Underpowered amp could harm speakers?

[Gossy]

I found nothing in the DIN standards regarding dynamic headroom...which is the specification to which I was referring as being without a standard. Do you have a link? I'd be very interested in learning more about this. Thanks!

Not so much a standard for 'headroom' but several different ones concerning what max powerhandling can be - a short text which matches what I remember somewhat http://www.radiolocman.com/shem/shem-cache.html?di=18850 in some regards to this thread.

There are different standard tests - unless the manufacturer specifies which was used you just don't know. A salesmen I knew in the 90's once tried to collate some information about this from the various manufacturers to be able to present his customers, and found it even varied from year to year or even model to model from one manufacturer he approached ^^

 

[musician2358]

I know, but the operator is in control if the amp clips or not, regardless of the attached speaker. That is the point. And, as already stated, it still depends on trying to exceed some capability. It doesn't just happen.

Actually, it usually does just happen. I think kcbluesman's point was that you can set the gain with no clipping, and a sudden spike in amplitude, especially at a high frequency, can crash through available headroom and clip.

M2

 

[Guest105]

Actually, it usually does just happen. I think kcbluesman's point was that you can set the gain with no clipping, and a sudden spike in amplitude, especially at a high frequency, can crash through available headroom and clip.

M2

I don't remember making that point, but it is certainly true that unless you have enough in reserve to accommodate the highest possible peaks in the music you're listening to, clipping will ensue.

However, in general I would say that minor clipping (of a few especially high transients, for example) will have no effect on your gear, and very little on your sound quality.

I think that this issue is the one that drives the "can't have too much power" crowd (which I tend towards, by the way). Rather than worry about clipping, we just make sure we have enough power to assure that it doesn't happen (or, at worst, is minimal).

Of course, we must be judicious with the volume control to avoid overpowering our speakers...but the same can be said for those using lower-powered amps. Note that at full clipping (which one is never likely to experience in a real-world situation, other than perhaps being so drunk that your hand cranks the volume to full blast just before you pass out), a 35 watt amp will put 70 watts of continuous power across the full frequency spectrum. Not many tweeters can handle that kind of load!!

Regardless of how far into clipping you are - or even of whether or not you are clipping at all, cranking up the volume continues to push more and more power into any as-of-yet unclipped frequencies (and into the drivers that carry them).

 

[musician2358]

I don't remember making that point, but it is certainly true that unless you have enough in reserve to accommodate the highest possible peaks in the music you're listening to, clipping will ensue.

Maybe I read more than you intended into the statement:

It seems to me that a 100 watt amp running at 50 watts of continuous output has only 3dB (6dB to rail) left to accommodate peaks in the music.. peaks which are often 12dB up, or higher

And while I agree with the statement:

However, in general I would say that minor clipping (of a few especially high transients, for example) will have no effect on your gear, and very little on your sound quality.

I would add that it is a game of probability. Compression tames the peaks on most recordings, but, a horn solo on a jazz recording can have some major transients that hit high notes with a crescendo for instance, and the amp gain may be set for the main part of the recording.

M2

 

[Guest105]

Maybe I read more than you intended into the statement:



And while I agree with the statement:



I would add that it is a game of probability. Compression tames the peaks on most recordings, but, a horn solo on a jazz recording can have some major transients that hit high notes with a crescendo for instance, and the amp gain may be set for the main part of the recording.

M2

If you are listening to the Telarc 1812, 20dB is probably not enough headroom. And higher quality recordings are not heavily compressed...which is why I target 12-15dB of headroom when matching amps to speakers, and both to my SPL requirements.

Setting the gain for the main part of the recording -that is, such that the average SPL is at the desired listening level - is precisely what one should do, at least in my opinion. If at that listening level you've still got 12-15dB to peak output, I think you're good. YMMV!

The quoted statement was intended to make abundantly clear that running an amp at a continuous output anywhere close to its rated output will put it in clipping (with any sort of normal musical signal).

 

[Gossy]

All amplifiers clip some of the time. The only time when damage to voicecoils and such can occur is when the amplifier is outputting powerlevels near the driver max wattage levels, and a hard clip happens - then the cooling action of the driver's movements is lowered, and the sudden increase in temperature can cause the coil or it's former to deform.

Tweeter damage usually happens due to their limited excursion: cooling action in them is limited. Is why in many designs they employ ferrofluid to assist heat disappation.

 

[whoaru99]

Duplicate deleted.

 

[whoaru99]

Actually, it usually does just happen. I think kcbluesman's point was that you can set the gain with no clipping, and a sudden spike in amplitude, especially at a high frequency, can crash through available headroom and clip.

M2

No, it doesn't just happen...in the context of what I actually said in that post, which was rebuttal to your post about speaker load. Changing the speaker to one more difficult load doesn't inherently create clipping. There still needs to be something that pushes the signal into the red, so to speak.

In re-reading it I see where that point could be cloudy.

 

[Guest105]

All amplifiers clip some of the time. The only time when damage to voicecoils and such can occur is when the amplifier is outputting powerlevels near the driver max wattage levels, and a hard clip happens - then the cooling action of the driver's movements is lowered, and the sudden increase in temperature can cause the coil or it's former to deform.

Tweeter damage usually happens due to their limited excursion: cooling action in them is limited. Is why in many designs they employ ferrofluid to assist heat disappation.

Nope.

 

[BilboBaggins]

I want to jump in here, and I haven't read the full thread yet. But I need to say it because I think it is real and relevant to this discussion.

When I first got back into audio, I started off with a McIntosh MAC1900 receiver. Very nice! I wanted really nice speakers to match it. So I bought a pair of Celestion Ditton 66 speakers. I was never really impressed with them. I found that the bass was muddy and seemed uncontrolled and the highs were harsh. I used an EQ to try to tame this, but was never happy with it. So eventually, I traded the Ditton 66s for a completely restored Sansui 9090. Double the power! And I replaced the Ditton 66 with Ditton 44s. Same issue, just not as in my face. I was not happy with the sound. I eventually added a pair of Kef Concerto's to the mix. These sounded amazing at the sellers place, but lackluster at my place on the 9090.

I was starting to think I needed more power even though everyone on the forums were telling me I should be fine. I bought a Kenwood Supreme 500. Voila! Problem solved. The Kef's came to life and so did the Ditton 44s to a lesser extent. The 44s moved on and the Concerto's are still here.

My interpretation is that neither the MAC1900 nor the Sanui 9090 had enough reserves for these 4 Ohm speakers. The Kenwood does. I have moved into bigger and much more powerful amps and it is wonderful. I am driving my 50 watt Concerto's now with a 200 WPC amp and it sounds great! I have driven Ditton 15s on my Bose 1801 and they sound great on it. I had a Sansui B2101 200 WPC amp and the kef and the Ditton's didn't sound great. The B2101 is an 8 ohm amp and works great with many 8 ohm speakers, but not all. I replaced my Yamaha M-40 with the Sansui B2101, to my regret. I hadn't tried the Sansui on those 4 ohm speakers before I moved the Yamaha onward.

With the big amps I never have to worry about the speakers being too much for the amp.

 

[Gossy]

Yes, plain simple power available at the impedance speakers are rated for is a pretty simple rule of thumb to get your speakers to shine!

Amp>speakers or amp<speakers w regards to clipping -so much information when relayed without understanding is transformed into urban myths.

 

[Guest105]

That's for sure.

 

[musician2358]

Changing the speaker to one more difficult load doesn't inherently create clipping. There still needs to be something that pushes the signal into the red, so to speak.

Actually, the point about impedance is that it can do just that. A given source that doesn't clip in any meaningful way at a given amplifier gain control setting with 8 ohm speakers could clip on the same source and same gain setting with 4 ohm speakers, as significantly more current is running through the amp and things are running hotter. Clipping can be a persistent phenomenon when, say, you are trying to draw more power than the power supply can supply, or a transient phenomenon when a component in the amp is driven out of spec for some, possibly brief, period of time and emits an out of spec pulse.

Recording studios typically over spec the power ratings of amps by a lot more than home listeners in part because they are working with loads that haven't been mastered yet and may have all sorts of untamed things going on (you can usually buy replacement tweeters from a pro monitor vendor years after the speaker went out of production for that reason).

Definitely don't crank the volume when listening to someone's home recording on soundcloud, it is likely more raw than a mastered CD released by a recording label.

M2

 

[whoaru99]

Actually, the point about impedance is that it can do just that.

M2

I'm quite clear on load/impedance, and stand by my previous point. Further to it, at some point one has to have enough cognition to realize a change has been made to 4 ohm speakers (or whatever), and to proceed with that knowledge.

 

[Gibsonian]

All the headroom stuff is not meaningful to me. It's just the manner on how the manufacturer decides to rate it's power. Wish they'd all rate them in the same manner.