A different question about power amplifier output (in the real world)

[Dood]

I think I know the answer to this already, but I'd be keen to find out which brands in particular have been tested.

I often read about the cheaper 1u power amplifiers having problems delivering their rated power output in one way or another. Either that manufacturers are telling porky pies, or that under extreme load the amplifier doesn't have sufficient cooling in order to keep it running under load. - Output sags after sometimes milliseconds or anything up to a minute of having to work hard.

So, I know that bass signal requires lots of power in the lows and I also know that just because we have a 1000W amp it absolutely isn't running at 1000W all the time (if at all) and the chances of a continuous bass signal for any more than 30 seconds is also very rare. (Unless I suppose the amps are being used for other instruments as well including bass synths playing long notes?)

My question is this - I suppose it's all about headroom. If you have an amplifier that on paper is capable of 2500W RMS in to 4 Ohms and in tests it is only capable of peaks of 2000W in the real world AND under hard load it drops down to say 1800W RMS continuous, does that then translate to an 1600W RMS amplifier that will never have any problems with delivering a big fat signal all the time?

To take it a step further - using a 3000W rated power amplifier that only ever has to deliver an absolute maximum of 1000W seems like a logical thing to do, especially as there are some 'cost conscious' models on the market. You could buy two of those and still have change for the cost of a premium unit such as PXN or Powersoft?

Edited August 1, 2015 by dood

[Roland Rock]

Real world you say?

I have a 1000w Type A power amp which easily handles the lows. Then I sell it, and with half of that money buy a cheaper Type B 3000w power amp which handles those lows just as well. Great.

My inner voice will be nagging me to buy the 3000w version of the Type A power amp, as it will be amazing.

I'm now skint with an aching back, but happy.

Edited August 1, 2015 by Roland Rock

[Dood]

[quote name='Roland Rock' timestamp='1438429482' post='2834462']
Real world you say?

I have a 1000w Type A power amp which easily handles the lows. Then I sell it, and with half of that money buy a cheaper Type B 3000w power amp which handles those lows just as well. Great.

My inner voice will be nagging me to buy the 3000w version of the Type A power amp, as it will be amazing.

I'm now skint with an aching back, but happy.
[/quote]

GAS aside! - That's always a problem. I used to use a massive all valve 300W head that sounded amazing but weighed what seemed like a ton at 3am in the morning. I'd still like to own it, but you know, my back also says no!!

[Guest bassman7755]

[quote name='dood' timestamp='1438428084' post='2834439']
To take it a step further - using a 3000W rated power amplifier that only ever has to deliver an absolute maximum of 1000W seems like a logical thing to do, especially as there are some 'cost conscious' models on the market. You could buy two of those and still have change for the cost of a premium unit such as PXN or Powersoft?
[/quote]

Well not so long ago the in-thing with bass rigs was to run a rack preamp and a massively specced PA amp running on tickover. My previous rig was an SVP into a QSC PLX 1602 running bridged and even at the loudest gig it rarely went above the first green LED on the output power indicator. Yes it was a great sound, but was it perceptibly better than the RH450 I use now ? - not that I can really notice TBH and certainly not worth carting around the extra 20kg of weight.

Edited August 1, 2015 by bassman7755

[Chienmortbb]

[quote name='Roland Rock' timestamp='1438429482' post='2834462']
Real world you say?

I have a 1000w Type A power amp which easily handles the lows. Then I sell it, and with half of that money buy a cheaper Type B 3000w power amp which handles those lows just as well. Great.

My inner voice will be nagging me to buy the 3000w version of the Type A power amp, as it will be amazing.

I'm now skint with an aching back, but happy.
[/quote] So as a Class A amplifier is around 10% efficient, you need 10000 watts or around 40 Amps of current at 230V. Three times what you can draw fro a single socket ( unless you are running three phase) it is August not April isn't it!

[Roland Rock]

[quote name='Chienmortbb' timestamp='1438456196' post='2834674']
So as a Class A amplifier is around 10% efficient, you need 10000 watts or around 40 Amps of current at 230V. Three times what you can draw fro a single socket ( unless you are running three phase) it is August not April isn't it!
[/quote]

Is your sarcastic gybe based on the fact that you think I was talking about Class A and Class B amps? I wasn't - I was referring to two different hypothetical types of amp labeled Type A and Type B.

Edited August 1, 2015 by Roland Rock

[mazdah]

Good Lord, I'm using one 150W amp and one 30w to gif regularly, art/prog rock and experimental. Hefty enough.

[Phil Starr]

[quote name='dood' timestamp='1438428084' post='2834439']
I think I know the answer to this already, but I'd be keen to find out which brands in particular have been tested.

I often read about the cheaper 1u power amplifiers having problems delivering their rated power output in one way or another. Either that manufacturers are telling porky pies, or that under extreme load the amplifier doesn't have sufficient cooling in order to keep it running under load. - Output sags after sometimes milliseconds or anything up to a minute of having to work hard.

So, I know that bass signal requires lots of power in the lows and I also know that just because we have a 1000W amp it absolutely isn't running at 1000W all the time (if at all) and the chances of a continuous bass signal for any more than 30 seconds is also very rare. (Unless I suppose the amps are being used for other instruments as well including bass synths playing long notes?)

My question is this - I suppose it's all about headroom. If you have an amplifier that on paper is capable of 2500W RMS in to 4 Ohms and in tests it is only capable of peaks of 2000W in the real world AND under hard load it drops down to say 1800W RMS continuous, does that then translate to an 1600W RMS amplifier that will never have any problems with delivering a big fat signal all the time?

To take it a step further - using a 3000W rated power amplifier that only ever has to deliver an absolute maximum of 1000W seems like a logical thing to do, especially as there are some 'cost conscious' models on the market. You could buy two of those and still have change for the cost of a premium unit such as PXN or Powersoft?
[/quote]
There are essentially three limits to an amplifiers power output. The voltage it can swing, the current it can provide and the heat it can dissipate. In any given amp one of these will be the important limiting element. (I'm ignoring any protection circuits or DSP here)

There are lots of ways to measure and calculate an amps output even if you stick to rms. they range from just measuring the voltage swing and calculating the power that voltage would drive through 8ohms to connecting the amp to a dummy resistor and running sine waves through it for a week and measuring the temperature rise. Different amps work differently so can scre better in some tests than others. It's relatively easy to switch huge voltages so class D amps score well here. Since in most applications the power is only needed for a fraction of a second it makes no sense to put in a huge power supply so this is often the limit. With transistor amps high voltage devices are more expensive and the transistors are prone to thermal runaway so heat is the problem even if the power supply is big enough. Valve amps were usually limited by current, not least because all the power would go out through the output transformer.

So, a watt is still a watt but there is a lot of room for confusion and even when standard tests and rms watts are used there will be amps that seem louder than others.

The other aspect of this though is the speaker. There is no single drive unit that can handle 1000W continuously at all frequencies, and if you use 1000W through multiple drivers then the sound levels developed are going to be so high as to be unusable on stage. These sorts of powers are only useful for PA arrays safely suspended way above the band and audiences heads.

To answer one of your questions. A 3000W rated amp limited by heat and power supply to1600W continuous could be treated as a 1600W amp except it could also put 3000W through your speakers for long enough to damage them. In fact real world speakers rated at, say, 800W rms 1600W peak probably can't handle 200W at some frequencies.

We are are all going to have to get used to class D amp ratings. Affordable power is now effectively unlimited, there are a few problems with power supplies but we are going to have to decide to be careful about what powers we wish for.

[Chienmortbb]

[quote name='Roland Rock' timestamp='1438459070' post='2834708']
Is your sarcastic gybe based on the fact that you think I was talking about Class A and Class B amps? I wasn't - I was referring to two different hypothetical types of amp labeled Type A and Type B.
[/quote] Yes. Sorry I misread you post. Blame it on my bad eye.

[Roland Rock]

[quote name='Chienmortbb' timestamp='1438679195' post='2836261']
Yes. Sorry I misread you post. Blame it on my bad eye.
[/quote]

No worries amigo

[Dan Dare]

People worry too much about quoted power outputs. They're pretty meaningless - usually arrived at over a limited range of frequencies into a stable load. Real world conditions aren't like that - a complex harmonic structure to the signal into a reactive, variable load (have a look at the impedance curve of a typical drive unit). The preamp with bridged PA power amp route is pretty effective, but you need enough speaker capacity - you won't shift enough air with a 2x10, no matter how much power you put into it. Suggest buying a trolley to cope with the weight.

[Chienmortbb]

[quote name='Phil Starr' timestamp='1438501882' post='2834879']
There are essentially three limits to an amplifiers power output. The voltage it can swing, the current it can provide and the heat it can dissipate. In any given amp one of these will be the important limiting element. (I'm ignoring any protection circuits or DSP here)

There are lots of ways to measure and calculate an amps output even if you stick to rms. they range from just measuring the voltage swing and calculating the power that voltage would drive through 8ohms to connecting the amp to a dummy resistor and running sine waves through it for a week and measuring the temperature rise. Different amps work differently so can scre better in some tests than others. It's relatively easy to switch huge voltages so class D amps score well here. Since in most applications the power is only needed for a fraction of a second it makes no sense to put in a huge power supply so this is often the limit. With transistor amps high voltage devices are more expensive and the transistors are prone to thermal runaway so heat is the problem even if the power supply is big enough. Valve amps were usually limited by current, not least because all the power would go out through the output transformer.

So, a watt is still a watt but there is a lot of room for confusion and even when standard tests and rms watts are used there will be amps that seem louder than others.

The other aspect of this though is the speaker. There is no single drive unit that can handle 1000W continuously at all frequencies, and if you use 1000W through multiple drivers then the sound levels developed are going to be so high as to be unusable on stage. These sorts of powers are only useful for PA arrays safely suspended way above the band and audiences heads.

To answer one of your questions. A 3000W rated amp limited by heat and power supply to1600W continuous could be treated as a 1600W amp except it could also put 3000W through your speakers for long enough to damage them. In fact real world speakers rated at, say, 800W rms 1600W peak probably can't handle 200W at some frequencies.

We are are all going to have to get used to class D amp ratings. Affordable power is now effectively unlimited, there are a few problems with power supplies but we are going to have to decide to be careful about what powers we wish for.
[/quote]Very good points Phil. To add my two pence worth,

Valves or Bipolar or Mosfet amps can all be made in Class A, B, AB or D (Class C are radio/tv transmitter power amps and E,G and H are all variations on Class B IMHO).

Firstly why do valve amps sound louder. Well to some extent it is a limitation of Valves that they need a transformer to drive low impedance speakers. It is easy to Tap a transformer to match the power output of the amp to the loudspeakers. So a 100W valve amp will give 100W into 4,8 and 16 Ohms of the trasformer is tapped at those impedances.
An amplifier without a transformer will suppy different powers into different loads (speakers). A perfect Solid State amplifier might give 50W into 16 Ohms, 100W intp 8 Ohms and 200W into 4 Ohms (It might even give 400W into 2 ohms. No marketing department would dream of calling this a 50W amp. It would be called a 200W amp.

So do a test with the 100W Valve V the 200W SS into 8 Ohms and guess what. They sound the same UNTIL the amps are pushed hard the soft clipping or self compression of a valve ourput stage kicks in. The valve amp will sound louder. If the 200W amp has a good compressor the differences will be subtle.


Coming on to the power supply this is a big limiting factor. The biggest single cost component in an SS amp, Whether Class AB or Class D is either the transformer or Switch Mode Power Supply. So an amp that can output 250W into ohms may be limited to a much lower level into 4 Ohms. Take a very good amp like the Hartke HA3500, an amp you know well. It is specified as 240W into 8 Ohms and 350W into 4 Ohms. Almost certaily the 4 Ohm figure is power supply limited.

Marketing departments are also guilty of using bogus ratings. A number of makers use the Anaview [b]Class D ALC0300, [/b]one manufacturer has used it for amps with a claimed outputs of 640, 420 and 320 watts. It is rated as 200W into 8 ohms and a recent change uprated the 4ohm output from 320 to 350 watts. So clearly 640 is twice 320, and 420 is the peak output specified by Anaview.

All these are specifies at 1% distortin figures but some makers quote at 10% distortion that can give a figure almost 40% higher but would be unuseable in real world use.

The giggest problem for me is that hardly anyone gives a true technical specification for their amps. so it really makes it hard to compare amp to amp even from the same make.

Many amps are quoted at their maximum output assumimg that heatsinking or fans are at the optimum level. Then use pathetic heat management that means the amp cannot possible reach its quoted output for more than a few seconds.

There are lies. damn lies and specifications.

So how do we compare amps? I would suggest always compare at 8 Ohms. So halve the output of a Solid State Class A/B or Class D Amps when comparing with a Valve amp. After that listen but remember that although the ear has no memory and is influenced by the eye. A big SVT looks louder than a small Class D amp but is it? How much is down to the "voice" of the amps being compared. Don"t compare a sound with the one you remember, it will be innacurate. The only valid comparisons are blind A/B test that happen at the same time and place.

[Handwired]

The type of cab will play a massive part in how loud the amp sounds.

[Guest bassman7755]

[quote name='Dan Dare' timestamp='1438691575' post='2836406']
... but you need enough speaker capacity - you won't shift enough air with a 2x10, no matter how much power you put into it.
[/quote]

Drivers vary greatly in their displacement capability for a given size. A top grade 2x10 with high displacement drivers could easily out perform a budget 4x10 although it would be somewhat less efficient.