Some Basic Audio Questions

Discussion in 'General Electronics Chat' started by WillS_70, May 19, 2014.

  1. WillS_70

    Thread Starter New Member

    Apr 26, 2014
    26
    1
    Hello.

    Regarding audio and amplification; I have absolutely no handle at all on certain concepts that are further obfuscated by some of the marketing "tricks" employed by sales departments. So, with my sincere apologies to all audio buffs, but please help;

    Similar threads, I wager, have probably been started before, but here goes anyway.

    1. Speaker / amplifier wattage issues:

    There's a lot of jargon (subs, tweeters, woofers, etc), which basically boils down to, far as I can see, the fact that each one of these speakers is actually a combination of two or more smaller speakers. I understand that the purpose of this is that one emits higher frequency sounds and the other lower frequency, and that there must be some filtering of the "composite speaker" input to ensure that each gets what it wants before emitting the sound. No problems understanding this so far, unless I have already strayed off track.

    Each one of these sub-speakers will have its own specifications, and whether they are wired in series or parallel within the "speaker" as a whole;

    a. How should I go about considering matching the speaker unit to an amplifier?

    b. For example, if a channel of an amplifier is capable of "putting out" 20W, what constitutes a suitable speaker to go on that channel? With this question I am after, first, what single speaker with no frills would be within a suitable range, and second, what "combination speaker", considering the individual speakers of the unit, would be suitable? I do understand that this question may well be about impedance matching.

    2. Pre-amplifier and circuits:

    This is something that really gets me. Why do you need a pre-amplifier? Let me go really basic here, please (note, my example may be completely inappropriate).

    a. Say I have a TEA2025 chip used bridged for mono, and I consider using an LM386 as a pre-amplifier for it. Say I am using this to amplify the sound from a crystal radio (I have already had success using just the TEA2025 for this purpose, but hesitate in using the LM386 for the aforementioned purpose as, honestly, I do not fully comprehend the concept). I feed the output of the 1W LM386 into the TEA2025, which can output 5W. What is the effect supposed to be, and why would I do this? Does it reduce distortion? Does it increase the peak volume, considering the same speaker?

    b. Further on. Is there any specific method for matching a pre-amplifier to an amplifier? For the sake of the argument, I elect to use an LA4440 amplifier, and I consider using an LA3161 as the pre-amplifier. Apart from the fact that the source voltages are the same, is this a suitable combination? Why? How would I know what other amplifier I could use in connection with the LA3161, or any other pre-amplifier?

    3. Tone control and equalization:

    a. So, treble and bass controls, or even the older fashioned single "tone knob", are coarse versions of the equalizer, far as I can tell. They filter frequencies with various capacitors and resistors, then accentuate one over the other with a potentiometer (please correct me if I am wrong, even though I have already done an experiment that worked under this "assumption"). Also as far as I can tell, this should be done on the input side, before feeding the signal into an amplifier. Am I right, or is this open for debate?

    b. There appear to be some dedicated ICs for equalization (LA3600, for example). So, it is the same question as 2b. How would I go about knowing what equalizer IC to use with a pre-amplifier / amplifier combination? When it comes to the equalizer chip, does this still go into circuit immediately after the raw input, ie; before the amplifier?

    Perhaps too many questions for one thread, yes, and probably laughable, too, but I thought I would ask them all together to attack all the issues on one front. If the answer is to throw in my direction some PDF that goes into some detail on these subjects, then that would probably suffice.

    Above all; Thank you!
     
  2. studiot

    AAC Fanatic!

    Nov 9, 2007
    5,005
    513
    0) Yes you have to beware the marketing departments and most especially their 'knowledgable' salesman.

    1) Yes what you say is basically true.

    2a) To understand the electrical side (amps and preamps) you need to understand what an electrical signal is and its properties. That is basically the difference between voltage, current, power and the (simple) relationships between them.

    2b) To understand the loudspeaker (transducer) side you need to understand the above electrical properties plus impedance.

    3) By all means discuss the whys and wherefores of preamps. Equalisation depends on the source and is of considerably less importance today. The rest of (3) also needs some understanding of the above electrical terms plus frequency.

    So how are we doing?
     
  3. bertus

    Administrator

    Apr 5, 2008
    15,649
    2,348
  4. wayneh

    Expert

    Sep 9, 2010
    12,153
    3,059
    1 a) Matching impedance and power rating is all most folks do. So you wouldn't attach a 5W speaker to a 100W amp, because of the likely result of blowing the speaker. Now, a 5W speaker and a 10W amp is within reason. Likewise it would not be wise to attach an 8Ω speaker to a 500Ω headphone amp, as the speaker would place too great a load on the amp. Conversely a pair of headphones might be destroyed by attaching it to a high power audio amp meant for driving 8Ω speakers.
    b) Personally, I'd be uncomfortable using a speaker that is not rated to handle the full rated power of the amplifier. It's not a hard rule at all.
    2. Pre-amplifiers bring low power sources up to "line level", and may be tailored to a particular type of input such as guitar, phonograph needle, microphone, and so on. All these can be brought to the standard line level, and then fed to a power amplifier. A pre-amp is also where the signal may be modified by filters, tone controls and so forth. Much easier in a pre-amp than at full power.
     
    absf likes this.
  5. #12

    Expert

    Nov 30, 2010
    16,343
    6,828
    Remember, some of this opinion.

    1) RMS power is the only valid specification.
    2) I buy speakers that can survive twice the wattage of the amplifier because, if everything goes wrong, the speaker will survive. Buying a 200 watt speaker for a 10 watt amplifier is a waste of money and will probably get you less performance than a more appropriately sized speaker.
    3)Multiple speakers in a single enclosure are already matched for loudness and impedance.
    4)The thing in the speaker enclosure that affects frequency is called a, "crossover". It's a passive network that directs the right frequencies to the right speaker.
    5) You need a pre-amplifier when the signal available is not enough to drive the output stage to its fullest potential.
    6) An LM386 is not a good pre-amp.
    7) You don't feed watts from one amplifier stage to another, you feed voltage levels until you get to the last stage, then that stage provides the power levels that qualify as serious watts.

    The usual method is to see if your input signal is strong enough to drive the power stage. If it is, you might stick in a tone control or an equalizer, but a tone control usually uses up so much voltage that you have to add an amplifier after it. That one is called a tone recovery stage. Active tone controls don't have such bad losses of amplitude.

    If your input signal is weak, add a nice, low noise pre-amp like a TL071.

    The deal is, first you get your signal up to a voltage that is pretty much safe from local noise, like about a volt. Then you do the modifications like tone controls. Then you send the results to a power amp stage.

    If you want answers about specific chips, it would save everybody a lot of time if you posted the datasheet for each chip you want answers for.

    OK. That's part of the answers.
     
    absf likes this.
  6. WillS_70

    Thread Starter New Member

    Apr 26, 2014
    26
    1
    My word, great answers that fill in some gaps neatly. Thank you all.

    I do like your approach to the questions. So how am I doing...

    I am conversant with fundamental electrical properties, yes.

    Where impedance is concerned, I understand that it is matching resistances of the speaker(s) to the output. The basic example is with my crystal radio experiments, to see if I have it right; one of my coils in circuit, after rectification and at the audio plug gives me a resistance of 2.3 M ohms. My earphone measures out at 1.9 M ohms, so I am good to go. Regarding the tolerance, its calculation or estimate, I am a bit more vague. Clueless, in fact!

    If a circuit and a speaker are not matched you will get either nothing (circuit too high) or that "woauwww" sound (speaker too high), I assume because it is letting the "wrong" frequencies through. Am I the right? There's a good article about impedance in bertus' link, which I am going to be reading right now. I am avoiding reading it before posting as I do not want to pre-educate my response to questions so far.

    I am generally conversant on frequency, too, as in I understand what it is, as in occurring in a circuit and (with much more familiarity) in radio waves, including modulations. That said, I would not be surprised if there were any new surprises, may the term of redundancy be valid!

    Great link, thank you.

    Okay, that is certainly one of the issues. One of the plans is to assemble a suitable speaker for my crystal radio to be feed by the successful experiment of using the bridged TEA2025 amplifier IC. What I plan might be a bit of overkill for an AM radio, but I wanted to do it that way as it is an opportunity for gaining some experience from scratch. The idea follows;

    Originally, I satisfactorily played the amplified signal through a 5W / 8 ohm speaker. I also have two additional speakers; an 8W and a 0.5W, both 8 ohms. I wanted to use both in a composite speaker and do some of that "crossover" #12 mentions. But I am confused here. Where I have no worries about feeding the signal raw to the 8W speaker, is it possible to filter enough out of the signal that the 0.5W speaker can be used? Or is that necessary? And, is this the way it is done? According to the TEA2025 datasheet, it can handle from 4 to 32 ohms, depending on the voltage. This means that my speaker configuration will determine what the source voltage should be, I assume, and considering this, leaves me a virtually free hand at wiring the two speakers either in parallel or series. What is better, for my purposes (ie; two dissimilar wattage speakers, one above and one below the amplifier output)?

    Thank you very much, #12. Excellent, concise information in that post. Point 2 in your post was already experienced performance-wise, as to be safe at one stage I played a 1W amplified signal through a 55W speaker. Buzzy and muffled. I more or less knew intuitively never to try and play a large amplified signal through a low wattage speaker, though from these posts I now know it is possible, within reason.

    Thank you all again.
     
  7. #12

    Expert

    Nov 30, 2010
    16,343
    6,828
    Your two speakers are a dreadful mis-match. If you put them in series, the 1/2 watt speaker is going to be the limit on your power because you shouldn't drive more than 1/2 watt through it, and the 5 watt speaker can only get 1/2 watt. If you put them in parallel, you shouldn't turn the power up higher than the 1/2 watt speaker can handle, and so, the 5 watt speaker can't get more than 1/2 watt. Forget about it. Bad idea.

    The reasons 2 speakers are connected is to either get more air moving for good bass response or to cover a wider frequency range than one speaker can do. A good bass speaker just can't get to 15,000 Hz. While you're at it, the 2 or 3 speakers in the set must have similar efficiency so you can hear all of them. That's where you have to get picky.
     
  8. WillS_70

    Thread Starter New Member

    Apr 26, 2014
    26
    1
    Thank you! That was just the the explanation to cause a facet of the problem to suddenly catch the light of understanding and shine. Makes me "face-palm" and ask myself why I could not see it that way to start with, as there is no mystery to it, after all.

    Idea forgotten. Big thumbs up.
     
  9. crutschow

    Expert

    Mar 14, 2008
    13,052
    3,244
    Just my two cents to cover some of the same ground:

    A preamp is just to amplify a low level signal, such as from a microphone, to a high enough voltage to driver the power amp. There is no significant power amplification involved. Also typically included in a preamp are volume and tone controls, etc.

    There is no "impedance matching" involved in connecting a speaker to a power amplifier, at least not in the conventional electrical sense. Modern solid-state amps have a very low output impedance (<1Ω) so you never "match" that impedance to the speaker's. But the amps do have a minimum value of impedance they can safely drive (due to maximum current and power dissipation considerations). They don't have an upper impedance limit (the amp will operate fine even with no load).
     
    absf, endolith and WillS_70 like this.
  10. WillS_70

    Thread Starter New Member

    Apr 26, 2014
    26
    1
    Slightly off topic;

    Something from my own post baffles me. That I put the units in MΩ is not a typo; it is the reading I got from my multimeter, which is a manual range selection type, the selection of which jumps from 200 KΩ to 2 MΩ on the resistance scale. Seems a bit high from what I have read, for a crystal radio circuit.

    Just wondering. Something is not right.
     
  11. alfacliff

    Well-Known Member

    Dec 13, 2013
    2,449
    428
    is that a digital multimeter? if not, the resistance indicated will not mean a whole lot when measureing a diode. and the "impedance " of a circuit is not measured with a multimeter, it is a complex combination of inductance, capacitance and resistance which varies with frequency. so just treat "high impedance" as anything over 5k Ohms, and low impedance as less. its like measureing a ceramic microphone, an open circuit with an ohmmeter, usually 10 to 50 kOhm impedance, sometimes 100k.
     
    WillS_70 likes this.
  12. endolith

    Member

    Jun 21, 2010
    27
    0
    It's called "average power", not "RMS power". It's derived from RMS voltage and RMS current, but it's the average of the power waveform. The RMS of the power waveform would be a different value, but nobody uses that for anything.
     
  13. Markd77

    Senior Member

    Sep 7, 2009
    2,803
    594
    Another option for you is bi-amping. Instead of amplifying the whole sound then passing it to a crossover to split it to two speakers, you put two different filters in between the preamp and two power amplifers. It is more efficient, but may be more expensive, although you can reduce the power of the amplifers a little because none is being wasted in the crossover. You can also use a lower powered one for the tweeter as they need less than bass for the same volume output.
     
  14. #12

    Expert

    Nov 30, 2010
    16,343
    6,828
    Guess again.

    I've been poking around on the internet and finding that there is yet another trend toward calling the RMS power something else. Right now, the trend is to call it the, "average" power. That thing which is called, "average" is still the RMS power according to:

    http://www.meyersound.com/support/papers/amp_power.htm

    When an amplifier is rated in RMS watts, this is a shorthand way of saying “average watts obtained by the RMS method.”

    However, RMS and average are two different things.

    http://www.rfcafe.com/references/electrical/sinewave-voltage-conversion.htm

    Vrms equation - RF Cafe ≈ 0.707 Vpk
    and
    Average voltage equation ≈ 0.636 Vpk


    Nobody uses RMS? Simply not true unless the nobodies you are referring to are the nobodies that are educated in the field of mathematics.

    Manufacturers and salesmen have intentionally confused the issue since vacuum tubes were powered by batteries. Now, after decades of proper usage, they are doing it again. I have provided links to two different definitions of, "average", and they contradict each other. Meyer Sound says you are just hanging another label on RMS power and RFcafe spells out the difference in mathematical terms.

    Rather than hang a different label on a well known mathematical formula, only for use in audio, I prefer to use the original term, lest it get confused with the real meaning of, "average".
     
  15. endolith

    Member

    Jun 21, 2010
    27
    0
    It's not a guess; it's a mathematical definition:

    P_{avg} = V_{rms} \cdot I_{rms}

    https://en.wikipedia.org/wiki/Root_mean_square#Average_electrical_power

    Yes, most people will know that this is what you meant, but that doesn't make it correct.

    ...which is why you shouldn't use one when you mean the other.

    Nobody measures the RMS of the power waveform. They measure the average of the power waveform, by measuring the RMS of the voltage waveform. This means that the peak power of a sinusoidal signal is 2x the average power, for example:

    https://i.imgur.com/MBJCEan.png

    "RMS power" and "watts RMS" are incorrect usage, and always have been, unless you're really measuring the RMS of the power waveform (but you're not).
     
    Last edited: May 27, 2014
  16. #12

    Expert

    Nov 30, 2010
    16,343
    6,828
    1)You're right. It isn't a guess, it's a mathematical formula:
    Average voltage equation ≈ 0.636 Vpk

    2)The second quote isn't what I meant. It's what Meyer Sound meant.

    3) Exactly! You shouldn't use the word, "average" when you are pretending it means the same thing as RMS.

    4)Nobody measures the RMS of the power waveform.
    Then why are True RMS meters available from high quality manufacturers?
    Because the averaging meters can not produce a true value for anything except a sine wave.

    5) How old are you and what level of math have you accomplished that you think RMS power has always been an incorrect way to tell how much power a sine wave can produce?
     
  17. wayneh

    Expert

    Sep 9, 2010
    12,153
    3,059
    Right or wrong, these terms have a legal definition in the U.S. established in 1974 by our FTC. They're useful to the degree that you trust the guy quoting it to be obeying the law. It doesn't much matter what term is used if they're not.
     
  18. endolith

    Member

    Jun 21, 2010
    27
    0
    We're talking about power, not voltage.

    Because they're volt meters, not power meters. They measure the RMS of the voltage waveform. They do not measure the RMS of the power waveform.

    The RMS of the voltage is proportional to the average power, not the RMS power.

    Doesn't matter. People with impeccable credentials are still wrong about things all the time. The important thing is what I'm saying, not who's saying it.

    The term "RMS power" is incorrect when you mean "average power". You agree with that, right? Well, when people report power ratings of equipment, they are measuring "average power", not "RMS power".

    When you measure the RMS voltage of a waveform, that RMS voltage is proportional to the average of the power waveform:

    P_{average} = V_{rms} \cdot I_{rms} = \frac{{V_{rms}}^2}{R} = {I_{rms}}^2 \cdot R

    Here's a less abstract example:

    You have a power amp with ±28.3 VDC rails. It can therefore output a maximum sine wave with an RMS voltage of 1/sqrt(2)*28.3 = 20 V. You apply this 20 Vrms to a 4 ohm resistive load. When the sinusoid swings through 0 V, the instantaneous power dissipated in the load is 0 W. When the sinusoid is at its maximum peak of 28.3 Vpeak, the instantaneous power dissipated by the load is 28.3^2/4 = 200 W. (When the sinusoid is at its negative peak of -28.3 Vpeak, the load is also dissipating 200 W. The power waveform is always positive for a resistive load.)

    The average power dissipated in the load over an entire cycle is half of the peak power: V^2/R = 20^2/4 = 100 W. This is the number that people care about and report in specifications; the average power, derived from the RMS voltage.

    It's also possible to measure the RMS of the power waveform, which would be \frac{\sqrt{\frac{3}{2}}}{2} \cdot P_{peak} (=122.5 W in this case), and this would correctly be called "the RMS power", but no one does this. So stop calling it "RMS power" when you mean "average power".
     
    Last edited: May 27, 2014
  19. rogs

    Active Member

    Aug 28, 2009
    279
    37
    You'll have one hell of a job trying to convince people though......there's added support for your cause here:

    http://www.hifi-writer.com/he/misc/rmspower.htm

    if it helps at all?.....
     
    Last edited: May 27, 2014
    endolith likes this.
  20. endolith

    Member

    Jun 21, 2010
    27
    0
    Well they'll just have to be happy being wrong, I guess. This is a mathematical fact, not a matter of opinion.
     
Loading...