Can an experienced speaker builder tell a driver Qts just by looking at it?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.

PRR

Member
Joined 2003
Paid Member
If you know how (what box) it worked good in, you can guess.

Bare naked driver, you can't guess Qts unless you short the voice coil. The "ring time" of the bass resonance directly gives the Q. HOWEVER it is typically VERY short and VERY low frequency. Even A/B-ing two similar drivers it is hard to judge.

I used to try to do this with bulk surplus speakers. There's better ways to waste your time.
 
PRR - "short the voice coil" - I assume same as an amplifier with very low output impedance.

OK so you cant just guess very well at it, even looking at the speaker in hand. What if you have an arbitrary waveform function generator and oscilloscope to use?

I'd assume then you could hit the speaker with some voltage pulse - maybe a half-sine at the speaker Fs - and watch the current ring on the scope?

I assume the various Qts values correspond to the common "underdamped", "critically-damped", etc systems. I thought I read a Qts of 0.5 means "critically-damped", so I assume the higher values would imply ringing. Which should be observable on the scope after a good strike.

I recently switched amps, to my little Jolida chip based with the tube front end. It makes a slight pop at turn off, only really heard because the Lii F15s are so sensitive. Being just behind the speaker as I reach for the power switch, I can definitely hear the "Toong" resonance of the cone. Their Qts is 0.7. (I suppose a speaker with a 1.0 value would really ring when fed such a pulse) I realize that because the amp is on its way off, probably not ringing into a short, but an open. I'm sure that extends the ring time from what it otherwise would be.

I realize having a programmable function generator and oscilloscope lends itself well to just measuring Qts in the normal way. I realize the time domain measure - such as counting cycles - inst going to be as accurate as a method using a continuous frequency. But I think it's far more intuitive as to whats actually going on with these Q values, especially after reading something like "the Qts is so high, the speaker is its own enclosure - no baffle necessary". I wont begin to comprehend how that could be, without a more empirical understanding of how speakers with the various values behave mechanically. Same with the cabinet designs that correspond to different speaker Q values. Or people saying they dont particularly like the sound of high Qts woofers...

Thanks!
 
Last edited:
Hi Joe,


the most elegant step away from guesswork is DATS. It cost´s you a little bit more than a nice meal for two with carefully selected wines and will be your friend for many years, you´ll learn a lot, you´ll discuss with other friends, you´ll develop your loudspeaker skills... it will change your attitude and will make you look better.
Ok ok, it´s just a very valuable tool, don´t underestimate it. Highly recommended.


All the best


Mattes
 

Attachments

  • Xizi Fast 5 No. 1.doc
    132 KB · Views: 24
PRR - "short the voice coil" - I assume same as an amplifier with very low output impedance.!

No, you actually short the voice coil using e.g. a piece of wire between the terminals. There is no other electrical equipment connected or involved, but rather you strike the cone with your finger to generate an "impulse" and then you listen with your ears close to the cone to try and discern how much ringing there is and at what frequency. It's better than nothing at all, but highly empirical at best.
 
Hi Lojzek,


yes, you´re right - if you have the right soundcard. I have a Steinberg UR12, couldn´t make it work for TSP, so had to decide whether to buy a new card or a DATS. Many many other solutions out there as well, but for people not so keen on computer measurements (like me) the DATS is a good choice, as it´s easy to work with. YMMV of course...


All the best


Mattes
 
The problem with the battery click test is the convenient VC shorting requirement. With a decent SS amp connected and driving it with a "click" signal, that would take care of the short requirement.

Why the short? Why didnt "they" define it with a 4 / 8 ohm resistor across - like, same as what the speaker impedance is supposed to be?

A current transformer could observe the signal in the short, after a fingertap. 'Luck finding such a probe flat through 10 Hz. A resistor would be so much easier to get; maybe 0.1 ohm, lots of scope channel gain and call it "close enough" to a short.
 
Last edited:
Administrator
Joined 2004
Paid Member
The click test would not use a shorted voice coil. I suppose you could put an 8 ohm resistor across it to damp it some. But either way, it's not the best test. A woofer tester is. It's the best thing to use if you want an accurate Qts number.
 

PRR

Member
Joined 2003
Paid Member
> Why didnt "they" define it with a 4 / 8 ohm resistor across - like, same as what the speaker impedance is supposed to be?

Because it complicates the math for no good reason.

Because an "8 Ohm" speaker is NOT 8 Ohms, and particularly at bass resonance.

Qts alone is not an interesting number. Fs/Qts is interesting. Qts with Vas is interesting.
 
I still have to believe the short across the speaker terminals most closely emulates a SS amplifier (with high damping factor, i.e. very low impedance). That's what would happen during some low frequency transient followed by silence - the speaker cone would still be in motion with the amplifier output @ 0V - a virtual short. But not all amplifiers are SS with very low output impedances; they used to have a "damping factor" control which changed the amplifiers output impedance (I believe by changing the feedback)

I've read GM say you can increase Qts by placing a resistor in series with the amplifier output. Makes perfect sense to me, if this Qts value represents "overdamped, critically damped (the 'ol 0.5 value...) and underdamped. So the cone in motion has a less solid "load" for the VC motor to "push against" with the additional resistor, so it'll flop about a little more than loaded into a dead short. Or, move a bit more toward an underdamped system.

Ever see one of those neodymium ball magnets fall slowly through a copper tube? That's how I'd expect a speaker cone to react given a force displacement with the VC shorted. (Exaggerating a bit there, perhaps if the VC was just a single turn of copper foil... ) It seems the prescribed way gives the speaker its best chance to ring at little as possible.

Connect it to a different amplifier type - or place a resistor in series - anything that upps the impedance the speaker sees back into the amp - will loosen the reins on it and allow it to ring a bit more, a little less damped; higher Qts. The old damping factor control allowed a user to adjust this. (I always wondered what that was for!)

I wonder what happens to Qts - defined into a VC short - when you connect the speaker to one of the LM1875 current source output topology amplifiers. I'd imagine in that system, it's nowhere near the same value, actually, anymore. So maybe a cross between voltage / current source, i.e. a definable output impedance via a "damping" control would give "Want the speaker to have a little more Qts? Just turn the dial on the back of the amp!" the user some control over this important driver parameter.

Sorry for the long reply, just trying to work out what Qts means physically and how the shorted VC aspect of the measure relates to the driving amplifiers output electrical characteristics. If the speaker enclosure is a resonant system reinforcing a speaker's Fs, i can begin to understand a speaker with a high Qts - lots of ring to it at Fs - is its own enclosure. Perhaps just not a very efficient one...
 

PRR

Member
Joined 2003
Paid Member
Qes on a high-Z source is obviously infinite.

Then Qt is limited by mechanical losses. (And acoustic radiation but this tends to be "none" for practical purpose.)

At the time the papers were new, "all" amplifiers were zero-Z and Qts was usually dominated by Qes. If you want to violate these assumptions you have to step-back to less-simple math.
 
It seems the quest for a zero-Z amplifier output has been with us since well before I was born.

With that achieved, the math is simplified and everyone's on the same page so to speak? Versus you cant tell what's going to happen - depends on the amplifier design the speaker happens to be connected to?
 

Attachments

  • Low-Z-Amp.jpg
    Low-Z-Amp.jpg
    265.7 KB · Views: 38
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.