Amplifiers perform audibly differently under different speaker loads?

[Phelonious Ponk]

This is a common theme around here, and one we have very recently touched on in the epic Objectivist/Subjectivist thread. I'm not really even doubting that this happens, really, I'm just hoping, expecting, that it is much less prevalent acceptable than I think I'm hearing. It has, several times, been used as an answer to how speakers can measure the same and sound dramatically different. I added the "dramatically" myself, because by the time you listen to one amp, take it out of the system, hook up the other amp, calibrate the volume to match the first amp and listen again, any difference you're going to be certain of hearing is, by necessity, going to be a serious one. Given that, and the assumptions that a) we're talking about well-designed solid state amps that are not deliberately colored in FR, and b) we're talking about well-designed speakers that are not wildly esoteric fringe products with very limited compatibility, a few questions:

Aren't quality amplifiers specifically engineered to avoid this problem?

Aren't many (most?) high-end amplifiers, in fact, over-engineered so they will be able to drive the widely variable loads of most high-end speakers?

If a variation in speaker load audibly changes an amplifier's output, would that change in the amp's performance not be measurable?

And if this condition is as common as we seem to think it is, shouldn't amp manufacturers recommend compatible speakers; shouldn't speaker manufacturers recommend compatible amps? Or at least warn us off of bad matches?

Wouldn't a BIG selling point for amp makers be an amp that is able to drive any reasonable load without veering off course, without compromise in performance? Isn't it?

And just to be sure I understand, you guys are saying that, if you switch from, say, Magicos to Revels, the audible output of your Bryston or Krell or even your Emotiva is going to change? And you find this acceptable?

Tim

 

[RBFC]

And just to be sure I understand, you guys are saying that, if you switch from, say, Magicos to Revels, the audible output of your Bryston or Krell or even your Emotiva is going to change? And you find this acceptable?

Tim

I feel a less-loaded question would be: Is it a reality that limitations in electronic components and design cause amplifiers to sound different when presented with varying loads?

Lee

 

[Phelonious Ponk]

I feel a less-loaded question would be: Is it a reality that limitations in electronic components and design cause amplifiers to sound different when presented with varying loads?

Lee

I'd opine that your version is as loaded as mine. Maybe more. It assumes that this is a given, that the issue, which I'm sure is real, has not been, cannot be, addressed by good design and implementation and, therefore, just simply be accepted as reality, even in the best solid state amps. Maybe that's the case. I'm not sure amp manufacturers would agree. And how do you even get to this conclusion? When you make the switch from Magicos to Revels, of course it's going to sound different. How do you know the amp's contribution to that difference?

Tim

 

[mep]

How about some amplifers perform audibly different with difficult impedances and phase angles? It's that power supply Tim.

 

[A.wayne]

Yes mostly PSU / output drive limitation from too weak an PSU and/or exceeding outputs SOA, causing thermal drift and increased distortion ..


Horse, course syndrome ..


Regards,

 

[bretdago]

This is a common theme around here, and one we have very recently touched on in the epic Objectivist/Subjectivist thread. I'm not really even doubting that this happens, really, I'm just hoping, expecting, that it is much less prevalent acceptable than I think I'm hearing. It has, several times, been used as an answer to how speakers can measure the same and sound dramatically different. I added the "dramatically" myself, because by the time you listen to one amp, take it out of the system, hook up the other amp, calibrate the volume to match the first amp and listen again, any difference you're going to be certain of hearing is, by necessity, going to be a serious one. Given that, and the assumptions that a) we're talking about well-designed solid state amps that are not deliberately colored in FR, and b) we're talking about well-designed speakers that are not wildly esoteric fringe products with very limited compatibility, a few questions:

Aren't quality amplifiers specifically engineered to avoid this problem?

Aren't many (most?) high-end amplifiers, in fact, over-engineered so they will be able to drive the widely variable loads of most high-end speakers?

If a variation in speaker load audibly changes an amplifier's output, would that change in the amp's performance not be measurable?

And if this condition is as common as we seem to think it is, shouldn't amp manufacturers recommend compatible speakers; shouldn't speaker manufacturers recommend compatible amps? Or at least warn us off of bad matches?

Wouldn't a BIG selling point for amp makers be an amp that is able to drive any reasonable load without veering off course, without compromise in performance? Isn't it?

And just to be sure I understand, you guys are saying that, if you switch from, say, Magicos to Revels, the audible output of your Bryston or Krell or even your Emotiva is going to change? And you find this acceptable?

Tim

Tim, I think you have called out at least 6 of what should be the ten commandments in regards to amplifier expectations. At least for a product you would want to be sold to the greatest amount of potential customers. What is interesting barring a few, is speaker manufacturers taking it easier on amplifiers at the expense of overall performance. I still believe the best speakers have traditionally been a little harder to drive. In defense to some of these amp manufacturers, price is a huge consideration in amp design and I believe if you are buying a 28K loudspeaker you should have an appropriate amplifier. You can't expect a 4k amp to have the necessary supply, output satge, and heatsink/cooling to meet the demands of a high end full range speaker. I would hope products in the 15-40k range could meet that speakers demands however and satisfy your criteria above.

 

[jkeny]

Tim,
I believe this is easy to answer:
Look at the measured output of a range of amplifers into simulated loads in Stereophile.
Then use the measured variations in impedance across the frequency range that any speaker represents.
Now tell me if there won't be measureable variations in output spectrum beteen different amplifier/speaker pairs?
Is it audibly different? Well when you have mapped the spectrum variations in output you might have a better handle on this, I guess?

 

[Robh3606]

Aren't many (most?) high-end amplifiers, in fact, over-engineered so they will be able to drive the widely variable loads of most high-end speakers?

I kind of doubt it. In the Pro world yes. How many "audiophile amps" can safely drive a 2 ohm load?? How many are even rated to drive a 2 ohm load??

I use a Crown XTI 2002 on my subs. It can deliver 1000 watts into 2 ohms and it's a mid power amp there are many that can easilly double that.

Rob

 

[rbbert]

I think there is little question that 2 amps driving speakers without passive Xovers have a better chance of sounding alike than two amps driving the typical very complex passive Xovers found in most good speaker systems today. TAS' recent roundtable of amp designers suggested (to me, at least) that keeping their amplifiers' output stable (i.e. unchanging except for amplitude) into the myriad of different speakers on the market is one of the major challenges facing a high-end amplifier designer/manufacturer.

 

[Phelonious Ponk]

I kind of doubt it. In the Pro world yes. How many "audiophile amps" can safely drive a 2 ohm load?? How many are even rated to drive a 2 ohm load??

I use a Crown XTI 2002 on my subs. It can deliver 1000 watts into 2 ohms and it's a mid power amp there are many that can easilly double that.

Rob

I'm sure I'm missing something here, but let's use Crown's current XLS 2500 as an example:

1200 watts per channel, stereo, at 2 ohms.
775 " " 4 ohms.
440 " " 8 ohms.

Is it that you can't make an amp that behaves well under widely varying loads, or is it just that most high-end amp manufacturers don't? Is it "high-end?" No. It's class D, weighs 11 lbs and costs about $600, street. Definitely doesn't qualify for "high-end." Is it "high fidelity?" That's a very different question in my view, but let's assume it's not. Let's assume it's midfi. If Crown can pull this off, in midfi, for $600, high end can't do it for $6,000? $16,000? The people paying the most have to except the most inconsistent performance?

But like I said, I'm sure I'm missing something.

Tim

 

[Phelonious Ponk]

Tim, I think you have called out at least 6 of what should be the ten commandments in regards to amplifier expectations.

I assure you it was dumb luck. What are the other 3?

I would hope products in the 15-40k range could meet that speakers demands however and satisfy your criteria above.

I would hope so too, but that doesn't seem to be what many audiophiles believe. They seem to believe that even very high-end amplifiers behave very differently, audibly very differently, when faced with different loads.

Tim

 

[rbbert]

Behaving differently yet delivering rated power (at a randomly chosen frequency like 1 kHz) into a static 2,4,8,16 ohms (or whatever) are two very different things.

 

[bretdago]

I assure you it was dumb luck. What are the other 3?



I would hope so too, but that doesn't seem to be what many audiophiles believe. They seem to believe that even very high-end amplifiers behave very differently, audibly very differently, when faced with different loads.

Tim

I assure you it was dumb luck. What are the other 3?



I would hope so too, but that doesn't seem to be what many audiophiles believe. They seem to believe that even very high-end amplifiers behave very differently, audibly very differently, when faced with different loads.

Tim

Tim,

I think the other 3-4 could be left up to debate... I admire you being outspoken about this topic. I am proud to say my amps double below 2 ohms maybe even 1 if I had a big enough variac to test. My math says I should be able to double to 1 and still be in the soa of the output devices and stable. This was a fundemental spec in all of my fathers designs at Krell and I see no reason at the moment not to continue the tradition. I do believe there are certain types of circuits that sound really good, but have a lot of trouble below 4 ohms and maybe not well suited to every loudspeaker. I do think some of it has to do with size and profit margin however. The difference in cost/size/weight in just the power supply between a 800va transformer based vs a 1600 or 2600 is pretty steep, but certainly well within the budget of some of these mega buck products.

 

[Raffles]

Wouldn't a BIG selling point for amp makers be an amp that is able to drive any reasonable load without veering off course, without compromise in performance? Isn't it?

Quad has always made a variation on this claim ("unconditionally stable into any load") but at the same time I understand that some early models of the Quad 405 protected themselves very early and so the rated power output fell in low impedance loads. At least Quad have always been open about the specs, though, providing diagrams in the manual that show the maximum output rating against load impedance and duration. I believe the current models have much more intelligent protection and are very good into any load. Quad also tend to be fairly conservative about frequency response, limiting it to not much wider than the maximum possible human hearing range, which apparently helps keeping to the stability claim. They have been known for reliability and used in recording studios, the BBC etc. However, I think that this has given Quad a 'professional', even utilitarian, image that the hi fi fraternity have steered clear of. Very reliable, but not 'musical'. (Complete rubbish of course!)

 

[Phelonious Ponk]

Behaving differently yet delivering rated power (at a randomly chosen frequency like 1 kHz) into a static 2,4,8,16 ohms (or whatever) are two very different things.

What's behaving differently, then?

Tim

 

[rbbert]

What's behaving differently, then?

Tim

The actual frequency response and distortion spectrum at least.

 

[amirm]

I'm sure I'm missing something here, but let's use Crown's current XLS 2500 as an example:

1200 watts per channel, stereo, at 2 ohms.
775 " " 4 ohms.
440 " " 8 ohms.

Is it that you can't make an amp that behaves well under widely varying loads, or is it just that most high-end amp manufacturers don't? Is it "high-end?" No. It's class D, weighs 11 lbs and costs about $600, street. Definitely doesn't qualify for "high-end." Is it "high fidelity?" That's a very different question in my view, but let's assume it's not. Let's assume it's midfi. If Crown can pull this off, in midfi, for $600, high end can't do it for $6,000? $16,000? The people paying the most have to except the most inconsistent performance?

But like I said, I'm sure I'm missing something.

Tim

One note: you can't compare pricing for commercial products for consumers. Commercial products are spec'ed by a customer and or their designer/engineer. The dealer then just supplies it with little cost of sales. As a result, the dealer margins are small. Consumer sales is very different. Customer comes in, expects a showroom/listening room. Spends hours there sometimes. Wants a salesman to talk to it. Often wants to take the gear home to listen to. If he doesn't like it, he expects to be able to return it. All of this means that there are much higher margins. The higher margins pushes up the MSRP. The higher MSRP in turn depresses the volume. With much lower volume, the price has to go even higher to make sense for the manufacturer.

To give you an example, there is a commercial video processor that is sold to cinemas. It sells for $6K or a bit less. Demand was generated for it for consumer market. The manufacturer took out the fan to make it quite and made a consumer version which is otherwise identical. The price now? $16K!

So it is hard to talk about retail pricing here. It does not reflect true apples vs. apples comparison.

 

[Phelonious Ponk]

One note: you can't compare pricing for commercial products for consumers. Commercial products are spec'ed by a customer and or their designer/engineer. The dealer then just supplies it with little cost of sales. As a result, the dealer margins are small. Consumer sales is very different. Customer comes in, expects a showroom/listening room. Spends hours there sometimes. Wants a salesman to talk to it. Often wants to take the gear home to listen to. If he doesn't like it, he expects to be able to return it. All of this means that there are much higher margins. The higher margins pushes up the MSRP. The higher MSRP in turn depresses the volume. With much lower volume, the price has to go even higher to make sense for the manufacturer.

To give you an example, there is a commercial video processor that is sold to cinemas. It sells for $6K or a bit less. Demand was generated for it for consumer market. The manufacturer took out the fan to make it quite and made a consumer version which is otherwise identical. The price now? $16K!

So it is hard to talk about retail pricing here. It does not reflect true apples vs. apples comparison.

Fair enough. I allowed for an extra margin of up to $15,400. I thought that might be enough. Does the consumer version of the video processor overheat?

Tim

 

[DonH50]

I wasted the afternoon fiddling around on what I thought was a ten-minute task... Enjoy!

Example Amplifier Responses into Speaker Loads

Plese do not bring this debate there; keep it here. I only presented something techie for consideration.

 

[treitz3]

Thanks for doing that, Don. To answer your question on that link? Sometimes. Not with all amplifiers I have heard but it most assuredly can be audible IME. A perfect example would be the Carver Research Lightstar amplifier with both the Carver Original Amazings compared to The Platinum Series Amazings. Clearly audible.

Tom