Underpowering the cabinet

Discussion in 'Amps and Cabs [BG]' started by 4String_Ninja, Sep 9, 2000.

  1. 4String_Ninja


    Aug 14, 2000

    I have a genz benz 610xb and i have to wait a little while before i get my Gallien Krueger 1001rb, so right now i'm using a gk 400rb (240 watts) and the cab holds 750 watts. Could I damage the cab or amp by feeding a big cab a smaller amp?

  2. Only if you drive the amp to the point of clipping - that can cause speaker failure.

  3. what's clipping the amp???? I've many ppl say stuff about it but I have no clue...

  4. ihixulu

    ihixulu Supporting Member

    Mar 31, 2000
    getting warmer
    Basically clipping the amp usually happens when you turn everything up and try to play for a while. You will hear a very nasty LOUD electronic sound that sounds, for lack of a better analogy, almost like a death-ray pulse from a sci fi flick. If you hear that, and keep going with it, you have a good chance of frying your speakers regardless of their power handling capability.

    I am not that technical but my understanding is that the power amp can't cope with the demands placed on it and changes the waveform of the output to square waves, which results in impossibly rapid changes in speaker direction. The speaker can't handle the waveform, equivalent to trying to change direction 180 degrees on a dime in a car travelling 90 miles per hour, and gets mechanical failure.

    If I am wrong in my explanation, please feel free to correct/clarify. I'd like to know this better as well.
  5. i'll state it like somone else on this board stated the sound "your speaker starts to sound like its farting" it just makes a sound that even if you dont know what it is, there is no way you could ever like that sound, and you'd know somthing is going wrong.
  6. I've heard my cab fart a few times and it would keep doing that for like 10 min or until I turned it up a notch and it would start working normal again.. that's why I'm saving up for a new cab =)

  7. MikeyD


    Sep 9, 2000
    Hi. This is my first post here in talkbass, so I hope it comes out clearly. Clipping is a term coined for the process by which an amplifier "clips" the peaks of a (typically voltage) waveform because the waveform has exceeded the amplifier's voltage capability. If you have a sine wave, for example, clipping will essentially put a plateau on its peaks, and you can often hear this as unpleasant distortion. Because the clipped waveform spends more time *at* the limiting voltage than a non-clipped waveform, the average power delivered to the speakers will be higher. The ultimate clipped sine wave approaches a square wave in shape - and can deliver twice the RMS power of the unclipped sine wave.

    Clipped waves contain much more high frequency power than non-clipped, which is why tweeters are often burned out when excessive clipping happens. This gets pretty technical (there's a lot of physics - electrical, mechanical, thermal, and magnetic stuff going on). There are certainly speakers that can handle square waves with no problem at all. Many synthesizers and distortion pedals can generate square waves, for example. It's just that you have to remember that a square wave can (a) deliver twice the power of a sine wave (which is smooth-sounding fundamental tone) of the same amplitude, and (b) deliver that power at higher frequencies, so any tweeters you have may be in jeopardy.

    You mentioned mechanical failure due to fast reversals - in a way, yes, but what happens is: most woofers won't exactly follow the voltage command of a square wave and will, in effect, filter the fastness out of the waveform - thus turning the unused energy to heat in the coil. This is especially true if no crossover network feeds the higher frequencies to tweeters. So, yes, some woofer coils can burn up if they are fed clipped waves for too long - IF those clipped (or square) waves are at or above the woofers' power rating. I hope this helps clarify some things.
  8. You really know your stuff. Couldn't have said it better. :D:D:D

    I'd like to add this: if your amp clips, you can recognize it by a clearly visible wavering of the speaker cones. This is due to the unsymmetrical character of clipping. And because of that the amp's DC uncoupling will act up and DC signals will reach the speakers. Which shows up as a slow wavering movement. Which is VERY dangerous for speakers, because it can rip out (literally) the cones.
  9. MikeyD


    Sep 9, 2000
    Thanks. I see you are a resident tech "guru" here, Joris, so I'll ask you how clipping can produce DC (assuming the voltage rails of the amp's power supply are equal and opposite). Be as technical as you wish. I don't know how clipping necessarily is asymmetrical. I know it produces odd harmonics, but I still think there's very little DC produced, from a Fourier series standpoint. I've made a similar comment on one of the usenet newsgroups about amps generating DC. Please enlighten me on this. Thanks!

  10. Nice to meet someone who's got the same cuppa tea (for most people this stuff isn't their cuppa tea). And (a delayed) welcome to Talkbass!

    I'll tell you how I found out. I used to play through a home made amp (should be a picture of it the "Show us your amps" thread). It proved too low in power for my use. So it was constantly clipping, and the speakers made these strange movements, which at that time I couldn't explain. I thought it was normal. But then I bought a new amp, which didn't let the cones move the way the other amp did. So I took the (now spare) amp apart en hooked up multimeters and a scope on it. The scope showed a normally clipped signal, and the power rails stayed very much symmetrical at +/- 40 volts, dropping both just a bit when the output clipped. Couldn't find it. Must be the bass then. Refit a newly designed preamplifier with a 5 position gain stage and a very sharp (5th order Bessel 25 Hz) subsonic filter. I figured the bass was generating the wavering. I was wrong. Hooked up the scope again. Then I looked closer at the signal and saw that only the positive signal halfs were clipped off. These positive halfs were high peaks, and the negative ones were wide and flat. The signal would average at 0 volts that way, unless it clipped. I thought that that must cause an offset shift, and when I checked the DC uncoupling of the output stage, I saw a voltage of about 100 mV max. at the uncoupling capacitor. The gain of the output stage was 40, so 4 V of wavering DC came across the output terminals. I increased the cap's value by a factor of 10, but the wavering didn't disappear, it got slower, resulting in a more DC-like behavior, causing the DC protection to engage. Decreasing didn't work either: rumble was the result. By means of double checking I drove my new amp past clipping level and the result was (guess what) wavering speaker cones. Then I started looking at other speakers and concluded that almost every amp does this. I guess an amp must be DC coupled (not decoupled) in order to remain stable during clipping.

    I made a classical mistake: I thought of the bass guitar's signal as a simple (and thus symmetrical) sinewave. Just for fun I connected my bass to a spectral analyzer and saw harmonics on places I'd never expect them, and very many of them, at strange distances, even with the treble down. This must cause the unsymmetrical signal.

  11. MikeyD


    Sep 9, 2000
    Interesting. Thanks for your response. I'm not really an electrical engineer, per se (more mechanical), but I know a little bit. If I understand you correctly, you haven't actually put the voltmeters and 'scope on the new amp's output stage. I wonder if you are making a hasty generalization based on seeing certain cone behavior. I also wonder if your homemade amp's DC isolation should have been "amended" a bit: maybe the capacitor wound up maintaining a DC charge that should have been bled off to ground (i.e., the topology of the circuit may have needed a little tweaking).

    I have some schematic diagrams of power amps here, and some appear to have a ground connection through a resistor on the back side of their DC uncoupling cap. I can't say much more, because I haven't done "amp design" things in many years. However, it would be interesting to figure out the actual cause of this phenomenon, then to see what kinds of amps may be susceptible to it.

    I note that 4 volts of DC content will only contribute about 4 watts into a 4-ohm load, so at that level, a little extra heating of the coil and slight offset of the cone may not be a serious problem compared to the power content of the harmonics resulting from the clipping.

    The harmonics you see on a spectrum analyzer should not themselves have anything to do with DC. A Fourier series is simply an addition of sinusoids. But there may be some nonlinearities in the circuit that could generate DC if something unusual is happening. Maybe some of the asymmetry has to do with the voltage "envelope" of the bass note (including its overtones): the initial peaks get clipped, but the later ones don't, so the net effect is asymmetric... but somehow the cap. is retaining a charge that it shouldn't be retaining. Or perhaps something funny is happening with the amp feedback when ultrasonic frequencies are present during clipping. If you have any further insight about what's going on in the circuit, that would be great. We may be making a bigger deal than necessary about this, however, if the DC component tends to be a fairly low voltage, as you've indicated.

    Oh - just had another thought: I wonder if intermodulation (IM) distortion could be happening such that subharmonics are produced (y'know, the A+B and A-B frequency synthesis thing). Maybe the wavering is subharmonic content at very low (but not really DC) frequencies, such as a few Hertz. What do you think? Your voltmeter might read low AC frequencies as DC, depending on the sample period or integration window...

  12. Finally, someone I can learn things from!

    Thanks for the suggestions. However, I must point out that I didn't design the power stage of the amp myself. I just built it. Only the preamp is mine.

    true, but what I'm referring to is a mechanical overload of the suspension system of the drivers. With 4 volts of (almost) DC on it, a driver's cone will move several millimeters, maybe even a centimeter. Add the signal from the amp and you're talking about overstress. Not to mention serious (unsymmetrical) distortion. Some drivers will only move 2 mm linearly. And... it must have been more than 4 volts, come to think of it. A wavering of 2 mm wouldn't be noticable. I saw at least a centimeter of movement.

    I didn't try to look at the unbalance problem over a spectrum analyzer, I was just having fun with my bass.

    About your remarks regarding unsymmetry: I think you didn't understand my correctly. Because the positive half is clipped, the signal is already unsymmetrical. The average value is no longer 0 volts ( a matter of math). As a result, the decoupling cap gets the resulting (almost) DC, divided by the gain of the whole stage. The amp is merely trying to compensate for the unbalance. The DC offset is inevitable, unless the amp is DC-coupled. In which case, in turn, the power rails are unequally loaded, and maybe that will cause other problems.

    My homemade amp is an almost-class A type with an idle current of several Amperes. I think IM distortion is unmeasurable on this amp, even at clipping level....
  13. MikeyD


    Sep 9, 2000
    Once again, thanks for your response, Joris. I understand about DC being a byproduct of asymmetrical clipping; but I interpreted your earlier post as saying that the rails were symmetrical. Sorry for the confusion.

    At this point I remain unconvinced that a majority of amplifiers put out significant levels of DC when they are pushed into clipping. By "significant", I mean a level that can cause thermal or mechanical stress of the drivers. You may have experienced an isolated case of this situation, and I don't see enough evidence from the measurements or observations you have mentioned that DC is truly a factor worth being concerned about in *most* cases - at least relative to the other bad things that clipping can do (i.e., generate additional harmonic power almost equal to the level of the undistorted signal).

    My position would be that clipping is primarily a problem of an amplifier's delivering power in excess of its RMS rating, and that excess power may jeopardize tweeters in particular or woofers whose power handling is not up to the added burden. Again, I don't think substantial DC output is a typical consequence of clipping.

    If you can cite some papers or more in-depth technical information on anomalous DC output of amplifiers, I'm sure that would be helpful. Thanks for the exchange of ideas! It has been fun and enlightening.
  14. wreckall


    Aug 21, 2000
    Sorry to interupt the tech talk, but just a quick story. I once had a peavey 118 powered by a crate spa 200 amp (seriously under powered, I know now.) Which on the first gig I used it on displayed some irratic quivering motions when I stoped playing. The next day I realized the speaker wasn't working at all anymore. The dealer replaced it. And even checked out my poweramp. The amp was ok but he said the coil looked like it was fryed from dc voltage. "too bad that dealer wasnt half as knowlagable as one of you two" Anyway I sold the cab, cause I thought it was junk. Hugh, now I now.
  15. Thanks, wreckall, for sharing that. It supports my case.

    MikeyD, I know I can't prove my case. I'm not trying to either, nor am I trying to convince you of my point. We're simply having a very interesting discussion. I don't have leads to books about the subject. I only have my own experience. At least 4 amps I have seen, do this wavering business.

    The extra heating of the coil (the case of Wreckall) might still be due to the (almost) DC. If the coil is out of range of the front pole plate of the speaker (only takes a few mm), its behavior changes. The inductance almost drops to zero, because it has now become an air coil, instead of a heavily cored one. If I'm not mistaken, that leads to a lowered impedance, especially high frequencies. The high frequencies produced by the clipping amp now are fully dissipated in the low inductance coil, while causing no actual movement (no pole plate nearby to generate Lorentz forces).

    On the other hand, I've had only one speaker cabinet burn out with a wavering clipping amp. All the others stayed alive. Maybe because the speakers involved had seriously higher power capacity than the driving amp.

    Just some thoughts.
  16. mikemulcahy


    Jun 13, 2000
    The Abyss
    Or it could be the Wrigglin pin got stuck in the skinootin valve.
  17. Originally posted by Hambone in another thread:

    Must be it....

  18. reel big bassist

    reel big bassist

    Mar 27, 2000
    I've heard several things about clipping
    and I wonder if you could enlighten me.

    I've heard that it's O.K. to clip a tube amp, and
    you won't damage the amp or the speakers.
    Is this true? If so, why won't the amp or speakers
    be damaged?

    In this thread we are talking about clipping
    in the power amp correct?
    Would you damage anything by clipping the pre-amp?

    Greg P

  19. Clipping the preamp isn't as destructive as clipping the power amp. But when you set up your amp to have the preamp clip just before the power amp does, you have the same problems. But you should always be able to set your preamp gain in a way that it never clips.

    Preamp clipping is commonly used by guitar players, called distortion or overdrive.

    The speakers can't be damaged by a clipping amp (ss or tube) IF the power limit of the speakers isn't crossed. But a clipping amp CAN produce twice the rms power it's rated at when it heavily clips.
  20. Munjibunga

    Munjibunga Total Hyper-Elite Member Gold Supporting Member

    May 6, 2000
    San Diego (when not at Groom Lake)
    Independent Contractor to Bass San Diego
    I'm very suspicious of the capacitor that you guys mentioned. Seems to me, if it doesn't bleed fast enough (maybe because of some stray resistance creating an RC circuit), you would get a buildup of charge on it, putting a DC component into the signal. Am I stupid? Don't hurt me.

    Also, 4 volts DC doesn't seem like a heck of a lot to worry about, since there might be 100 volts AC driving the speaker, eh? It just doesn't seem to me that a bias of 4 VDC will displace the cone significantly. How much displacement does a 9-volt battery cause? Still stupid? OK, then, forget I said anything.