Power Amp Output Questions

[Tinnkindling]

I have noticed that some solid state power amps, except those with output transformers, have different power ratings depending on the speaker load impedance. For example, an amp may be rated at 100 wpc @ 8 ohms, and 150 wpc at 4 ohms. This makes sense to me, as the lower load impedance allows the amp to deliver more power.

What I want to know is why some amps have a much higher "spread" in those ratings. For example, one amp may be 100 wpc tag 8 ohms, and 150 wpc at 4 ohms. Another amp may be 100 wpc at 8 ohms, and 200 wpc at 4 ohms.

What is the difference between the two amps I just mentioned?

 

[Chris Brown]

More power into lower impedance loads implies an amp that is very good at delivering current, likely a beefier power supply.

 

[Guest105]

Ideal amp will double its output as impedance is halved. Very few do it, due to limitations in power supply (the costliest section of the amp).

 

[onemug]

What is the difference between the two amps I just mentioned?

Most likely the manufacturers manipulation of the specs.

It's a great selling point to spec an amp at 100 into 8/200 into 4. There is truth there (at least, there better be) in that the amp will put out 200 into 4, but in the "vast" majority of actual bench test specs I've seen, the maker is underrating the 8 ohm spec so it gives the illusion of that magical "doubling" phenom.

I see a lawsuit coming if/when a maker says 100/8 and it only puts out 75. Who's going to sue if they say 100 and it puts out 150?

 

[fred soop]

Even with the "perfect" power supply, the delivered power can't quite double with the load resistance cut in half, because the required current also increases by SQRT(2), resulting in increased losses in the amplifier output such as emitter resistors, possible switch or relay contact resistance, and even the internal emitter resistance of the output transistors.

There have also been amplifiers with electronic load limiting that do not allow the current to increase by SQRT (2). The Heathkit AR-15 receiver is such an example. The 8 ohm power is 50 watts per channel, dropping to 45 watts with a 4 ohm load. Meanwhile, the Harman Kardon Citation 22 and 24, (with no output transformers), have a switch on the back to select the optimum load impedance. This actually switches voltage taps on the power transformer and reconfigures the protection circuit to optimize the amplifier for 100 watts (Cit 24) or 200 watts (Cit 22) per channel into either load impedance.

 

[Guest105]

Even with the "perfect" power supply, the delivered power can't quite double with the load resistance cut in half, because the required current also increases by SQRT(2), resulting in increased losses in the amplifier output such as emitter resistors, possible switch or relay contact resistance, and even the internal emitter resistance of the output transistors.

There have also been amplifiers with electronic load limiting that do not allow the current to increase by SQRT (2). The Heathkit AR-15 receiver is such an example. The 8 ohm power is 50 watts per channel, dropping to 45 watts with a 4 ohm load. Meanwhile, the Harman Kardon Citation 22 and 24, (with no output transformers), have a switch on the back to select the optimum load impedance. This actually switches voltage taps on the power transformer and reconfigures the protection circuit to optimize the amplifier for 100 watts (Cit 24) or 200 watts (Cit 22) per channel into either load impedance.

Yes, there are all kinds of variations on design and topology....but by far the most common reason for variation in output at reduced impedance is that it is due to differences in the quality of the power supply.

 

[E-Stat]

What is the difference between the two amps I just mentioned?

While the power supply is largely to explain the difference as others have mentioned, it may also have to do with limitations of the output stage itself. Many amps have protection circuitry to limit voltage and keep devices from overheating when pressed with a challenging load.

Other amps, however, are designed from the outset to handle low impedance and/or reactive loads and obviate the need for such. They typically do "double down" (or nearly so) without causing any reliability issues.

 

[bowtie427ss]

Ideal amp will double its output as impedance is halved. Very few do it, due to limitations in power supply (the costliest section of the amp).

Please name one amplifier that doubles down with a halving of impedance, and include the independent(not manufacurer's) test results.

Claimed specs are mostly bullshit, do not use them for doubling down citation.

 

[whoaru99]

The Crown CE4000s I use on subs claim to double from 8 ohms (600wpc) to 4 ohms (1200wpc), but only 50% more at 2 ohms (1800wpc), and you have to be running at 200V or on the AC mains more to get the latter (universal SMPS power supplies). My guess is those are reasonably accurate being Crown, and a pro audio product where they don't take much shine on numbers games.

I had some power tests done a long time ago by "Bink" Knowles that included the CE4000, unfortunately I can't find it now. One part I recall for sure was the stellar low voltage performance. The amp retained effectively full rated output down to 96VAC, for all you saggy line voltage folks.

 

[JoeESP9]

bowtie427ss:
See the September 1994 Sereophile review of the Krell KSA-100. It produces 20dBW at 4 and 8 Ohms. That classifies as doubling down by any standard. If you have questions about a dBW rating do some research. The test results are by John Atkinson. Even if you don't like Stereophile you have to acknowledge that his test reports are accurate.

There are many other amplifiers that pass this test. Most of them are Class A and none of them are inexpensive even used.

 

[bowtie427ss]

Even if you don't like Stereophile you have to acknowledge that his test reports are accurate.

Have used them for my own citation in the past.

It is also one of the sources where it first became grossly obvious to me how much manufacturers "massage down" 8 and 4 ohm ratings such that the advertised specs will show doubling down all the way to 2 ohms.

Another Krell, the FPB 300 is rated to double down at 4 and 2 ohms with a claimed output of 300, 600, and 1200 watts per channel respectively into 8, 4, and 2 ohm loads. Sterophile's own test results measured the FPB 300 at over 580 watts into 8 ohms before clipping, 930 watts into 4 ohms before clipping. At the time of their test, they did not have the ability to verify 2 ohm continuous output.

My entire point is that manufacturer's ratings that double down into 4 and 2 ohm loads are 99% advertising manipulation.

 

[Guest105]

Please name one amplifier that doubles down with a halving of impedance, and include the independent(not manufacurer's) test results.

Claimed specs are mostly bullshit, do not use them for doubling down citation.

Having a bad day? No need to pass it on to me. I'm not here to provide citations for you.

Please note my answer, which was quite specific:

"Ideal amp will double its output as impedance is halved."​

If you disagree with this, fine. Cite your sources, and I will cite mine.

 

[E-Stat]

My entire point is that manufacturer's ratings that double down into 4 and 2 ohm loads are 99% advertising manipulation.

Another series of brands that comes 99% close to doubling down is those fathered by Nelson Pass - be it Threshold or Pass Labs. The very first Threshold 800A from 1975 was designed to drive the reactive load of the Dayton-Wright electrostat - a load that shuts down many a pro amp.

I have an '81 Stasis 3 purchased new that nearly doubles down. With it's 105 joule power supply and 32 - 150 watt output devices for 100 watts/channel into 8 ohms (clips at slightly over 200 at 4 ohms), it has no difficulty driving the Acoustat electrostats that I've been using ever since.

 

[whoaru99]

Right, but the point is how much is the actual 8 ohms and possibly 4 ohm output sandbagged on paper to give the appearance of doubling down?

On paper my Levinson 23 also doubles from 8 to 4 and nearly again to 2. However, by test, it puts out a fair bit more than rated power at 8 ohms. So, if it was rated straight up, it wouldn't appear to double down.

 

[E-Stat]

Right, but the point is how much is the actual 8 ohms and possibly 4 ohm output sandbagged on paper to give the appearance of doubling down?

Since you're responding to the OP and didn't quote text, I don't have a clue as to who you're responding to.

In my case, the Stasis 3 clipped in a test by Audio Magazine at 125 watts @ 8 ohms and something over 200 @ 4 ohms.

 

[JoeESP9]

You asked if there were any that had independent testing that supported the manufacturers claim. I gave you one. There are others.

While there may be a lot of exaggeration regarding amplifier outputs I seriously doubt if it's in the 99% group.

As far as the "claimed output of the FPB 300, what do you find dishonest about it? Krell says it produces 300, 600 and 1200 Watts per channel. Stereophile measuring 580 and 930 respectively says to me that they have met their claims, at least into 4 and 8 Ohms
.
Note: Krell's ratings do not say, "at clipping" they only say 300 Watts into 8 Ohms and 600 Watts into 4 Ohms. Please explain what is dishonest or misleading about that. The amplifier clearly produces that much output Wattage. If it produces more than 300Watts at 8 Ohms that's just gravy.

Unless it produces less than 300 Watts at 8 Ohms and/or doesn't double that into 4 Ohms they haven't lied or misled anyone. The amplifier produces what it's claimed to produce. What's your beef with that?

If I make an amplifier and claim that it produces 100WPC at 8 Ohms and 200WPC at 4 Ohms and it actually produces 125WPC at 8 Ohms and only 200 WPC at 4 Ohms where have I lied or misled anyone? It doubles down with the power rating I specified.

If an amplifier doubles down on the advertised and claimed output levels where and how is that misleading?

You're arbitrarily adding "at clipping" whereas the manufacturer did not do so.

Incidentally, the Krell I cited does double down at clipping. The dBW figures that John Atkinson measured support that.

 

[whoaru99]

You asked if there were any that had independent testing that supported the manufacturers claim. I gave you one. There are others.

While there may be a lot of exaggeration regarding amplifier outputs I seriously doubt if it's in the 99% group.

As far as the "claimed output of the FPB 300, what do you find dishonest about it? Krell says it produces 300, 600 and 1200 Watts per channel. Stereophile measuring 580 and 930 respectively says to me that they have met their claims, at least into 4 and 8 Ohms
.

It's dishonest insamuch as it's not the true output of the amp and intentionally misleading to give the appearance of doubling down which is an often promoted "feature" of amps. In addition, it promotes dishonest comparison to other 300 watt amps that are rated more honestly. Why do 300 watts of "high end" power kick ass on 300 watts of mid-fi power? Because the 300W of "high end" power is actually almost 600 watts (in this example).

If those figures are correct, and if that amp doubled down honestly it would be rated 580, 1160, and 2320. That's quite a departure from 300, 600, and 1200.

When this is brought to the attention of many they dismiss it because "you're getting more than you paid for". No you're not, you're getting what you paid for. They just lied to you about the amp to play on the "doubling down" mindset.

 

[Dr. Morbius]

In the car audio biz they called those "cheater" amps as they would produce a good bit more power than what they were rated.

 

[E-Stat]

It's dishonest insamuch as it's not the true output of the amp... If those figures are correct, and if that amp doubled down honestly it would be rated 580, 1160, and 2320. That's quite a departure from 300, 600, and 1200.

In the real world, speakers don't always act like test resistors.

 

[whoaru99]

In the real world, speakers don't always act like test resistors.

They do close enough. I just posted about this a day or two, maybe three ago.