Please explain a speaker’s "impedance curve" | The Canadian Guitar Forum

Please explain a speaker’s "impedance curve"

Discussion in 'Amp Building/Technical/Repair' started by greco, Apr 16, 2018.

  1. greco

    greco Gold Member

    Joined:
    Jul 15, 2007
    Location:
    Kitchener, Ontario
    This is a graph for an Eminence Lil' Buddy - 8 ohm - Guitar Speaker.
    Here are some specs for this speaker:

    SPECIFICATIONS

    Nominal Impedance: 8 ohms
    Power: 50w rms
    Freq. Response: 80.0 Hz- 5.0 kHz
    Sensitivity: 99.1 db 1w/1m
    T/S PARAMETERS
    Fs: 149.0 Hz
    Re: 6.31 ohms
    Le: 0.46 mH
    Qms: 9.06
    Qes: 0.93
    Qts: 0.84
    Vas: 10.0 L, 0.35 cuft
    Vd: 29.0 cc
    Cms: 0.05 mm/N
    BL: 11.7 Tm
    Mms: 22.0 grams
    EBP: 160.0
    Xmax: 0.8 mm
    Sd: 366.1 cm2

    Could someone please explain how to interpret the orange line (the speaker’s impedance curve) and how (or if) it could be useful when purchasing a speaker.
    (I understand the red line)
    [​IMG]

    I found this article interesting and helpful...
    Speaker Geeks: Why Graphs Matter and When They Don’t | Premier Guitar

    Thanks in advance.

    Cheers

    Dave
     
  2. GTmaker

    GTmaker Gold Member

    Joined:
    Apr 24, 2006
    Location:
    Guelph, Ontario
    greco likes this.
  3. High/Deaf

    High/Deaf Gold Member

    Joined:
    Aug 19, 2009
    Location:
    LM,BC,CAN
    The fact that a speaker has a rated impedance (reactive, varies with frequency) and not a resistance (resistive, the same at all frequencies) is what the graph is getting at.

    The two are similar in that they, through Ohm's Law, establish a current / voltage relationship. If you apply 'V' voltage across a certain resistance/impedance, it will draw 'I' current (I=V/R or I=V/Z). Unlike resistance (R), impedance (Z) isn't the same for all frequencies, it varies as the frequency does.

    If you were to sweep that speaker with a tone generator from 70 Hz to 5 kHz (a common guitar's bandwidth), it would present a different load (so different current draw for the same voltage) at different frequencies. In the graph you posted, it would vary from 8 ohms at 70 Hz to about 80 ohms at 150 Hz, to a low of 7.5 ohms between 250 and 530 Hz, and back up to just over 20 ohms at 5 kHz. For simplification, they are rated at the lowest, or close to lowest, point in the graph, probably 8 ohms in this case.

    Drivers have all sorts of different impedance curves. That's one of the things that makes attenuators, reampers and reactive loads difficult. They are hardware and can't model every impedance curve, they have to generalize. But reactive is still better than resistive, it comes much closer to modelling an actual impedance curve.
     
    nonreverb, greco and WCGill like this.
  4. greco

    greco Gold Member

    Joined:
    Jul 15, 2007
    Location:
    Kitchener, Ontario
    @GTmaker and @High/Deaf ..Thanks again. Very helpful.

    I read the article I linked more carefully and thoroughly this morning. This is about the practical application of the graphs..(bolded sections done by me)

    It may surprise you to learn that many amplifier manufacturers don’t put a lot of initial stock in these graphs when deciding which speaker to use for a given amp design. While they’re very useful in making a final decision between two candidates with very similar signatures, the real work takes place long before we ever get to the nitpicking graph comparison stage. That means we, the amp manufacturers, have made the most critical decisions about sound, application, and overall performance much earlier than that. Once we get close to final production design, we may pay a little more attention to the graphs to fine-tune the amp/speaker/cab relationship in ways that the player would be wise to think about before replacing the stock speaker with another unit.

    What a speaker graph shows is that most guitar speakers perform similarly on paper, and all bets are off when you install them in your amplifier. With that in mind, the first step in making a speaker purchase should be to establish your desired outcome. Then look at the manufacturer’s recommendations about the models you feel line up with your goals and fit your budget. Forums can be a valuable resource, but try to avoid getting into the weeds when soliciting opinions online by carefully vetting responses that address your specific application.

    After you’ve narrowed your search, it’s time to have a look at those graphs. Pay particular attention to the more prominent peaks they reveal. Those are the areas that are most subject to operational variables like output impedance and cabinet resonance. The higher the impedance, the greater the impact those peaks will have on your amplifier’s performance and, consequently, the more they will drive cab resonance. And since internal feedback (more on this in a future column), as found in most power amp designs, plays a role in output impedance. A power amp designed with low or no feedback—such as in AC30-style amps—will exhibit higher effective output impedance and further exaggerate those curvalicious peaks!
     
  5. cboutilier

    cboutilier

    Joined:
    Jan 12, 2016
    Location:
    Halifax, NS
    Impedence changes with freqency, and that graph shows you exactly how that one does it. I don't think I'd ever consider looking it for a guitar amp. That response curve might be important in the HiFi world, but I'm not sure.
     
  6. mhammer

    mhammer

    Joined:
    Nov 30, 2007
    Location:
    Ottawa, Ontario
    Thanks for that link. Very informative.

    Speakers will normally have an identified "resonant frequency". As I understand it, that frequency,, or range, will be noticeable, though how contrasting with the rest of the frequencies will likely vary. It is a joint function of several things: the specific impedance at this or that frequency, the mechanical properties of the speaker assembly (e.g., cone stiffness, ribbing, etc.), and the idealized cabinet. In other words, the properties of the voice coil itself (which is what impedance measures) will be realized to the extent that the physical conditions permit. The information about resonant peaks can also be used to tailor equalization, although that also necessitates having information about the room/space being listened in as well.
     
    nonreverb and greco like this.
  7. High/Deaf

    High/Deaf Gold Member

    Joined:
    Aug 19, 2009
    Location:
    LM,BC,CAN
    Yes, that's the way I see it, too.

    Guitar cabs are far from leading edge. They're basically a box that's big enough to fit the amp and/or speakers, while being small enough to be easy to transport and not be a hassle position it.

    Hifi designs take speaker box design much more seriously. They have a goal of trying to create a flat response - guitar cabs do not. They also just about always are tying at least 2 devices together via a crossover, and the crossover design takes the driver's impedance curves into account, because the drivers, crossover and box all work as a system.
     
    greco likes this.
  8. greco

    greco Gold Member

    Joined:
    Jul 15, 2007
    Location:
    Kitchener, Ontario
    I totally agree...but we just might want to suit up, just to be prepared...
    BTW...I'd prefer the sporty, slimmer fit of the one on the left.
    [​IMG]
     
  9. Granny Gremlin

    Granny Gremlin Gold Member

    Joined:
    Jun 3, 2016
    Location:
    Toronto - Riverside
    The important part(s) of the curve are the 2 peaks - these are the resonant frequencies ('free air resonance). The one in the bass region tells you how low a sound the speaker can usefully reproduce (depends a bit on cab type, but assuming not anything more advanced than sealed, open or ported, this will be the low cuttoff, if not a bit higher, e.g. in the case of open). The rise in the treble end explains the treble response rolloff. Basically the higher the Z, the lower the speaker efficiency at that freq.

    No, this is not what it shows at all; quite the contrary unless you're a simpleton (they are calling you simple FYI). Unless by "perfom similarly on paper" they mean (in a reductionist and oversimplified sense) that the Z curves are similarly shaped - that's like saying all people have the same basic shape so we can all perform similarly (e.g at sports). I mean of course they all have the same basic shape - they are all essentially the same type of device - a transducer with a cone attached. Yes all speaker Z curves will have upper and lower peaks, but they will not be of the same slope, max amplitude or in the same place; also there may be some nonlinearities, such as the weird speed bump at 2K the graph above (note how that is reflected in the F Response) - the devil is in the details as regards the low end resonance and the high end rolloff. I agree in the sense that response graphs (which are what most people look at) are of limited/overblown value (but still some - the above is not a 'typical ' response for a vintage style guitar speaker - those usually have a much larger rise in output in the upper mids). I can look at an F Resp graph and tell you what sort of application I expect the speaker to be useful in. I can sometimes get a vague idea of the tonality of the speaker, but it is not an exact science, rather educated guessing - there are some complications introduced by the various measurement methods used by different testers.

    What would be useful (especially for finished cabs vs raw drivers, but still) is square wave response graphs. What those tell you is how fast/sluggish the speaker is (dynamic response) and (especially important in the case of multiway cabs) how well in phase the speakers in the cab are with each other (time, alignmnet etc).
     
    Last edited: Apr 17, 2018
    greco likes this.
  10. greco

    greco Gold Member

    Joined:
    Jul 15, 2007
    Location:
    Kitchener, Ontario
    This is a square wave response comparison of 4 headphones as an example.
    Which of the 4 has the best graph and why?
    These measurements are graphed for 300 Hz. How many different frequencies would typically be measured and graphed?

    [​IMG]
    Do you have a sample of a square wave response graph for a guitar (or any) speaker?
     
    Last edited: Apr 17, 2018
  11. Granny Gremlin

    Granny Gremlin Gold Member

    Joined:
    Jun 3, 2016
    Location:
    Toronto - Riverside
    Not handy no - most makers don't provide those - you can google for them, but they are most often done by third parties, and not all speakers (not even close) have been tested. I can't tell you which headphone's response is best unless I am provided with the source wave to compare to which is the whole point. A proper square wave test will provide both the speakers measured output as well as the source signal.
     
    greco likes this.
  12. jbealsmusic

    jbealsmusic Occasional CEO Dealer

    Joined:
    Feb 11, 2014
    Location:
    Nepean, ON
    Came in to say some things, but apparently those things have already been said. Learned some new things by reading through the thread. I love this forum!
     
    loudtubeamps and cboutilier like this.
  13. cboutilier

    cboutilier

    Joined:
    Jan 12, 2016
    Location:
    Halifax, NS
    For me, the beauty of electric guitar tone is hidden in the imperfections. Thus, I care little for specs and rules. I just try shit, and see what it sounds like.
     
    Wardo, CathodeRay and jb welder like this.
  14. GussKolb

    GussKolb

    Joined:
    Mar 26, 2020
    Location:
    Virginia, USA
    Hi....The impedance curves are very important because:
    - You read with their resonant frequency speaker,
    - Selects the filter cutoff frequency,
    - Adjusted increase in impedance due to inductance coil speaker for increasing frequency,
    The more wavy characteristic impedance of the more difficult to match the appropriate amplifier. It is commonly said that the columns are difficult to drive.
    Have a nice day
     
  15. Eric Reesor

    Eric Reesor

    Joined:
    Jan 26, 2020
    Location:
    Victoria BC
    Good thread. I have always considered creating a guitar shaped tone wood cabinet driven by speakers and seeing what it can sound like. If you have pure sine wave sound created by a midi interface with no character it sounds flute like. It is the resonant characteristics of that which a speaker drives that influences the colour of the sound. No strings just a hopped up speaker cabinet that deliberately colours the sound. It could even be driven by a midi guitar, which I hate but the natural sounds created by wood boxes is very much my schtick.

    There is nothing to compare with playing a good acoustic guitar in a room filled with good acoustic guitars all in tune. That simple trick is a staple in the sale of high end guitars. ;) Loud speaker sound characteristics is a very diverse and interesting field of acoustics. Not as interesting as the long storied history of luthiership with the guitar as we know it today but certainly a field of considerable unexplored possibilities.
     

Share This Page