[jinu_joseph]

Hi Folks

Is there any way to calculate the efficeincy of an amplifier based on the theoretical specification values published.

I believe Class AB effiecieny can vary from 30% to 50% whereas Class D efficiency can go upto 75%, but is there any way to calculate this without acutally having to actually measure the input and output voltage/current.

Thanks

[vina]

Originally Posted by jinu_joseph

Hi Folks

Is there any way to calculate the efficeincy of an amplifier based on the theoretical specification values published.

I believe Class AB effiecieny can vary from 30% to 50% whereas Class D efficiency can go upto 75%, but is there any way to calculate this without acutally having to actually measure the input and output voltage/current.

Thanks

Class AB efficiency can be as high as 90% if you are driving high-only or low-only. For random signals with zero mean also this can be so if, for example, you are driving out a square wave.

Class D efficiency can be really bad if you are driving a very low amplitude signal.


That said - I have never seen any amplifier specify efficiency. If they do specify efficiency they will have to do so for driving a standardised waveform (usually a sinusoid) for a given amplitude. Best is to measure it directly.


That said - efficiency is a practically useless concern for non-techies (peak volume, distortion and heat generation are important), why are you worried about it?

[jinu_joseph]

i need the efficiency to compute the power draw from the battery/alternator

[vina]

Originally Posted by jinu_joseph

i need the efficiency to compute the power draw from the battery/alternator

Best is to measure it. For example some equipment may have class-D for bass and class-AB for treble.

Also power draw will depend on exactly what you are playing (not only on the volume) - there is something called the "crest factor" of sound (peak power to average power ratio) for most modern music it is between 9dB and 6dB, but for classical music it can be higher (20dB is very easily possible).

My advice - look at the PMPO rating and at least double it (in any case your electrical systems will have to work under peak load, not just the average load) preferably tripple it. If you know which ones are class AB and which ones are class D then again you can look at PMPO of each, and multiply the AB PMPO by 2.5 while the D's PMPO by 1.5 and add them together - this will take care of all classes of amplifiers along with their supply's losses.

[DerAlte]

Originally Posted by vina

... look at the PMPO rating and at least double it (in any case your electrical systems will have to work under peak load, not just the average load) preferably tripple it. ...

@vina-bhai, thoda discount please! At this rate we will be running insulated copper bus-bars to the amp.

To be very practical (your advice not to worry about efficiency was very good), since the manufacturer would have looked at current consumptions at 9.6, 12 and 14.4V, one can rely on a simple thing - the fuse rating - to know what is the expected max current. Even moderate overloads are taken care of by the fast blow fuse used (design time decision). So, rather than worrying about PMPO and RMS and efficiency of the Class of amp, one can directly work out the power expected to be drawn (during the loudest music passage) on the DC side rather than the AC side.

I have a feeling (I may be completely off) that @jinu_joseph's question had to go with going 'green'! I use an Alpine Class D amp (all channels Class D) because I wanted to keep the load light. To my old ears, the compromise on sound quality is undetectable, and I'd rather not waste power.

PS: You are right, manufacturers neither measure nor declare efficiency of the amp. Some of them do publish the output damping factor, but that is a different story altogether. Amp efficiency is measured in labs on a Class of amp basic, not model of amp basic. And yes, the input signal used in the lab is a sinusoid (not square wave) since the class of application is predominantly that. This measures steady state efficiency at different frequencies, not behaviour with transients.

[vina]

Originally Posted by DerAlte

@vina-bhai, thoda discount please! At this rate we will be running insulated copper bus-bars to the amp.

I was wondering where you were - this is your speciality area.

My advice was purely from the point of view of load on his electrical system - wiring is lesser concern as PMPO is rare.

I did some calculations - for a 100W PMPO off a 12V battery we get min 10A current for the PMPO and 20A to be absolutely sure. Normal household wire will be more than enough to tackle this but the resistance may be too much

Originally Posted by DerAlte

To be very practical (your advice not to worry about efficiency was very good), since the manufacturer would have looked at current consumptions at 9.6, 12 and 14.4V, one can rely on a simple thing - the fuse rating - to know what is the expected max current. Even moderate overloads are taken care of by the fast blow fuse used (design time decision). So, rather than worrying about PMPO and RMS and efficiency of the Class of amp, one can directly work out the power expected to be drawn (during the loudest music passage) on the DC side rather than the AC side.

Do music systems come with their own fuses? If so then that is the best estimator.

If not then the load will have to be found from max sound output power x 2 (or more to account for losses). Also this may not be a very good way of measuring - nothing beats a cheap multimeter for this kind of a thing.

Originally Posted by DerAlte

I have a feeling (I may be completely off) that @jinu_joseph's question had to go with going 'green'! I use an Alpine Class D amp (all channels Class D) because I wanted to keep the load light. To my old ears, the compromise on sound quality is undetectable, and I'd rather not waste power.

Class-D amplifiers these days (if you buy the expensive ones) are so good - they produce waveforms purer than CD recordings. Unless you are trying to play back a recording done at a professional studio recorded on a professional grade digital storage with no compression whatsoever (in which case good luck - no car stereo system comes close anyway) even a 10 year old's year can not detect the difference - a very sensitive oscilloscope can not detect the difference (CD has more noise than the class-D)

By the way, class-D equipment (not chips) should be cheaper - there is less thermal to worry about and less money to spend on power management chips - but I guess they sell it more expensive because "digital amplifier"

Originally Posted by DerAlte

PS: You are right, manufacturers neither measure nor declare efficiency of the amp. Some of them do publish the output damping factor, but that is a different story altogether. Amp efficiency is measured in labs on a Class of amp basic, not model of amp basic. And yes, the input signal used in the lab is a sinusoid (not square wave) since the class of application is predominantly that. This measures steady state efficiency at different frequencies, not behaviour with transients.

This is true for most equipment with one (very significant) exception - mobile phones (and tablets these days). Those guys test the heck out of it (though they still don't publish the electrical results). Samsung, Nokia and LG for example have standardised music they play and measure the power numbers.

these are used to determine the battery life - that they do advertise proudly - but that includes power of a lot of other components (decoders, RTC ...) for it to be useful to find out amp-only power.

[jinu_joseph]

The aim of all this discussion is to find out if the current Alternator that i have in the car is capable of driving my ICE unit without acutally draining my battery, inspite of the car running. In Mumbai i am stuck in a lot of bumper to bumper traffic which would mean that the alternator is not running at peak efficiecy/output, so in effect the alternator rated at 85 Amps might actually be producing only about 50Amps or even less in such traffic conditions.

This would mean that any draw over 50amps would actually be draining the battery inspite of the car being in running. Of these 50 Amps available around 15-20 amps goes for regular car electricals (lights, AC, etc). Considering that we are in good old rainy season, the wiper motor will draw another 10 amps (i guess) So that would mean I am left with about 20A for my ICE.

So if i have a setup like a focal solid4 for front and rear and Focal Solid 1 for sub. they both have 2 fuses of 25A each. So that a total max current draw of 100A. Even if i am using only half that power that is still a draw of 50A. That would mean that I would be drawing 30A from the battery (without it getting recharged) and considering that the battery is rated at 55Ah i wouldnt last more than a couple of hours on the road.

Is something wrong with my calculation or is this a correct potrayl of power requirement.

[navin]

Music is not a continous signal. Music has peaks and valleys and hence you will not have a problem. A senior Mod on TBHP (GTO) had a similar problem with his Civic which was resolved by replacing his amps with amps that can operate at even low battery voltages (sub 11V).

Since his morning drive is not at rush hour (no pun intended ) his battery gets a good charge in the morning. Similarly if you give your car a good long every other day you should be fine.

I have a couple of amps in my car too (similar draw to your Focal 1/4 combo) but since my system is used only 2 times a month or so (I dont drive often) the battery is well charged. In fact I often find that my in dash ipod has discharged (due to lack of use) and have to use my rear ipod (in the arm rest).

[vina]

Originally Posted by jinu_joseph

The aim of all this discussion is to find out if the current Alternator that i have in the car is capable of driving my ICE unit without acutally draining my battery, inspite of the car running. In Mumbai i am stuck in a lot of bumper to bumper traffic which would mean that the alternator is not running at peak efficiecy/output, so in effect the alternator rated at 85 Amps might actually be producing only about 50Amps or even less in such traffic conditions.

This would mean that any draw over 50amps would actually be draining the battery inspite of the car being in running. Of these 50 Amps available around 15-20 amps goes for regular car electricals (lights, AC, etc). Considering that we are in good old rainy season, the wiper motor will draw another 10 amps (i guess) So that would mean I am left with about 20A for my ICE.

So if i have a setup like a focal solid4 for front and rear and Focal Solid 1 for sub. they both have 2 fuses of 25A each. So that a total max current draw of 100A. Even if i am using only half that power that is still a draw of 50A. That would mean that I would be drawing 30A from the battery (without it getting recharged) and considering that the battery is rated at 55Ah i wouldnt last more than a couple of hours on the road.

Is something wrong with my calculation or is this a correct potrayl of power requirement.

If the fuses are rated at 100A total - your system CAN draw roughly 80A max - and usually will not draw more than 20A even if you are playing it very loud with class-D amplifiers.

For medium volumes the current will drop to 10A.

[jinu_joseph]

Ok now i am totally confused.. I read something about speaker sensitivity and things got more complicated. heres why.
My understanding of sensitivity is that it takes only 1 Watt of input power to produce an SPL of 91Db measured at a distance of 1 meter from the speaker.
1. Lets consider a typical setup this time only consider the speakers.
Front - Focal 165KR - rated at 80W RMS with a sensitvity rating of 91Db.
Rear - Focal 165KR - rated at 80W RMS with a sensitvity rating of 91Db.
Sub - Focal 33V2 - rated at 400W RMS/800W peak and 90db Sensitivity
Assumptions
- Ideal SPL that the human ear can sustain over long periods (85DB). This is the SPL that the listener should get. For our calculation lets take it to be 90Db.
- Every 3DB increase in SPL requires the power(watts) to be doubled
- For every doubling of distance from the speaker there is drop of SPL by 3db
- Average distance between Listener and front/rear speakers 2mtrs
- Average distance between Listener and Sub 4mtrs

1. For the Front and rear speakers to be able to produce 90db at a distance of 2 meters from the speaker which is the same as 93db at a distance of 1m we will require about 2W of input power. Because sensitivity of the speaker is 91db. Hence all 4 channels combined will require only about 8w of power from the Amp.
5. Now for Subwoofers. To produce 90db at a distance of 4 meters from the speaker which is the same as 96db at a distance of 1m we will require about 4W of input power. ( I find this result very suprising considering that it is the norm that subwoofer require a lot of power)

This would that my entire setup would require only 12W of output power from the amp. Even if i have to assume some losses on the way and double the Amp output power required its only 24W

So is there any substance in this matter.... or is this a case of partial information being fatal.

[DerAlte]

Originally Posted by jinu_joseph

... This would that my entire setup would require only 12W of output power from the amp. Even if i have to assume some losses on the way and double the Amp output power required its only 24W

So is there any substance in this matter.... or is this a case of partial information being fatal.

Your assumptions are within the ball-park, even though your calculation method is, well, unusual!
1. One normally sits within 1.33m from any speaker in a car
2. If the subwoofer is 4m away, it is actually in the car in front or the car following you! 4m is ~ 12'6"!!!
3. Looks like all your habits are within medically prescribed limits. Sometimes music is not music at 85dB
4. Numbers published by most speaker manufacturers are for guidance only, and it would be very difficult to reproduce those figures practically. Especially under non-ideal circumstances in a car
5. Yes, you have recognized it correctly, as you said "a case of partial information being fatal". If you really want a detailed calculation down to the first digit after decimal, you will have to account for another dozen or so factors

Recommended: "Don't worry, be happy" and enjoy your music in moderation without worrying about your battery failing on you.

[navin]

Originally Posted by jinu_joseph

Ok now i am totally confused..

This would that my entire setup would require only 12W of output power from the amp. Even if i have to assume some losses on the way and double the Amp output power required its only 24W

a. Speakers are not linear. They may produce 90db/1w/1 and even 96db/4w/1m but once the excursion of the woofer starts getting stretched it will compress the output.

b. Music is not a continous signal but a series of peaks and valleys where the peaks might be 20db higher the valleys. So your amp will be running at 1W and then be asked to produce 100W (20db) for a few milliseconds in extreme cases (modern electronic music and some orchestral scores like the 1812).

c. cabin gain and absorbtion ensures that the frequency response of a speaker in a cabin is far from linear.

[vina]

You are doing too much calculation friend. Don't worry this much - it is very hard to make your battery go dead by playing music while the engine is running (i.e. at least idling).

Originally Posted by jinu_joseph

Ok now i am totally confused.. I read something about speaker sensitivity and things got more complicated. heres why.
My understanding of sensitivity is that it takes only 1 Watt of input power to produce an SPL of 91Db measured at a distance of 1 meter from the speaker.

Can you tell me the source of this information? I'm asking because I have been looking for electrical-sound conversion metrics for some time without success.


Keep in mind that there are efficiency issues (frequency dependent - very low frequencies and very high frequencies both give poor speaker efficiency; smaller speakers = poor speakers in general as far as efficiency and linearity goes ...) that can increase the numbers for you.

Originally Posted by jinu_joseph

1. Lets consider a typical setup this time only consider the speakers.
Front - Focal 165KR - rated at 80W RMS with a sensitvity rating of 91Db.
Rear - Focal 165KR - rated at 80W RMS with a sensitvity rating of 91Db.
Sub - Focal 33V2 - rated at 400W RMS/800W peak and 90db Sensitivity
Assumptions
- Ideal SPL that the human ear can sustain over long periods (85DB). This is the SPL that the listener should get. For our calculation lets take it to be 90Db.
- Every 3DB increase in SPL requires the power(watts) to be doubled
- For every doubling of distance from the speaker there is drop of SPL by 3db
- Average distance between Listener and front/rear speakers 2mtrs
- Average distance between Listener and Sub 4mtrs

1. For the Front and rear speakers to be able to produce 90db at a distance of 2 meters from the speaker which is the same as 93db at a distance of 1m we will require about 2W of input power. Because sensitivity of the speaker is 91db. Hence all 4 channels combined will require only about 8w of power from the Amp.
5. Now for Subwoofers. To produce 90db at a distance of 4 meters from the speaker which is the same as 96db at a distance of 1m we will require about 4W of input power. ( I find this result very suprising considering that it is the norm that subwoofer require a lot of power)

This would that my entire setup would require only 12W of output power from the amp. Even if i have to assume some losses on the way and double the Amp output power required its only 24W

So is there any substance in this matter.... or is this a case of partial information being fatal.

the last line is correct.

You are assuming an overall efficiency of 50% (12 becomes 24) but speakers lose energy AND amplifiers lose energy AND modulators, equaliser, filter lose energy AND LCD displays etc. lose energy ....

My suggestion is if you really want to get down to it, just take a multimeter and measure it out. You should be able to lay your hands on a handheld one relatively easily.

Otherwise don't worry too much - all the numbers mentioned are for very high volumes (humans can sustain 90dB is not the same as they'll enjoy it).

To put thing in perspective (just to help you) - Get a cheap to mid-range Samsung phone, put the ring volume to its highest close your car and put it in place where you want to put your speakers.

Now play it out loud - it'll most likely be not loud enough, but it'll give you the idea of how much louder you want.

The Samsung phones output about 1W rms and consume about 2W from the battery in doing so - (Don't ask me how I know - I just know). For highly distorted audio they can output max 2W (if connected to charger, but don't count on it).

Now if you want double the volume, you'll need 4W, and so on.

[navin]

Originally Posted by vina

You are assuming an overall efficiency of 50% (12 becomes 24) but speakers lose energy....

Good call...taken from Loudspeaker - Wikipedia, the free encyclopedia

"Typical home loudspeakers have sensitivities of about 85 to 95 dB for 1 W @ 1 m—an efficiency of 0.5–4%."

[jinu_joseph]

Originally Posted by vina

Can you tell me the source of this information? I'm asking because I have been looking for electrical-sound conversion metrics for some time without success.

Here are the areas form where I gathered information.
Sound pressure - Wikipedia, the free encyclopedia
Damping of sound level with distance - decibel dB damping calculation calculator distance versus sound reduction free field loss - decrease drop fall in sound over distance versus dB sound at different distances microphone calculator distance drop ra
Relationship Between Watts and dBs — Reviews and News from Audioholics
http://www.integratedaudio.com/help/sensitivity.pdf
Peak SPL Calculator

Folks
On the whole i am finding it difficult to accept that there is does not seems to be a defined mathematical process to derive the input power required to drive a certain set of speakers at different SPL levels at different distances from the listener (even if it was assuming a pure sine wave as input, because that would always be on the higher side)