2-for-1: Samson Amp / Amp Clipping

Discussion in 'AV Receivers' started by Keith Hyde, Jan 30, 2004.

  1. Keith Hyde

    Keith Hyde Stunt Coordinator

    Joined:
    Dec 28, 2003
    Messages:
    106
    Likes Received:
    0
    Hi folks,

    Couple items I'm trying to learn about.

    First, I've read a lot lately about amp clipping on this forum, but don't really understand what that's all about. Anyone with a simple (or if you're energetic, lengthy) explanation?

    Second, been chatting about putting amps behind your speakers and the improvements in sound over your standard A/V receivers. I've been recommended a $200 Samson torroid-style amp, bridgeable, 120 watt, for each speaker ideally (mains and center mainly - don't really need it on surround). According to one, its a "true" 120 amps, 5hz to 100khz, that makes a load of difference.

    Here's the item specs:
    http://www.samsontech.com/products/p...D=84&brandID=2

    Here's my little brainstorm: instead of dropping huge $ into the biggest baddest A/V amp (like a high end Denon AVR, for ex.), why not get better sound by dropping $ into the amps, and buy the simplest Denon AVR just to keep you up with the new formats and decoders (ES, EX, etc.). Seems like a $ saver in the long run. This will culminate for me in about a year when I look at picking up some Studio 100's or near equivalent and will still be running my 75w.p.c. Denon AVR 1800.

    Thanks.
     
  2. John S

    John S Producer

    Joined:
    Nov 4, 2003
    Messages:
    5,460
    Likes Received:
    0
    Beware, Samsung amps, tend to hiss/make noise with the gains high. This seams universal on every one of them I have seen.

    Clipping most commonly ocurres when the input gain is to high for the early amp stages.

    Clipping also happens when an amp runs out of availble power for its needs, and when amp stages are overworked / driven / to hot / ect...ect...
     
  3. John Garcia

    John Garcia Executive Producer

    Joined:
    Jun 24, 1999
    Messages:
    11,571
    Likes Received:
    24
    Location:
    NorCal
    Real Name:
    John
    You won't necessarilly get better sound by going with a bare bones AVR with amps, but you have the right idea. Your sound is still very dependent on the DACs in your other gear, so if you are using the receiver to do the decoding, you are better off getting a mid level receiver with better DACs. You also have to consider interconnects between the receiver and amp(s) as they will somewhat influence the sound also (as well as added cost and power requirements, so it's not entirely a saver).

    Clipping will be enitrely determined by your overall system. Your specific speaker selection should be tailored to the amp and vice versa to prevent this.

    Running a full compliment of Studios off an 1800 would be asking for clipping, but 2 100s would probably be OK up to a certain point.

    Clipping will sound like the speaker is straining. The sound will become less clear, almost "fuzzy", possibly crackling.
     
  4. Keith Hyde

    Keith Hyde Stunt Coordinator

    Joined:
    Dec 28, 2003
    Messages:
    106
    Likes Received:
    0


    So how does clipping present itself? How would I know it if it is occuring or not, or to what extent?
     
  5. John S

    John S Producer

    Joined:
    Nov 4, 2003
    Messages:
    5,460
    Likes Received:
    0
    Ooops my bad, I did mean Samson... They have this issue with all their so called pro-amps....

    They work and perform ok, just irritating that they make so much noise when the gain is turned all the way up...


    Clipping doesn't always become audible.

    On this amp, you will see lights flash that indicate the incomming signal has become to hot, back it off some.

    These types of amps do light the light a little before actual clipping happens, so an occasional blip of the light is fine.
     
  6. Brad Wood

    Brad Wood Stunt Coordinator

    Joined:
    Nov 25, 2002
    Messages:
    88
    Likes Received:
    0
    A quick note on clipping. Clipping occurs when the voltage applied to the input is too great for the amplifier to handle. The danger in clipping is two fold. First, the amplifier sends a square wave to your loudspeaker. As a rule, loudspeakers don't like to reproduce square waves and damage can result. The other problem is unless your amp has some kind of overload protection or "soft clip" the clipped signal to your speaker is twice the peak wattage of your power amplifier. So if you have a 100 watt amp, the clipped signal is 200 watts. If you're speakers aren't prepared for that amount of power then you've got problems. Most people clip amps because they don't have enough watts to handle either the speaker's sensitivity or their own listening preferences. That's why most will say that it's more dangerous to under power a loudspeaker than overpower it.
     
  7. Keith Hyde

    Keith Hyde Stunt Coordinator

    Joined:
    Dec 28, 2003
    Messages:
    106
    Likes Received:
    0


    Well you just lead me into the next thing on my list to educate myself about - sensitivity and what the ratings mean / how they are interpreted.

    I know the spec sheets of speakers include a sensitivity rating, suitable amp power range, and max input power. So if I figure right... if I freak out my Denon AVR1800 at 75w.p.c. and it clips, I will momentarily feed my speaker 150 watts. So if I were to use the case of my Studio CC470 (90db/87db, 15-180 watt suitable amp power, 130 watt max input power) frequent clipping would feed this speaker about 20 watts too much which could lead to damage (in crossovers, magnets, wiring?) which is bad. But so long as I have enough watts of amp/preamp power going into the speaker to prevent the voltage strain at reference level and subsequent clip, then I'm good. In an ideal world, I would be feeding the speaker wattage just a shade under the max input power? For the CC470, I would theoretically get the best sound it is capable of producing running around 120 amps? I'm a little puzzled by the difference in "suitable amp power range" and "max input power."

    I understand watts as the "muscle" behind drivers - the more muscle available, the faster and cleaner the driver/s can do the job they are assigned - with a weaker "muscle" the drivers can still do the job but they don't do it as cleanly as they could and they also may be begging for power so much to effectively do their job that it strains the power supply and they are backslapped with an amp clip. If this is generally right, now I need to figure out where sensitivity comes into the mix - or perhaps that is the similar to the "demand" a driver needs for wattage to work as it is told - because I still can't picture it.

    And that's all taking the watt ratings of amps at face value - not even factoring in advertised "rated" and "real" watt differences as mentioned in other HTF threads.

    This is a very informative thread for me so far. Thanks everyone.
     
  8. John Garcia

    John Garcia Executive Producer

    Joined:
    Jun 24, 1999
    Messages:
    11,571
    Likes Received:
    24
    Location:
    NorCal
    Real Name:
    John
    I think you have a pretty good understanding of the issue already.


    Heat generation in the voice could from trying to reproduce the inaccurate waveform is where the damage occurs, and most often shows up in the tweeter's smaller VC wires. Studios are efficient, meaning they don't require a ton of power to give playback, but they do like a lot of power for modest to reference levels without distortion (result of clipping), depending on your room. Now, the problem comes when you are watching a movie you've never seen and an unexpected burst of low frequencies or a lot of activity in most or all of the speakers comes along and you are very suddenly fresh out of current - damage can be instantaneous.
     
  9. DavidLW

    DavidLW Stunt Coordinator

    Joined:
    Nov 21, 2003
    Messages:
    161
    Likes Received:
    0
    Don't be misled into believing that a clip signal at 20w is less dangerous than one at 100w. Even if your speakers are rated to handle 150w. Because of the nature of a square sine wave, when your speakers tries to reproduce the signal, the speaker cone will stay in one spot too long and fry the voice coil at that spot. Speaker cones likes to keep moving with the sine wave. Quality speakers will fare better than cheap speakers if the clip signal is of a short duration. But even the best speakers will fry with a clipped signal from a 10w Radio Shack amp if played long enough. It's like what Ralph Nader said about the Corvair, "unsafe at any speed". It is never the speaker that causes the amp to clip. The speakers just happens to be in the wrong place at the wrong time. The only way for a speaker to damage a amp is if it causes a shorts across the terminals.

    With speaker sensivity, the basic rule is that the higher the sensitivity the less power (wattage) needed to drive them to a certain level. Another words, speakers rated at 90db would need twice much power to sound as loud as speakers rated at 93db (it requires twice as much power to make a 3db difference). Most speakers are rated between 85db and 95db. Remember, this has nothing to do with how good a speaker sounds. A lot of expensive High End speakers are in the low end of the scale because people that can afford them can afford to buy a High End 200w to 400w per channel amp. On the other hand if you're limited to a less than 50w per channel amp (like many tube amp audiophiles) then getting speakers with sensitivity rating in the 90's becomes important. But how they sound should be your main concern. 100w per channel will probably drive 90% of the speakers availble louder than you care to listen to. Speakers rated at 200w will sound just as loud as speakers rated at 100w if they both have the same sensitivity rating and the same amount of watts driving them. But the 200w speaker can go louder because you can drive them at 150w. A watt is a watt. How "clean" and "fast" it sounds doesn't depend on how much "muscle" it has behind it but how well the amp is built. Another words 20 watts from a High End 50w per channel amp will sound "better" than 20 watts from a cheap 100w per channel AVR. But they will produce the same volume with the same speakers because 20 watts is 20 watts, no matter the source.

    Your theory that your speakers will sound the best with 120w is way off base. The amount of power delivered to your speakers is always changing with the source material (unless you're listening to a test tone). I venture to say that upper 75% range of your amp should be reserve for the loudest part of the source material. You do not want to be driving a 200w amp at 120w while listening to the dialog of a movie when you know there's going to be gun shots and car crashes coming up. The amount of wattage needed to go from normal speech to a gun shot blast is at least 10 fold. Another words if you're driving your speakers with 10w with the speech then you're going to need about 100w for the gun shot blast (of course were talking DVD not TV) in order for it to sound realistic. If anything, you should choose speakers that sounds their best when driven with about 15 to 25 watts. This is going to be the range of probably 75% of your source material. Don't buy high power handling speakers in anticipation of your amp clipping, rather invest in a powerful enough amp that you won't need to drive to clipping. It's better to have 100w speakers with a 200wpc amp than 200w speakers with a 100wpc amp.

    There may be a lot to learn but it shouldn't (and doestn't) take rocket science to put together a descent HT system. Just don't go off the deep end and buy a $200 AVR for a set of $2000 speakers or vice versa.
     
  10. TanT

    TanT Stunt Coordinator

    Joined:
    Dec 17, 2002
    Messages:
    135
    Likes Received:
    0
    Regarding the fan noice of the Samson amp, unless you sit right next to it and there is no sound in the room you can't hear the fan noice. My Samson S1000 has dual 2 stages fans. By the time it goes into the high speed mode these will be too much sound in the room for you to be able to hear the fan noice anyway.
     
  11. Peter Johnson

    Peter Johnson Stunt Coordinator

    Joined:
    Nov 12, 2000
    Messages:
    136
    Likes Received:
    0
    There are a few corrections that I think need to be made about exactly what occurs when an amplifier clips.

    Yes, the wave is "squared off" when the amplifier clips, the cone does not just sit in one spot such as was suggested.

    A square wave consists of (very) higher order harmonics. That is, when an amp clips a increasingly larger amount of high frequency energy is output to the speakers. The worse the clipping, the more high frequency power.

    A speaker that is ligitimately thermally rated to say 100W will not actaully be able to take say 100W at 15kHz. It will be lucky to take more than 10-20W of a pure, unclipped sine way. The tweeter, few of which take more than 20W, is receiving all of the power rather than the 100W being spread across the entire frequency range as occurs in normal music.

    When an amp clips more power is simply routed towards the tweeters, which is what causes the damage.

    Power is what kills speakers, not clipping. The speakers dont care what they are trying to produce. And no....the amp just doesn't instantaneously double its output when it clips as was suggested [​IMG]
     
  12. Brad Wood

    Brad Wood Stunt Coordinator

    Joined:
    Nov 25, 2002
    Messages:
    88
    Likes Received:
    0
    Peter,
    I've been a professional sound engineer for over 15 years and the clipping/doubling of power that I mentioned was taught to me by every pro that I've come across since I first learned how to plug in a mic cable. I'll have to consult my reference manuals, but I believe that to be true
     
  13. Peter Johnson

    Peter Johnson Stunt Coordinator

    Joined:
    Nov 12, 2000
    Messages:
    136
    Likes Received:
    0
    For a sin wave with an amplitude of 1V (ie 2V p-to-p), the RMS voltage is 0.707V. For an 8R load, this represents an average power transfer (ignoring output impedance etc) of 0.5/8 = 62.5mW.

    For a perfect square wave with an amplitde of 1V, the RMS voltage is 1V. This represents an average power transfer of 1/8 = 125mW, which is exactly double that of the sine wave.

    So yes, a perfect square wave does contain twice the power of a sin wave of equivalent amplitude. However, a square wave will never be realised, and the wave becomes more square as the output is increased towards infinity. It does not occur instantly. It is the increased power at the higher frequencies that kills the tweeters...
     
  14. Claude M

    Claude M Stunt Coordinator

    Joined:
    Jan 18, 2001
    Messages:
    239
    Likes Received:
    0
    the tweeter no longer "sees" the input power as a wave with a frequency, it sees it as a dc current
     
  15. Peter Johnson

    Peter Johnson Stunt Coordinator

    Joined:
    Nov 12, 2000
    Messages:
    136
    Likes Received:
    0
    Incorrect.

    A fourier analysis on a square wave will reveal zero DC component. A square wave consists of a set of sinusoids. A perfect square wave contains an infinite number, requires infinite bandwidth and hence doesn't actually exist. A finite rise and fall time is therefore required.

    A DC component in a time-varying signal will represent itself as an offset, not an oscillating "straight line" on an oscilliscope, such as a square wave. They do exist (measure the output of your amp) but are usually a few tens (or less) mV in amplitude. These affect the woofer and result in an decrease in power handling, although the effect is marginal so long as the offset is small.

    On another note, even if it was DC it would never reach the tweeter due to the crossover. It would (and does) go straight to the woofer. This DC has nothing to do with clipping.
     
  16. Keith Hyde

    Keith Hyde Stunt Coordinator

    Joined:
    Dec 28, 2003
    Messages:
    106
    Likes Received:
    0
    Hi folks. Didn't mean to drop out of a thread I started - was out of town a week for work. Thanks for all the technical replies - despite a little bit of disagreement the concept is much clearer for me now. Good thread.

    I do listen at reference/near reference levels on HT (Avia calibrated 85db, but often mellowed down to what would be about 75 on playback) - less juice on music. I'm a little concerned about my Denon 75 wpc. I guess it is difficult to tell whether clipping is occuring or not until you fry a tweeter? On my el-cheapo mains this is a significant possibility, as the build quality is presumably not a contender of any sort, if clipping is indeed ocurring. They filled a needed void many HT development phases ago and are now the ugly ducks of the crowd.

    Regarding whining on the Samson amps - I'm not sure if John S meant the hiss/noise was through the speakers, or from the unit itself such as from a fan. It was my understanding this was through the speakers hiss, not the latter. Clarification?
     
  17. John S

    John S Producer

    Joined:
    Nov 4, 2003
    Messages:
    5,460
    Likes Received:
    0
    Yes, those dang Samson amps, perform fairly well, but if you run the gain on them all the way up. (This is how I tend to run amps) you will hear hiss/noise out of the speakers attached.

    This was true of every Samson amp I used through last summer anyways.
     
  18. John S

    John S Producer

    Joined:
    Nov 4, 2003
    Messages:
    5,460
    Likes Received:
    0
    PS: About your Denon AVR,

    If you hear no audible distortion when driving your volume levels, you will be ok, I am sure.

    But then again, I have an experienced ear, I can tell audibly when there is any audible fidelity effecting clipping going on. (or I like to think so anyways) lol
     
  19. Keith Hyde

    Keith Hyde Stunt Coordinator

    Joined:
    Dec 28, 2003
    Messages:
    106
    Likes Received:
    0
    Thanks John. I can't stand the hiss, with a passion. I'll be cognizant of your observation if I am ever led towards Samson.
     
  20. Jerry Parker

    Jerry Parker Stunt Coordinator

    Joined:
    Jan 15, 2001
    Messages:
    174
    Likes Received:
    0
    For those of you interested in knowing what happens when an amplifier clips, I highly reccomend visiting this link: http://www.rane.com/pdf/note128.pdf

    It is by far the best resource I have read on amplifier clipping.

    A couple points I would like to bring up though are:

    A clipped amplifier does not cause the speaker to "hang" at the peaks or make it move incorrectly. It still follows the SAME path it would with a clean input signal. All the "distortion" that makes up the square portion is odd ordered harmonics from the fundamental. So if your fundamental was 40hz, the 3rd harmonic would be 120hz, 4th would be 200hz and so on until infinity.

    When the amplifier clips it does not immediatly create a square wave. You must overdrive the input by 20-30dB before you even get close to a square wave.

    A 200 watt amplifier is perhaps less dangerous than a 100 watt one ONLY IF YOU DON'T CLIP THE 200 WATT ONE! Chances are, the 200 watt amplifier will be clipped as well, and that will be even worse for your speakers.

    Most important of all though is that the speaker doesnt know the difference between a clipped power and clean power. All it knows is just power. Feed it too much to make it move past its mechanical or thermal limits and you could destroy it. Otherwise you won't unless the unit is defective. That RANE link goes into much more detail on how this occurs and especially on why tweeters can be destroyed.
     

Share This Page