who specifies the TRUE RMS ratings even some big bees dont..

no need to speak about the Rotel, krell, plinius etc.. they are just very good in term of power ratings... Like that... I dont find any of the manufacturers rate the equipments power supply...

cheers,
Sandeep

Hey !

Did you know that the Krell 400i ( 200 Watts RMS ) Integrated Amp does not mmet its power rating ?

On evaluation, the amp over heated and shut down, when being preconditioned.

The amp has relatively TINY Heat sinks, that too encased INSIDE the amp body.

Gives the amp a sleek look, but poor heat dissipation.

well Im not talking about integrated stuffs here. But If properly designed the system will be highly stable...

cheers,
Sandeep

Sorry, I kind of missed the point.

Are we talking about receivers as in the first post, or are we talking about krells and other 'high-end' audio?

RMS ratings, like most amp specs, are close to meaningless. It is only a statement of intent, not a truth cast in stone.

Again, power = voltage/load. The voltage output is defined by the input voltage, the gain and the supply voltage rails of the amp. The load (in a perfect world) defines the current required to drive it.

Speakers get damaged by basically two things (please feel free to add if you feel I've missed out something)

1. Inability to dissipate VC heat/energy. This has lots of primary causes like DC, overvoltage, clipping, etc, but the net effect is the frying of the VC.

2. Exceeding linear excursion limits consistently. this effect is normally audible before damage in LF drivers, and is sometimes heard before terminal damage. The usual cure is to reduce the power input. Bottoming out woofers even once can cause irreversible damage to the spider structure and diaphragm balance. This is usually a voltage-related effect.

High current amplifiers will not deliver more power to a load unless the load asks for it. In the case stated in the first post, 150 watts RMS is 34 volts at 4 amps into 8 ohms, not that difficult to manage unless you now start adding tough 4 ohm specs, add current loading for headroom (3db) and look at 2 ohm capability. Also remember a true 34 volt amplifier has a voltage gain of about 60 from a 600 mV signal, which is difficult to achieve with low distortion unless a lot of tricks are used, specially if it's a discrete design. Then comes heat dissipation. A 150 RMS class AB will be putting out a lot of heat, I'd say close to 100 watts - enough to light up a small village with some CFLs. It's not going to be that easy to get rid of it cheaply.

The basic principle to amp design is to estimate the acoustic output or SPL level required by the system as a whole. Then by using the efficiency figure, arrive at the electrical power required to drive the speakers (remembering to apply frequency derating and keeping it within the limits of the speaker). Then work out required voltage and current, selecting devices and topology, and last simulation and validation.

hi eric,
Ur post seems to be very matured. Anyway thanks for the information. Can I know your background in hifi?

Regarding amp Ive been studyin since the past 5 years...
yes 100watts of heat dessipation is not again simple but at the same time not that complex as well..
High current amps are very solid in construction and gives excellent tightness in bass region...and ofcourse if and only if the driver asks for such current then it will be useful... But in many situations it asks...

cheers,
Sandeep

Sonic_Master schrieb:

hi eric, Ur post seems to be very matured. Anyway thanks for the information. Can I know your background in hifi?

Oh me, I'm just a hustler who likes to read. I have no background in 'hifi', just been tinkering with electronics for the last 15 years and not just audio electronics. So don't mind me, since you surely have things under control.

However, just two small points I want to make. 100 watts of heat in a 30 degrees ambient with a heatsink c/w of 1, means your output devices will be at 130 degrees. Adding a fan should be able to drop it to about .7 c/w, and if you have massive heatsinks of .8 c/w and add a fan you can maybe get down to abut .5 c/w. That is still 80 degrees on the output devices, assuming the room temperature remains constant. I assume there is some kind of an enclosure and some of this heat will be inside the enclosure.

At that temperature coffee can burn your tongue, and 5 degrees above that is ordinary capacitor failure threshold. Also temperature derating curves usually come into effect at about 70 degrees C for most known semiconductors and passives. So hot is not very good. Normally we strive to keep temeratures below 50 degrees for all electronics. Not always possible, but a good target to aim for.

However since you will surely be running music through the amplifier and not sinewaves, there is not much of a chance of hitting peak dissipation more than 40% of the time (at a stretch, normally the figure is closer to 25%) so a bit of laxity can be shown for commercial design, when dimensioning heatsinking and power supplies.

A lot of designers take 70% as a good mark, and mass market manufacturers normally dimension their products at 50%. This is maybe why the Krell failed (My assumption) when subjected to full FTC testing (40% full power RMS sinewave + IM signal to represent music). I guess the spirit of the manufacturer is what counts here - whether they build the amp to full sine spec or music power.

If you are studying amplifiers may I recommend Douglas Self's excellent book as a reference. It's called the Power Amplifier Design Handbook and contains excellent concepts as far as amplifiers go. It also takes a lot of views on some of the bunkum surrounding 'Hi-Fi'.

There is no doubt that High current designs are the foundation for a lot of amplifiers and are a good way to counter some of the difficulties faced when driving tough speakers. However a lot of time is spent thinking about power supply design and not output stage design, as beta droop occurs at higher collector curents, and this affects output stage distortion as well.

There are lots of other considerations, but I fear this is already getting too technical for a hifi forum. Basically the point of my previous post is that music is vibration in air. We need to figure out the best way of getting the air to vibrate the same way as it would have given the original intrument/s, and it is my opinion that the most efficient way of doing so is the best. And in that context, an amp's power ratings alone mean very little. As a matter of fact someone on this board brought up the first watt concept, which Nelson Pass has exploited in his F1 design, where he states that the most important part of the power band is the quality of the first watt, since that is where we do most of our listening.

eric_clapton schrieb:

However, just two small points I want to make. 100 watts of heat in a 30 degrees ambient with a heatsink c/w of 1, means your output devices will be at 130 degrees. Adding a fan should be able to drop it to about .7 c/w, and if you have massive heatsinks of .8 c/w and add a fan you can maybe get down to abut .5 c/w. That is still 80 degrees on the output devices, assuming the room temperature remains constant. I assume there is some kind of an enclosure and some of this heat will be inside the enclosure. At that temperature coffee can burn your tongue, and 5 degrees above that is ordinary capacitor failure threshold. Also temperature derating curves usually come into effect at about 70 degrees C for most known semiconductors and passives. So hot is not very good. Normally we strive to keep temeratures below 50 degrees for all electronics. Not always possible, but a good target to aim for.

Hi Eric,
When I touch the upper surface of my 60W per ch power amp, it's temperature is never really hot - within 5 deg of the room temperature. Does this mean that the amp runs cold? Or that the heat is trapped inside the amp (meaning the electronics is much hotter than the body)?

The amp has cooling vents only on its bottom. The top surface which I touch is completely sealed. It does not have a cooling fan, just a heat sink bar.

Neutral,

In all probability, it means your amp heat dissipation characteristics maybe very good or it just does not generate too much heat to have a lot of heat sinks and cooling fans...

Secondly, your speakers may not be posing a very difficult load to the amp so it runs at a decent temp....nothing to worry about....

manek.

Manek schrieb:

Neutral, In all probability, it means your amp heat dissipation characteristics maybe very good or it just does not generate too much heat to have a lot of heat sinks and cooling fans... Secondly, your speakers may not be posing a very difficult load to the amp so it runs at a decent temp....nothing to worry about.... manek.

Thanks Manek. I wasn't worrying. (We both know the Pulz amp is decent). What I was asking is whether the external surface temperature of an amp is closely related to the internal temperature of the electronics.

The 70 deg that Eric was talking about (I assume of the output transistors) would be very noticeable if the entire body were to heat up. In a non-AC home, such heating would be very unpleasant, especially in a Bombay summer.

Do you know which parts of an amp really heat up the most when an amp is in operation: transistors, transformer, power supply caps, pots?

Neutral schrieb:

What I was asking is whether the external surface temperature of an amp is closely related to the internal temperature of the electronics.

In a well-designed amp, absolutely. There should not be more than a couple of degrees difference between the internal temperature of the output devices and the coupling point to the heatsink. One of the measures of amp construction is thermal coupling between components and heat dissipators.

Neutral schrieb:

Do you know which parts of an amp really heat up the most when an amp is in operation: transistors, transformer, power supply caps, pots?

Most of the time it is the output devices, but certain supply components can get uncomfortably warm. Certain designs require a lot of dissipation (compared to their component size) from voltage regulators, for example, which are typically the hottest components inside CD players.

Sometimes a lot of resistors can run fairly warm, and sometimes transformers. There is really no hard and fast rule, but when you're giving it the stick, then it's usually the power supply and output transistors that take the brunt of the heat.

Manek schrieb:

Secondly, your speakers may not be posing a very difficult load to the amp so it runs at a decent temp....nothing to worry about....

I would think that's mostly it. Plus you must be running at decent volume levels.

Neutral schrieb:

The 70 deg that Eric was talking about....

would be of a very powerful amp running at full tilt. Uncommon unless you're talking professional (live/stage) use, and then you're looking at bigger and better heatsinking or Class D. I suspect in home use with normal speakers by the time the heatsinks get that warm you'd be on the operating table getting your ears removed.