"Did you measure this or is it your estimated value from rail voltages?"
No, it actually clipped at about 160W/8R with both channels driven between 200hz~2Khz.
"this is plain wrong."
What part?
More than thirty years ago I thought clipping was bad, I even helped create the urban legends about DC and harmonics.
Rane has a good note on clipping and tweeters, ESP has a good note on low frequencies produced from asymetric signals.
"Yep, tweeters hate clipping."
Explained in the Rane note.
"(Woofers also, as DF becomes zero)"
You're right about the loss of feedback causing the output impedance to rise, but that is not what causes the damage. DF is a red herring, JBL has a note about that, and the ESP site has a clue as to what actually causes the woofer damage.
No, it actually clipped at about 160W/8R with both channels driven between 200hz~2Khz.
"this is plain wrong."
What part?
More than thirty years ago I thought clipping was bad, I even helped create the urban legends about DC and harmonics.
Rane has a good note on clipping and tweeters, ESP has a good note on low frequencies produced from asymetric signals.
"Yep, tweeters hate clipping."
Explained in the Rane note.
"(Woofers also, as DF becomes zero)"
You're right about the loss of feedback causing the output impedance to rise, but that is not what causes the damage. DF is a red herring, JBL has a note about that, and the ESP site has a clue as to what actually causes the woofer damage.
Hi,
When an amplifier clips the distortion increases enormously.
All of the distortion is at a higher frequency than the fundamentals that were in the original signal.
The clipped signal now has a higher proportion of higher frequencies than the unclipped signal.
The passive crossover splits the signal into upper and lower passbands.
The teble driver feels the higher proportion of higher frequencies as extra heating and start to wilt. Increase the clipping further and the treble driver will burn out.
Lets' take a hypothetical example.
A tweeter that has a crossover at about 3kHz will take about 5% of the average music signal.
100W of wideband noise will arrive at 95W into the lower frequency drivers and 5W into the treble driver.
Now clip the signal so that distortion goes up to 20%. That is 100W of the original signal and about 40W of harmonics.
Much of the harmonics are going to be in the treble passband and so let's assume that only a quarter of the harmonics are above 3kHz. This will be 10W of harmonics. But the crossover has already sent 5W of high frequency fundamentals to the treble driver.
So the poor treble driver now feels the effect of 15W of treble power i.e. 3times what it would normally have to handle and this is only for a quarter of the 20% distortion. These numbers can be much worse for a badly clipping signal.
To make things worse, the little treble driver has a very low thermal inertia and blows quite quickly when clipping occurs.
this bit is plain wrong.
When an amplifier clips the distortion increases enormously.
All of the distortion is at a higher frequency than the fundamentals that were in the original signal.
The clipped signal now has a higher proportion of higher frequencies than the unclipped signal.
The passive crossover splits the signal into upper and lower passbands.
The teble driver feels the higher proportion of higher frequencies as extra heating and start to wilt. Increase the clipping further and the treble driver will burn out.
Lets' take a hypothetical example.
A tweeter that has a crossover at about 3kHz will take about 5% of the average music signal.
100W of wideband noise will arrive at 95W into the lower frequency drivers and 5W into the treble driver.
Now clip the signal so that distortion goes up to 20%. That is 100W of the original signal and about 40W of harmonics.
Much of the harmonics are going to be in the treble passband and so let's assume that only a quarter of the harmonics are above 3kHz. This will be 10W of harmonics. But the crossover has already sent 5W of high frequency fundamentals to the treble driver.
So the poor treble driver now feels the effect of 15W of treble power i.e. 3times what it would normally have to handle and this is only for a quarter of the 20% distortion. These numbers can be much worse for a badly clipping signal.
To make things worse, the little treble driver has a very low thermal inertia and blows quite quickly when clipping occurs.
Hi Andrew,
I didn't work your numbers, but yes. I agree. That is what takes out tweeters and sometimes midrange driver units also.
The other thing that hasn't been mentioned is the average RMS continuous rating a woofer has in a 100W system rated woofer. I've seen 30W woofers (less than 1" voice coil former) where the glue simply let go. Speaker power ratings leave an awful lot to be desired.
-Chris
I didn't work your numbers, but yes. I agree. That is what takes out tweeters and sometimes midrange driver units also.
The other thing that hasn't been mentioned is the average RMS continuous rating a woofer has in a 100W system rated woofer. I've seen 30W woofers (less than 1" voice coil former) where the glue simply let go. Speaker power ratings leave an awful lot to be desired.
-Chris
PB2 said:
This is a good point - the crossover of a KEF 104/2 however takes up the almost the whole cross-section of the box. It is mounted on the bottom cover, where the feet of the loudspeaker is also mounted. Inductors are placed 90 degrees from each other (as much as possible) although there are many of them.
OTOH these inductors are all metal-core based, so too much power into a particular range can definitely cause saturation...
Cheers!
AndrewT said:
It's helped significantly by the fact that the T33 is ferrofluid-cooled, and that the magnet stucture of a T33 is fairly large. However I totally agree that regardless of that, it's still fragile - (and the ferrofluid keeps drying up!!!)
If you think a T33 is fragile, you should try the same trick on their UNI-Q series...
Still, my speakers are KEF. I like 'em.
Cheers!
AndrewT said:
Well, that Rane paper says it isn't.
http://www.audiovisualdevices.com.au/downloads/rane/note128.pdf
But Jim Lesurf says it's probably a combination of effects:
http://www.st-andrews.ac.uk/~jcgl/Scots_Guide/audio/clipping/page1.html
Jump to the conclusions here: http://www.st-andrews.ac.uk/~jcgl/Scots_Guide/audio/clipping/page6.html
Looks like another reason to use a light bulb as a limiter.
With respect to ferrofluid... I used that in commercail large format drivers when it came out and found it to be totally worthless. Its claim to cool and provide larger power handling capabilities proved to be a joke. Just more snake oil that left me with a bigger mess to clean up when a diaphram failed.
With respect to ferrofluid... I used that in commercail large format drivers when it came out and found it to be totally worthless. Its claim to cool and provide larger power handling capabilities proved to be a joke. Just more snake oil that left me with a bigger mess to clean up when a diaphram failed.
Chris,
The only speaker ratings that are worth any consideration are those that show a long term pink noise handling factor. I believe I mentioned this before.
With respect to the bridge mode switch and your amplifier load. There is a lot of misconception here probably caused by the dawn of the car audio craze. Most and I will repeat this....MOST home audio amplifiers are to be driven with a 8 ohm or higher load.
There are a FEW amplifiers that can actually drive a speaker load lower than 8 ohms and stay together.
Rule of thumb from a commercial standpoint is NEVER use a load lower than 8 ohms in bridge mode. Rule number (2) is never count on the resistance of the speaker wire as part of the load. This again is another commercial application note because you can pick up a few ohms of resistance in a long speaker run that would make your load to be larger. Strange things can and do happen sometimes when trying to drive a load like this.
Personal opinions about this thread... I am wondering if the poster of this thread isn't sitting back laughing at all of us that didn't consider he was bridging his amplifiers and paralleling up his load. Shame on us!
I offered a solution to the tweeter problem and very few looked at it seriously. I have pointed out it has been used religiously by me and others in the sould reinforcement field that people depend on for solutions to their problems. Its simple and it works without any negative side effects.
Power ratings for speakers are and will continue to be jokes until the industry adopts the commercail/pro audio standards and rates by long term pink noise.
The papers on clipping by Rane and others would seem(in my opinion) to say that an attitude of more is better in terms of amplifier power. This should be taken with an idea that some reasonable attempt to overdrive a speakers abilities is to be taken into consideration.
Rambling on if I may....
Years ago in my theatre system in my basement I managed to find and rebuild (5) Altec Lansing 9440 amplifiers. These are a Very good sounding amplifier for those that haven't experienced them. Anyway, I bridged all (5) of them to allow right,center,left, and the two surround speakers to have an abundance of power.
The speakers across the front were Klipsch Heresy speakers and the surrounds were in-wall Sonance speakers. Needless to say the sound was very punchy and quick. I played them at some extreme levels when watching movies. No damage was ever suffered by any of the speakers. Oh, the Klipsch were rated at 100 watts and the in-walls probably less than this. The Altec Lansing 9440 bridged at 800 watts each. Maybe more is better?
Hi Joe,
Amp design has more to do with that quick, punchy sound. Witness some amplifiers that are more punchy that higher powered models. So your amps are good to begin with.
I also agree that the car sound market has turned out more idiots than normal. When they hit real sound gigs we start to have trouble.
-Chris
I agree with you, there didn't seem to be reason to post on this though. Poly switches can be used, but their tripped resistance (higher than new) should be considered. I like to inconvenience people who abuse their product. 😀
Amp design has more to do with that quick, punchy sound. Witness some amplifiers that are more punchy that higher powered models. So your amps are good to begin with.
I also agree that the car sound market has turned out more idiots than normal. When they hit real sound gigs we start to have trouble.
-Chris
Re: Nad/Kef
You do some very fine looking work there at your website, Hemiguy. You'd probably like Alfa engines, they're hemi's as you probably know. I have a nice 6 cyl DOHC engine left from my spare parts that's fairly rare, it's small 2600 cc. This motor with 3 2bbl Webers and cams reground to Alfa Veloce specs would be real nice. There was a factory Weber option, but without the cams.
Your question is an interesting one, and I have a few more comments because there are some interesting unique aspects to your setup. One thing to keep in mind that I often repeat, is that what makes a system sound loud is distortion, give a system more power and we tend to keep turning it up. I see people write that their 30 W system plays plenty loud, no it produces lots of distortion that makes it sound loud. Your system has many features that keep the distortion down. Soft clipping, lots of peak power, and the woofer section is a bandpass type. The bandpass box acts as a filter which simplifies the crossover, however it also filters out distortion from the bass section, hamonic but not intermod by the way. So even your speaker soft clips, at least in the bass.
You chose real nice speakers, however they've gotten to be classics and there are better models to get what your looking for. You obviously like accurate sound, you just need a more rugged system. I suggest you give a listen to PSB Stratus Golds, and you'll probably want to add some serious subs.
Your thoughts on selling the KEFs are a good idea if you can get enough for them. If not, you might want to consider a tweeter upgrade and crossover rebuild.
Best Regards,
Pete B.
You do some very fine looking work there at your website, Hemiguy. You'd probably like Alfa engines, they're hemi's as you probably know. I have a nice 6 cyl DOHC engine left from my spare parts that's fairly rare, it's small 2600 cc. This motor with 3 2bbl Webers and cams reground to Alfa Veloce specs would be real nice. There was a factory Weber option, but without the cams.
Your question is an interesting one, and I have a few more comments because there are some interesting unique aspects to your setup. One thing to keep in mind that I often repeat, is that what makes a system sound loud is distortion, give a system more power and we tend to keep turning it up. I see people write that their 30 W system plays plenty loud, no it produces lots of distortion that makes it sound loud. Your system has many features that keep the distortion down. Soft clipping, lots of peak power, and the woofer section is a bandpass type. The bandpass box acts as a filter which simplifies the crossover, however it also filters out distortion from the bass section, hamonic but not intermod by the way. So even your speaker soft clips, at least in the bass.
You chose real nice speakers, however they've gotten to be classics and there are better models to get what your looking for. You obviously like accurate sound, you just need a more rugged system. I suggest you give a listen to PSB Stratus Golds, and you'll probably want to add some serious subs.
Your thoughts on selling the KEFs are a good idea if you can get enough for them. If not, you might want to consider a tweeter upgrade and crossover rebuild.
Best Regards,
Pete B.
Hemiguy said:
I suspect overdrive/ gain mismatch.
It looks like your power amps are overdriven, regardless of the output level. It may be anywhere between source to power amp's input. Please let me know if that resolves your problem.
It looks like your power amps are overdriven, regardless of the output level. It may be anywhere between source to power amp's input. Please let me know if that resolves your problem.
Fuse the tweeters. They are typically a much lower wattage than the woofer. When dome tweeters (if you have one) are overdriven they overheat and the sound will sound scratchy.
My speakers are rated for 100W. The woofer is rated for 100W, but the tweeter is only rated for 35.
I overdrive my tweeters too.
My speakers are rated for 100W. The woofer is rated for 100W, but the tweeter is only rated for 35.
I overdrive my tweeters too.
Hi,
Kiss is exceptional in having a tweeter that can take 35% of the rated input power.
Many speakers can only pass 5% of the rated input power to the tweeter.
Take care to listen for a clean (unclipped) sound. Clipping destroys tweeters.
Kiss is exceptional in having a tweeter that can take 35% of the rated input power.
Many speakers can only pass 5% of the rated input power to the tweeter.
Take care to listen for a clean (unclipped) sound. Clipping destroys tweeters.
Soft clipping
On an NAD 2200, you have an indicator that stays on all the time that the soft clipping circuit is enabled. This does not light only when "going into soft clipping". It becomes a clip indator if the soft clipping feature is disabled (by rear- panel switch). It sounds like you are actually hard-clipping the amp. This will certainly take out tweeters. I'm ignorant about the efficiency of your speakers, but concerned about your hearing. We are talking about lotsa watts in a livingroom!
On an NAD 2200, you have an indicator that stays on all the time that the soft clipping circuit is enabled. This does not light only when "going into soft clipping". It becomes a clip indator if the soft clipping feature is disabled (by rear- panel switch). It sounds like you are actually hard-clipping the amp. This will certainly take out tweeters. I'm ignorant about the efficiency of your speakers, but concerned about your hearing. We are talking about lotsa watts in a livingroom!
oops
I mixed up a couple of amps, but not affecting my previous statement too bad. I have another amp (Peavey) that uses the same light to indicate clipping or compression, depending on status. The 2200 uses different lights for soft clipping on and overload- which can only flash for clipping if the soft clipping circuit is switched off in back.
I mixed up a couple of amps, but not affecting my previous statement too bad. I have another amp (Peavey) that uses the same light to indicate clipping or compression, depending on status. The 2200 uses different lights for soft clipping on and overload- which can only flash for clipping if the soft clipping circuit is switched off in back.
another problem with tweeters and clipped amps, are bad clipping behavior. in many amp designs the outputs and/or the drivers go into saturation, and the transistor can hang on the rail, even after the amp should be off of the rail. see the pic below for an illustration. this type of clipping behavior is very bad for tweeters, especially if the slew rate at the end of the "hangover" is very high, as this generates large amounts of high order harmonics.
with NAD soft clipping, this should never happen, since the soft clipping circuit is supposed to clip the input signal at a level slightly lower than the clipping level of the output stage. the amp never gets enough signal to actually clip the amp if the soft clip circuit is adjusted properly. the soft clip circuit is supposed to act as an input limiter
with NAD soft clipping, this should never happen, since the soft clipping circuit is supposed to clip the input signal at a level slightly lower than the clipping level of the output stage. the amp never gets enough signal to actually clip the amp if the soft clip circuit is adjusted properly. the soft clip circuit is supposed to act as an input limiter
Attachments
Hi,
even soft clipping will modify the shape of the music envelope and pull down the peaks.
This has a similar effect in that it generates a higher proportion of harmonics in the output signal.
I guess the predominance is 2nd and 4th rather than the very high and many odd order harmonics.
The NAD will still send extra HF energy to the speaker if it is clipped whether soft or hard clipping.
The same advice as always:- avoid clipping (unless you're earning a living at reproducing bad music for a disco).
even soft clipping will modify the shape of the music envelope and pull down the peaks.
This has a similar effect in that it generates a higher proportion of harmonics in the output signal.
I guess the predominance is 2nd and 4th rather than the very high and many odd order harmonics.
The NAD will still send extra HF energy to the speaker if it is clipped whether soft or hard clipping.
The same advice as always:- avoid clipping (unless you're earning a living at reproducing bad music for a disco).
- Status
- Not open for further replies.