speaker driver and hook it up to a "voltage source"amp and current source amp.....  you will often find the current source amp has a flatter response.

 

 

I believe the OP was referring to commercial (i.e. “regular”) speakers and amps.

 

If you made an amp that had a linear response from voltage input to current output, flat across the frequency range, and you hooked that up to a regular speaker, you’d get a frequency response plot (input voltage to sound pressure level) that would look similar to the speaker’s impedance curve. I.e. not pretty

 

[pwstereo]

The ability of the amplifier to deliver the required output voltage into different impedance loads is where current is important.

Say a power amp can deliver a max of 70 V peak to peak, into 8 ohms (resistive load) that will need to be able to deliver 8.75 A.

If the load drops to 4 ohms, the current required at 70 V will now be 17.5 A.

 

An amplifier that has a lot of capacity to deliver current will manage better on loads that drop to low impedance.

[Guest Eggcup the Dafter]

16 minutes ago, Ittaku said:

Tell me exactly how one is supposed to encode music as a current?

Amplifier topologies for current-drive | Current-Drive - The Natural Way of Loudspeaker Operation (current-drive.info)

I have no idea if this site or page make sense. I just know it exists.

[davewantsmoore]

27 minutes ago, pwstereo said:

OK, so I'm being a bit silly now, but if my 240V power outlet is switched on with nothing plugged into it, the power station isn't pushing electricity out into the room. But if I plug in my fan heater it is drawing almost 10 amps on high.

 

Correct.... because there is nothing connecting the power terminal to allow the power to flow through.

 

As much current will flow as the wall socket can push..... part of this is the ability for it to hold up the voltage..... part of this is the impedance of whatever the current is flowing through.

 

 

The question coming back to a speaker is what do you encode the music signal into.   The voltage.... or the current.    You can do either..... but when you make the voltage represent the music signal..... then you allow the resulting current to be distorted by the speaker.

[Steffen]

4 minutes ago, Eggcup the Dafter said:

Amplifier topologies for current-drive | Current-Drive - The Natural Way of Loudspeaker Operation (current-drive.info)

I have no idea if this site or page make sense. I just know it exists.

 

There is nothing magical about this, it is perfectly doable. However, it will not work well unless used with a speaker that has been design to have a flat current-to-sound-pressure response (rather than a flat voltage-to-sound-pressure response, like most speakers). There is also nothing “natural” about it, unless you can argue (and demonstrate) that it is easier to make a loudspeaker with a flat response to current, than one with a flat response to voltage.

[davewantsmoore]

47 minutes ago, Ittaku said:

Tell me exactly how one is supposed to encode music as a current?

 

https://en.wikipedia.org/wiki/Operational_transconductance_amplifier

 

... or more generally (I realise the above link is "opamp")

https://en.wikipedia.org/wiki/Transconductance#Amplifiers

Edited March 16, 2021 by davewantsmoore

[aussievintage]

43 minutes ago, davewantsmoore said:

Think of the (voltage) potential difference as a force.... so the amp is doing the pushing, rather than the speaker doing the pulling.

 

Yep.  That's why the call it electromotive force.   It's what makes the electrons to move down the wire.

 

37 minutes ago, pwstereo said:

OK, so I'm being a bit silly now, but if my 240V power outlet is switched on with nothing plugged into it, the power station isn't pushing electricity out into the room.

 

That's why it's other name is a potential. It has the potential to do work.  With nothing plugged in however, you are asking a force to move an infinite obstacle.  i.e. the impedance is infinite.  Hence nothing moves and no electricity flows.

 

In a fluid analogy, you have put a cork in the end of the pipe. 

 

Edited March 16, 2021 by aussievintage

[davewantsmoore]

20 minutes ago, Steffen said:

I believe the OP was referring to commercial (i.e. “regular”) speakers and amps.

 

If you made an amp that had a linear response from voltage input to current output, flat across the frequency range, and you hooked that up to a regular speaker, you’d get a frequency response plot (input voltage to sound pressure level) that would look similar to the speaker’s impedance curve. I.e. not pretty

 

Yes, very much so ..... why do (some) tube amps sound different?!  

It's definitely digressed from the original topic.

[Steffen]

3 minutes ago, davewantsmoore said:

 

The question coming back to a speaker is what do you encode the music signal into.   The voltage.... or the current.    You can do either..... but when you make the voltage represent the music signal..... then you allow the resulting current to be distorted by the speaker.

 

Why would voltage drive distort but not current drive? As you said earlier, one source of distortion in speaker drivers is that the impedance depends on the deflection of the voice coil (among other things of a mechanical and pneumatic nature). Thsi would affect voltage and current drive, as far as the power output is concerned:

 

P = U^2 / Z

P= I^2 * Z

 

Z is in both, and when variable/non-linear, will mess with the sound output either way.

 

[davewantsmoore]

5 minutes ago, Steffen said:

There is also nothing “natural” about it

I think you've missed the point.

 

 

5 minutes ago, Steffen said:

unless you can argue (and demonstrate) that it is easier to make a loudspeaker with a flat response to current, than one with a flat response to voltage.

 

Why have you linked "natural" and "easy"  ?!

 

If it's the "easiest" way (what does that even mean, hehe) .... then it's "natural"?

 

The speaker moves because of current flow.    You can choose the current which is going to flow..... or you can choose the voltage, and let whatever current flow.    Which one sounds more natural?!

 

 

As I said to you before..... plug a speaker driver into an amplifier(s) and vary the voltage/current drive..... or as you said, look at the response to a voltage source and convolve it with the impedance curve.   It is often the case that current drive will product a flatter response.

 

.... but as speaker designers have often chosen their boxes and filters for voltage...... then the choice is often made. 

[davewantsmoore]

5 minutes ago, Steffen said:

Why would voltage drive distort but not current drive?

 

Yes. Because the impedance isn't stable (see link to chart posted before)

 

5 minutes ago, Steffen said:

Thsi would affect voltage and current drive, as far as the power output is concerned:

 

P = U^2 / Z

P= I^2 * Z

 

Z is in both, and when variable/non-linear, will mess with the sound output either way.

 

 

No.   We are not interested in P.

 

 

When X amps flows out of the amplifier .... those same X amps flows through the VC.....  it is not affected by Z.

 

HOW MANY amps want to flow from the amplifier will be affected by Z .... as you already noted.....  This is where you use feedback..... to ensure that when the music signal wants the amp to output X amps...... that the amps actually DOES irrespective of Z.

 

In a voltage amplifier .... the feedback is on the voltage (music) signal.   The voltage will always be what you want it to be (thanks to feedback) ..... but that leave the current up to "chance".

[davewantsmoore]

1 hour ago, RoHo said:

So if the data isn't available is it actually possible to make an "informed choice"?  

Follow the advice of your speaker manufacturer.

[Steffen]

9 minutes ago, davewantsmoore said:

 

Why have you linked "natural" and "easy"  ?!

 

If it's the "easiest" way (what does that even mean, hehe) .... then it's "natural"?

 

“Natural” was a word used in the linked article, I question it’s appropriateness in this context.

 

By “easier” I meant the ability to either make a better sounding speaker at the same cost, or a same sounding speaker at a lower cost.

 

12 minutes ago, davewantsmoore said:

The speaker moves because of current flow.    You can choose the current which is going to flow..... or you can choose the voltage, and let whatever current flow.    Which one sounds more natural?!

 

 

I think that’s a naive misrepresentation of physical reality. Voltage and current at a load are intrinsically linked. One isn’t better or more natural than the other.

 

4 minutes ago, davewantsmoore said:

No.   We are not interested in P.

 

 

On the contrary, P is all we’re interested in. P is (minus some thermal losses) the work that the membrane performs, pushing air around. It is (minus losses) the sound output of the speaker.

 

[Ittaku]

1 minute ago, davewantsmoore said:

Follow the advice of your speaker manufacturer.

On the surface this would appear to be good advice, but in my experience, the vast majority of speaker manufacturers "hide" just how hard their speakers are to drive properly. I would say "take your speaker manufacturer's advice with a grain of salt" - look at the nominal impedance of the speakers, and find an amplifier that drives half that impedance comfortably without its power actually falling.

 

26 minutes ago, davewantsmoore said:

Yes, very much so ..... why do (some) tube amps sound different?!  

It's definitely digressed from the original topic.

On the contrary, I think this is very relevant to the topic as tube amps are the epitome of "low current" amplifiers, or at least more accurately "minimal current necessary" amplifiers.

[pwstereo]

21 minutes ago, davewantsmoore said:

The speaker moves because of current flow.    You can choose the current which is going to flow..... or you can choose the voltage, and let whatever current flow.    Which one sounds more natural?!

This is a good point and makes what you've been saying make a lot more sense.