Should a Good Amp Double Its Power? (1 Viewer)

[Yogi]

Yessss!!! Numbers on paper rarely tell how the amp will sound to your ears. Nothing beats hearing with your own ears in your own room with your own equipment.

 

[Robert Cowan]

hehe, speaking of which, im gonna turn off the computer and put on a cd

 

[Kevin C Brown]

I'd be interested in seeing some actual test results from a reasonably priced amp that doubles down.

 

[Yogi]

I know this is not a reasonably priced amp by a long shot, but here are the actual results of a levinson 436 amp which is think is quite a bit overbuilt:


The amp is rated 350/700/1400 @ 8/4/2 ohms but in reality it does 500/890/1500 @ 8/4/2 so as you can see the manufacturer overstates the high impedance ratings considerably so that the amp would seem like it doubles its output for every halving of impedence.

Again there cannot be an amp that would truly behave as a voltage source, IMHO.

 

[Robert Cowan]

when i finally get mine built ill throw it on my test bench and see how it works hopefully it will do well.

 

[Kevin C Brown]

Yogi- I really like ML stuff (shoot, you know I have the Proceed setup ), but ... If you look at those results as percentage increases in power output, the McCormack results I have in post 5 actually beat it. But see that's the problem (sort of). I have never seen any amp even within 2x of the McCormack's price range do anywhere close to what it does. (The McCormacks actually get really well reviewed for sound quality too.) That's why when people say that "quality amps should double down", in fact, most don't do anything close to that. But IMO, that is a secondary effect on sound quality anyway. (I.e., only if *not* doubling down limits the dynamics for low impedance or hard to drive loads.) I am *guessing* that that's Chu's point.

Robert- I'd certainly be interested...

 

[Yogi]

Here are the DNA225 results from the same article in Stereophile:



So you can see the Mccormack 225 doesnt even double into half load let alone exceed it like the Levinson. The measurements you are talking about are tonebursts at 1KHz and will always be substantially higher than continuous RMS output like I was talking about.

I am not saying that the Levinson would sound better than the Mcormack its just that the ML stuff is overbuilt to the hilt. Fully balanced from input to output meaning basically two identical amplifiers with opposite phase per channel. 2.4 KVA powersupply/channel. In fact in one of stereophiles reviews it said that the amplifier couldn't be tested into continuous 1ohm load as the wall voltage sagged so much making the measurement invalid. Now thats some serious draw and even then it can't really double its output for half impedance. But then those amplifiers require taking a second mortgage on one's home:frowning:

BTW I am a big ML fan myself inspite of being a tube lover. I have Proceed Amp2 powering the two fronts of my HT setup which will soon be replaced by a Proceed HPA2 which IMHO is ML sound at a budget price. Infact the new 4XX series amps are just the HPA amps rebadged as ML because the same design and manufacturing people that built the Proceed HPA gear made them too close to the Levinson amps and so their sales were being cannibalized by Proceed and hence the discontinuation of the Proceed series, but now I am digressing.

 

[Kevin C Brown]

Yogi- Yes, the results I quoted were from a slightly different test, but you got the gist.

I guess my point was that it's easy to say that quality high performance amps should double down, and if you look at a lot of amp's specs, you'd believe that's true, but in reality not a lot of amps come even close.

And actually here's a question. Let's say I design an amp that does 200W into 8, into 4, into 2, and into 1 ohm. Let's say I don't have the output designed as a perfect voltage source, but simply a constant power source. So now I use this amp with an 8 ohm set of speakers and compare it to a set of 4 ohm speakers (same efficiency). The amp will play louder with the 8 ohm speaker, right? So it is important to increase power to some degree into lower impedance loads, right?


Sidenote: Proceed ... Mark Levinson ... I looked at the 431, 432, 433. I was kind of thinking, might be an upgrade path for someone who likes Proceed so much. :b But the "cheapest" one, the 431 goes for $6000 list. Nice to dream, anyway! I did get confirmation at least that the 433 will be 200 x 3. (Not on the web site yet.) So kind of like the HPA 3. And it's kind of interesting too, that all of the HPA & BPA series and the AMP 5 are all the same basic design (from the AMP 2 & 3), just differing numbers of channels, and power capabilities. So it's kind of neat (to me) to realize that Madrigal Labs had the same basic design in production for almost 10 years ('94 to '03). And now with the 431/2/3, that same design, obviously evolving somewhat, will continue.

 

[Michael R Price]

It's impossible to actually double the power into half the load, because of the internal resistances and voltage drops in the amplifier, which are usually constant (or increase at higher currents). I did the calculations for the amplifier circuits I am using, which use a standard Class AB Mosfet follower run off +/-75V, and even with a perfect power supply, the amp could deliver something like 270W into 8 ohms and 520W into 4 ohms. Not quite a doubling of power. Then there are a lot of other losses to deal with, which I didn't take into account. So, IMO, it's a pointless goal, and any amp that can show large increases of power into lower impedance loads is good enough.

 

[Chu Gai]

I-T4000 1250W @ 8 ohms, 2000W @ 4 ohms, 1800W @ 2 ohms. 20-20kHz less than 0.35% THD, 3 year no fault, transferable warranty. No double down though but seems like it ought to handle most loads. Five grand though...ouch and no idea how noisy it is. But it is within 2x of the McCormack price though which is really a nice f*cking amp IMO. Both ought to drive those little baby ventriloquists just fine, don't you think?

I just realized...duhhh, Crown is now a Harmon company so I wonder if the I-mode will make its appearance in other products that they carry.

 

[John Kotches]

Yogi,

Thank you for making the point I would have made myself. Many amps "double down" via specsmanship, ie understating output at 8 ohms (and maybe even 4 ohms) just to get it to double down "on paper".

As an example, Bryston could make their amplifiers double down overnight by simpling changing the 8 ohm rating to 250watts/channel

Cheers,

 

[Yogi]

You said it in one phrase, what I was taking paragraphs to say

 

[Shiu]

That's close enough to call it "double down". Pretty hard to do better than that!

 

[Robert Cowan]

something for everyone to consider...

many of these places benchmark amps until they clip. that is, they test it separately at 8/4/2 ohms until the amp clips. each is seen as an indepentant test...

lets say the amp clips at 240 watts into 8 ohms, 420 into 4 ohms, and 812 into 2 ohms. this amp isnt doubling-down is it? well, yes. if this amp is spec'd at 200/ch, it is. it can do 200/400/800... but it clips a bit beyond those specs.

the test would be, can you send the exact same signal into an amp, measure it with 8/4/2 ohm loads, and does it double then... not measure where it clips at each impedance.

 

[Chu Gai]

What prevents fudging in that case? I mean, you could probably take a bottom of the line Sony or even a KLH for that matter and choose a nice benign 8 ohm power of say 5 or 10 watts, no?

 

[Yogi]

I think thats what Chu is asking, if an amp can really double its output to clipping since thats the only reference cutoff point. If there is no hard cutoff you could say every amp doubles its power depending on how you spec it.

But we won't know what exactly Chu means by his question since he is not talking:frowning:.

 

[Kevin C Brown]

Sorry Robert. That isn't "doubling down" to me either.

The whole point to doubling down, is the amp acting as a perfect voltage source. Specs have nothing to do with it. Actual performance via real measurements is what matters.

 

[Robert Cowan]

after i wrote that, i thought a bit more about it, and i guess i have to add that the amp should at least double at its rated frequency...

so, its its a 200w/ch amp, it should do at least 200/400/800 before clipping.

BUT, i do understand what you guys are saying about being a perfect amp and doulbing perfect. ill have to think about that. but, in that case, no amp will be perfect, there will be a few losses here and there. and ultimately, every amp has its optimum impedance.

 

[Chu Gai]

And so long as the amp can deliver what a speaker calls for...right?

 

[Robert Cowan]

i think that would be fair...

if the speaker isnt demanding more power than the amp can supply, and it delivers stated power, doubling from 8 - 2 ohms, sure...

regarding resistance and losses. i think there is a certain amount of consistancy with this. the amp will have a certian internal inefficiency in all resistances. meaning, at 8 ohms, the amp might have an internal resistance which will make it actually produce x amount less than its theoretical. this might be the same loss at all loads. in any event, it could be seen as a constant, thus ruled out.