or Connect
AVS › AVS Forum › Audio › Audio theory, Setup and Chat › Watts and Current
New Posts  All Forums:Forum Nav:

# Watts and Current - Page 6

Quote:
Originally Posted by KralNoj

Nice job on the chart.
No problem. Nice discussion back there. What you presented was clear, my questions were to the other guy, but thanks for the elaborations.

I had a heck of a time for a while getting that chart to work right. Nothing integrated properly. Eventually I realized I forgot the 4/pi.

Being gluten free, I claim that as the reason I forgot the pie...

jn

### AVS Top Picks

Ouch, off to bed without dessert!
Still think pi is a joke. Should have been defined as the equivalent of what is now 2pi. One revolution, one pi, equations everywhere become simpler.
Quote:
Originally Posted by Bigus

Still think pi is a joke. Should have been defined as the equivalent of what is now 2pi. One revolution, one pi, equations everywhere become simpler.

you can just use tau as 2 pi. One tau is one revolution
Quote:
Originally Posted by jneutron

Ohms law....no wonder...

As I said, your numbers do not appear consistent. Rather than bash me, please go back and read your post, it is not internally self consistent. You have mixed arbitrarily peak voltages and rms voltages, so it is not possible to follow the argument you presented. That is why I asked about the 300 watt number, your writing lacked the content necessary to understand what it was you were trying to say.

It is a trivial thing to make an amp clip hard one Vce sat below the rail voltage, nowadays it's typically an RDS on value as it were.. It sometimes takes finess to make it soft. I learned that back in the early 80's when I made some power amps, and IR was only starting to make lots of hexfets, few of them P channel.

edit: It was actually more difficult forcing the output transistors out of saturation. At high slew rates, this can cause cross conduction rail to rail, goodbye outputs..

jn

You are reading lots of stuff into my words that just simply isn't there. Not having said either peak or rms, there's nothing to mix. Analytically, either peak or rms power works exactly the same, for a sine wave, and they have a fixed mathematical relationship (IIRC somewhere areound RMS=0.7peak. Ohm's law is what it is whatever emoticon you care to put after it.
It's rather silly to talk about square wave RMS voltage when you do not know if you have anything close to a square wave. I suspect that in sound reproduction, the 36 percent (or greater) THD necessary to achieve square waves is unlistenable, and thus never approached. But I haven't experimented with it.

It seems pretty clear that the clipping light on an amp conveys no real meaning. What input level trips it? what would the outptu be for that input level, in volts or watts or percent distortion? All utterly unknown, all very likely different at least from manufacturer to manufacturer, if not from amp design to amp design even within a single manufacturer's line. It certainly does not necessarily, or I expect, very likely, mean you are approaching square wave clipping.

The point is that when a modern well designed transistor amp is"clipping" at one percent distortion it is NOWHERE NEAR sitting on the voltage rails. At least if its power supply is designed robustly. We know receivers don't have robust power supplies, and many guitar amps don't either because some players like what power supply sag does. Put simply, clipping isn't square wave generation. There's no question that as THD increases, the odd harmonics increase. There's not real question that that phenomenon is capable, in some circumstances, of damaging a tweeter. But IMO there's just no way a person should believe that once an amp is above clipping (defined as one percent distortion) the amp can't make any more power, or that beginning just after the one percent THD point on the ower versus distortion durve the distortion-generated harmonics are even a drop in the power bucket in any frequency range.

At one percent distortion, by definition, the distortion is 40 dB below the signal. So the distortion products, from a power perspective, are one ten thousandth of the total power. Not of much importance. At 10 percent distortion, by definition, distortion products are 20 dB below the signal, so one one hundredth of total power (actually a little less becuase total power has become 101 percent of the pure signal power, but the difference isn't important. I'm struggling to see how, if I've pushed my 100 watt amp to the point it's making 200 watts at 20 percent distortion, the 2 watts (out of 200) attributable to the distortion products are the whole reason that the amp might blow a tweeter.

Using a real example, the Bryston 9BSST has been tested to be just over 150 watts at one percent distortion. http://www.hometheater.com/content/bryston-sp3-surround-processor-and-9b-sstsup2-amplifier-ht-labs-measures. In hits 10 percent distortion at about 180 watts, it looks like to me. At ten percent distortion, the distortion products, at 20 dB below the original signal, constitute 1.8 of those watts.
Edited by JHAz - 7/4/13 at 6:54am
Here is an interesting little article I'd forgotten about. http://mixonline.com/mag/audio_sound_vacuum_tubes_2/. Just looking at the pictures is a hoot. They're testing power tubes in guitar amps. What's fascinating is that the first pic, at 12 percent distortion LOOKS worse to me than the next one, at 17 percent distortion. I didn't reread the entire article but I think the tests are all single ended (one tube) class A.

Really just for fun, for those interested. AFAIK, solid state devices aren't going to distort a sine wave differently on the positive half of the wave than on the negative half, as these tubes all do.

The thin I've never been able to find is scope traces of familiar real world guitar amps being cranked. Would be fun to know what the THD of a typical Les Paul into a typical Marshall JTM45 is, and see the shape of a sine wave distorted at the same input level. Maybe it's out there. Maybe I'll stumble into it eventually. Surely the folks at Kemper )supposedly amazing profiling amps) know . . . but maybe they see it as a trade secret. Of course, you'd have "distortion" from the tone control esttings, plus distortion in the preamp tubes, and the phase splitter, as well as the output tubes, to account for to perfectly recreate one setting on any specific guitar amp . . .
.
Incidentally, the resistor-capacitor load example was suggested by Otala and Huttinen in their paper “Peak Current Requirement of Commercial Loudspeaker Systems”, that appeared in the June, 1987 issue of the Journal of the Audio Engineering Society. I know a good idea when I steal it:)
Oops, sorry I forgot to include the quote. this is in respons to Heinrich S' thread starter
Hi guys,

Would like to know from those more experienced, what the nitty gritty is with current and wattage. I've been told that some amplifiers have lots of wattage but not much current, others have low wattage and high current. What is important in an amplifier : High wattage or high current? Thanks.
Quote:
Originally Posted by JHAz

You are reading lots of stuff into my words that just simply isn't there. Not having said either peak or rms, there's nothing to mix.
That is why I asked you for clarification. The discussion included peak voltage, as in supply rails, and rms voltages, as in specified power delivered to a load. You tossed about such a mix of numbers and provided new ones with no supporting text, that it was not possible to determine what your statements meant. Your written verbage lacked coherence.

Instead of answering, you alluded to my needing to learn ohms law.. A knee jerk reaction showing you in poor light.
Quote:
Originally Posted by JHAz

Analytically, either peak or rms power works exactly the same, for a sine wave, and they have a fixed mathematical relationship (IIRC somewhere areound RMS=0.7peak.
But peak or rms voltage is of concern when discussing power delivered to a load, and clipping. Which was why I asked.

Did you just say .7 IIRC???? Your not an engineer, are you. The number is .7071, it is 1 over the square root of 2, which is 1.414.

I know of no engineer who does not have those numbers burned forever into the brain...
Quote:
Originally Posted by JHAz

Ohm's law is what it is whatever emoticon you care to put after it.

I was actually looking for an emoticon which expressed this: ""really, your kidding, right? I've been a practicing EE for 40 years now, so perhaps you should leave the knee jerks on the cutting room floor.""...emoticon. Couldn't find one, so settled for that.

Clearly, you missed the point..
Quote:
Originally Posted by JHAz

It's rather silly to talk about square wave RMS voltage when you do not know if you have anything close to a square wave. I suspect that in sound reproduction, the 36 percent (or greater) THD necessary to achieve square waves is unlistenable, and thus never approached. But I haven't experimented with it.

Not silly at all. What it does is point out the upper bounds of the harmonic energy that can be delivered to the system simply by clipping of a specific amplifier with a specific rail voltage. For example, total annihilation of the signal into a square will produce NO MORE than 38% of the rated energy of the amp as harmonics. a 100 watt rated amp driven hard will produce 38 watts maximum harmonic energy.

If the frequency of the drive signal is high enough, all of that 38% can hit the tweeter. If not, it's easy to calculate what part will hit the tweeter.

As I said, it is the upper boundary of the energy the tweeter could receive. By using that number, it is easier to design a system where the tweeter is more likely to survive. In the example of my two way system, I upped the tweeter crossover to 5Khz third order to help prevent too much harmonic energy from hitting the tweeter. For my personal use, it works very well. When I loan the system out, all bets are off.

Ah, by the way. With a woofer crossed over single order at 3K, and a tweet 3rd order at 5K, there is a designed in hole in the response. When I setup and play, I will mod the 11 band graphic to produce the desired sound. Given the typical spl I work at, and the length of time I play or audience participates, I've found over the years that the hole is actually a blessing. The 1K to 5K content tends to contribute heavily to listener fatigue. By suppressing it a tad, the music can be listened to considerably longer. Again, this is dance music, not hardcore rock and roll. There, the lead guitar and vocals play a heavier role..

Oh, I have indeed experimented with it.
Quote:
Originally Posted by JHAz

It seems pretty clear that the clipping light on an amp conveys no real meaning. What input level trips it? what would the outptu be for that input level, in volts or watts or percent distortion? All utterly unknown, all very likely different at least from manufacturer to manufacturer, if not from amp design to amp design even within a single manufacturer's line. It certainly does not necessarily, or I expect, very likely, mean you are approaching square wave clipping.

In the QSC rmx 1450, my amps, the clipping light is driven by circuitry which compares the signal drive with the output result. As such, it illuminates when the amplifier is no longer faithfully reproducing the input waveform. It is also independent of rail voltages, so if the supply caps are small, the clip circuitry will work independent of that, as well as during line droop or brownouts.

Quote:
Originally Posted by JHAz

The point is that when a modern well designed transistor amp is"clipping" at one percent distortion it is NOWHERE NEAR sitting on the voltage rails. At least if its power supply is designed robustly.
The robustness of the supply does not change the system as you state. Output stage drive and performance doesn't quite work the way you are alluding to. And jumping back and forth between tube and SS amp types also is of no use.

jn
Edited by jneutron - 7/8/13 at 7:25am
Jneutron, you seem to know a lot more about this topic than Arnyk. This is why following these kind of discussions is difficult because so-called experts chime in and present statements which are called "myth" which are later busted by someone who has an even greater understanding of the relevant topic.
Quote:
Originally Posted by Heinrich S

Jneutron, you seem to know a lot more about this topic than Arnyk. This is why following these kind of discussions is difficult because so-called experts chime in and present statements which are called "myth" which are later busted by someone who has an even greater understanding of the relevant topic.

Forums can be very difficult as a result.

That said, arnyk is a very learned individual, with a huge amount of expertise. He is indeed an expert in a significant amount with respect to audio and systems, so his opinion cannot be casually dismissed. (That is my personal opinion of course..)

My ultrasonic tweeter failure stuff is outside the realm of most pro audio people only because the pro's cannot waste their time troubleshooting a silly 25 dollar tweeter diaphragm. They are field replaceable, and that is where the repair effort needs to go, as time is money. I on the other hand, was not in a limited time position, and had the equipment available to perform further examination. Also, as he essentially pointed out, he doesn't stress his systems like I do. I beat the living daylights outta my stuff on the road at times, kinda like a crash test dummy...

He and I sometimes clash, but that's personalities. He's still a good, expert resource despite what I say about him....

jn
Edited by jneutron - 7/8/13 at 2:04pm
Jneutron, you have considerable experience yourself. I have a question or two for you :

The power supply size is what can improve headroom. So the power supply you get in a 80W amp and the one in a 500W amp. What would happen if you took the power supply out of the 500 watt amp and put it in the 80 watt amp? You would get more headroom? But it wouldn't be 80 watts, correct? f

Headroom is limited by the power supply, generally or is it more complicated than that?
Quote:
Originally Posted by Heinrich S

Jneutron, you have considerable experience yourself. I have a question or two for you :

The power supply size is what can improve headroom. So the power supply you get in a 80W amp and the one in a 500W amp. What would happen if you took the power supply out of the 500 watt amp and put it in the 80 watt amp? You would get more headroom? But it wouldn't be 80 watts, correct? f

Headroom is limited by the power supply, generally or is it more complicated than that?

often the amplifier circuit is designed for a certain maximum current, so just putting in a larger power supply could make no difference, as the amplifier won't put out more current than it's designed to. Also, each transistor has it's own individual power handling which means there is usually some type of protection circuit so the transistors don't catch on fire. There are many other factors such as the fuse on the AC input of the power source, set gain of an amplifier, and design goal of an amplifier.

It's possibly if the only thing holding an amplifier back is the power supply, that upgrading it could make the amplifier more powerful.
JW, well put.

In general, I suspect the supply and the circuitry in modern day amps are pretty matched. As JW said, upping the supply could be dangerous.

Sometimes people will add lots of caps to the supply to make it stiffer, but raising the supply voltage is a no-no, somethings going to fry.

jn
Okay, so if you take the power supply out of the theoretical 500 watt amp and put it in a 80 watt amp you could potentially have a dangerous situation on your hands. So the power supply on it's own ... isolated is not the be-all source of headroom? Apologies for making you guys run around, I'm stupid and I'm trying to better understand this with my limited experience.
Quote:
Originally Posted by Heinrich S

Okay, so if you take the power supply out of the theoretical 500 watt amp and put it in a 80 watt amp you could potentially have a dangerous situation on your hands. So the power supply on it's own ... isolated is not the be-all source of headroom? Apologies for making you guys run around, I'm stupid and I'm trying to better understand this with my limited experience.

Correct. The amplifier circuit is usually designed with a specific power supply size in mind. Most likely a larger power supply would either blow the fuse on the amplifier or give no improvement at all. If for some reason the amplifier was designed with a larger power supply in mind, but they decided to change the power supply to a smaller one to save money, putting a larger one in could posibly give more power, but it is very unlikely that would be the case.
Quote:
Originally Posted by JWagstaff
The amplifier circuit is usually designed with a specific power supply size in mind. Most likely a larger power supply would either blow the fuse on the amplifier or give no improvement at all.

Headroom is not in the amplifier circuit. It is limited by the power supply.
Quote:
Originally Posted by Heinrich S

Okay, so if you take the power supply out of the theoretical 500 watt amp and put it in a 80 watt amp you could potentially have a dangerous situation on your hands. So the power supply on it's own ... isolated is not the be-all source of headroom? Apologies for making you guys run around, I'm stupid and I'm trying to better understand this with my limited experience.
Since more power = more supply voltage, you are guaranteed to have an incompatible situation. As noted, you will do do serious damage to the 80 watt amp given that it was designed to run at much lower voltages.

What you want is a higher current, not higher voltage. Then you will have your increased headroom assuming it was power supply limited (and not say, the protection circuit in the amp).
Quote:
Originally Posted by OllieS

Headroom is not in the amplifier circuit. It is limited by the power supply.

Yes, that is what I'm saying. But say you have a 300 VA power supply with 40 V rails, and you put in a 500 VA power supply with 40 V rails, you will have more headroom, but the amplifier will not necessarily be able to use it because it could have a fuse that will blow right at 300 VA or the currrent limiting circuit could not allow more.

In some cases if the amp is only limited by the power supply, and you put a 500 VA 40V rail power supply in place of a 300 VA 40V rail power supply you will get more headroom.

Basically what I'm trying to say is you can't just drop a larger power supply into an amplifier without looking at all the variables to make sure it will be safe or give an improvement.
So if I am following this correctly................

With a properly designed amplifier, the power supply should be able to provide the required voltage during max load without sagging? This would mean that the current capability needs to be there, because without the power supply to maintain the voltage without sag, it must have the available current behind it.

I noticed a huge difference in sound and performance/sound quality with my Paradigm Studio 60's when going from a standard 110W Receiver to a "High Current" (I know marketing term) 110W amplifier.

This is an interesting read being that I have a background with Switchmode Power Supplies and how they regulate voltage with respect to static and dynamic loads.
Arnyk/Jneutron, I have a question for you. Sorry to hijack the thread, but it's somewhat related. As far as power requirements go, assuming reasonable speaker sensitivities and seated distances, would you reckon 10 times the average power would be a good and safe bet for dynamic swings?

So if an AVR on the test bench using typical sine waves can deliver 80-100 watts at low distortion, which is pretty typical nowadays, using a reasonable speaker sensitivity of 88-90 dB and seated 2-3 meters away, the average power on idle with real world musical signals would be, I assume, a few watts at most?

Sine wave testing is not indicative of real world usage of amplifiers using real world music, but how much additional overhead could you expect with real world music signals? If you could perhaps elaborate more on that, it would be appreciated.

Thanks!
Quote:
Originally Posted by majek 60

So if I am following this correctly................

With a properly designed amplifier, the power supply should be able to provide the required voltage during max load without sagging? This would mean that the current capability needs to be there, because without the power supply to maintain the voltage without sag, it must have the available current behind it.

(Numbers provided for the non technical people)

Yes. If the supply can't hold up the rail, the output voltage may suffer. If I make a 42 volt supply for a 100 watt amplifier into 8 ohms requiring 5 amperes peak, I've built in no room for sag. If I used a 1 farad capacitor (very big) for the rails, when the diodes are not replenishing the capacitor the peak output current of 5 amperes will cause a droop of 5 volts every second, or 5 millivolts every millisecond. If I use a 1 millifarad capacitor (1000 uf), the voltage will drop a thousand times faster, or 5 volts per millisecond.

The power line in the usa is 60 cycle, and a full bridge supply will replenish every 8.3 mSec. So the 5 volt per millisecond number very quickly reduces the supply voltage to mud. The supply must not fall below the output voltage plus what the output transistors require for 5 amperes operation, or the output voltage will drop below what the signal should be.

To make the supply better, I can either make the supply bank huge, raise the supply voltage so the droop doesn't make the rail very low voltage, or a combination of both which is what is generally done.

The tradeoff is that higher voltage means the transistors will dissipate more heat meaning bigger heatsinks and transistors, and need to be higher voltage. Cost tradeoffs.

Large caps get expensive, and take up room.. Again, tradeoffs.
Quote:
Originally Posted by majek 60

I noticed a huge difference in sound and performance/sound quality with my Paradigm Studio 60's when going from a standard 110W Receiver to a "High Current" (I know marketing term) 110W amplifier.

I would guess you ran into output current protection issues.

The swtpc 250 had output current limits as well as a foldback circuit for transistor dissipation limits.

jn

ps..had to fix the rate error...duh, math ain't my strong suit...
Edited by jneutron - 7/11/13 at 7:01am
Quote:
Originally Posted by goneten

Arnyk/Jneutron, I have a question for you. Sorry to hijack the thread, but it's somewhat related. As far as power requirements go, assuming reasonable speaker sensitivities and seated distances, would you reckon 10 times the average power would be a good and safe bet for dynamic swings?

So if an AVR on the test bench using typical sine waves can deliver 80-100 watts at low distortion, which is pretty typical nowadays, using a reasonable speaker sensitivity of 88-90 dB and seated 2-3 meters away, the average power on idle with real world musical signals would be, I assume, a few watts at most?

Sine wave testing is not indicative of real world usage of amplifiers using real world music, but how much additional overhead could you expect with real world music signals? If you could perhaps elaborate more on that, it would be appreciated.

Thanks!

arnyk is better suited to answer that. My typical applications do not in general have headroom considerations but simply as much power as can be provided.. But as arnyk pointed out, even systems like my little one at 300wpc, will use maybe 10 watts in normal use for normal listening of quality recorded music. Nightclub/dance applications do not use quality recorded music. And when my amps had analog meters, I could hear the pointer hitting the upper stop peg.

jn
Edited by jneutron - 7/11/13 at 7:02am
Quote:
Originally Posted by jneutron
arnyk is better suited to answer that. My typical applications do not in general have headroom considerations but simply as much power as can be provided.. But as arnyk pointed out, even systems like my little one at 300wpc, will use maybe 10 watts in normal use for normal listening of quality recorded music. Nightclub/dance applications do not use quality recorded music. And when my amps had analog meters, I could hear the pointer hitting the upper stop peg.

Your input is appreciated, as always. Thanks, I'll wait for Arnold to reply when he has the time. If have a question concerning the super position principle then I'll come knocking on your door.
Quote:
Originally Posted by goneten

Your input is appreciated, as always. Thanks, I'll wait for Arnold to reply when he has the time. If have a question concerning the super position principle then I'll come knocking on your door.

Uh oh, no getting kinky...

BTW, I just remembered... When I had my system set up in my living room, during normal listening levels, the amp light which indicated a signal normally didn't illuminate. It was set to turn on at 300 milliwatts, and it was a 300 wpc amp. So I had lotsa headroom, 3 orders of magnitude..

The system was about 99 db sensitive.

jn
Quote:
Originally Posted by jneutron

Uh oh, no getting kinky...

BTW, I just remembered... When I had my system set up in my living room, during normal listening levels, the amp light which indicated a signal normally didn't illuminate. It was set to turn on at 300 milliwatts, and it was a 300 wpc amp. So I had lotsa headroom, 3 orders of magnitude..

The system was about 99 db sensitive.

jn
Do you have any experience you want to share on equipments designated for consumer market?
When we speak of output current limit protection, this is what we speak of.

This is how the tiger 250 is protected.

The blue transistors bypass the output transistor drive current when they conduct. They get their info from an emitter stabilizing resistor through another resistor, both hilitedyellow. When there is too much current through either R39 or (40), Q14 (15) will conduct. This is the current limit feature of the outputs.

The resistors 43 and (44) hilited red, are foldback resistors. They connect to the output, and when the output is closer to one of the rails, less current will be required to turn off the opposite outputs.

For example, if the output is far negative, Q14 will require less voltage from R39 to turn on. The circuit lowers the current limit when a transistor has more voltage across it.

If I wanted to goose the output current ability for current headroom, I could change either R43/(44) or R41/(42). But these parts exist to protect the output transistors, so changing them carries a risk.

jn

ps..added the () to distinguish the negative side elements from the positive side.
Edited by jneutron - 7/11/13 at 1:18pm
Arnyk, are you still alive?
Tiger, from SW? Blast from the past...

The peak-to-average ratio of music was measured ages ago (AES) and was typically 17 dB, a 50:1 power ratio. It was a big deal when CDs first came out and people (recording engineers) had to deal with the fact that 0 dBFS actually was clipping compared to the usual practice of building in 6-10 dB'ish of headroom over 0 dB on a tape machine's Vu meters. If your system is perking along at 1 W average you need 50 W for max peaks. Different music may have different crest factors; pop is probably less, some orchestral and organ probably a little more. I have read on here ratios of 20 to 30 dB for movies (100x to 1000x in power) but have no personal data.

These days I worry about my hearing and tend to keep things quieter. My system can play 105 dB or a hair above and that is plenty.

Also note most folk won't notice a short-term clipping event in the middle of a bunch of loud stuff, those peaks you probably won't miss. Constant clipping sounds harsh but these days AVRs and speakers are such that I suspect clipping is far less an issue than most think. Back when 80 - 85 dB/W/m speakers were common, and 25 W/ch receivers were mid- to upper models with 100 W/ch amps considered large, clipping may was more an issue.

All IMO ('cept the AES data which is seared into my memory but I am no longer a member so can't reference the actual paper) - Don
Quote:
Originally Posted by DonH50

Tiger, from SW? Blast from the past...

Yup. I still have one pair left, still works, combined them into a 3RU chassis with internal fan cooled heatsinking. Also 3 SL-220's, 2 SL-5100's, and a dynaco 400 all working (the belts on the 220's actually fell off and left a sticky residue on the base, a salvadore dali turntable.. totally funny.. Gonna sell off the dynaco and 2 of the 220's.
Quote:
Originally Posted by DonH50

I have read on here ratios of 20 to 30 dB for movies (100x to 1000x in power) but have no personal data.
For movies, I have to agree. They tend to overdue the dynamics with the explosions.
Quote:
Originally Posted by DonH50

These days I worry about my hearing and tend to keep things quieter. My system can play 105 dB or a hair above and that is plenty.

Agreed. I don't play anything at home loud, just pleasant. When I do a gig where I have to mix vinyl or cd's, I'll keep the phones on (full ear cup for protection), and run their level quite low.

The normal home system is probably 85 db sensitive with 40 or 50 wpc, more than enough. I'm seriously entertaining building basshorns again, I loved the ~104 db spl numbers, but didn't want to store the cabs. Used to run 4 speakerlab K's, loved those things..

jn
New Posts  All Forums:Forum Nav:
Return Home
Back to Forum: Audio theory, Setup and Chat

### AVS Top Picks

AVS › AVS Forum › Audio › Audio theory, Setup and Chat › Watts and Current