Does someone know what the peaks causes in the typically 2-3Khz region? A time ago i red this is the resonance of the spider and surround but since i have no spider it is incorrect because i have also two peaks. cone material resonance or maybe the coincidental frequency, break-up? This is important to know otherwise you can't eliminate them without a filter.
Little peaks yes but it is audible (I didn't think it was that notable in a test setup)
The region 2-3Khz is very important it has to be flat.
Little peaks yes but it is audible (I didn't think it was that notable in a test setup)
The region 2-3Khz is very important it has to be flat.
Hi Jef, its hard to tell just from impedance plot. Could be very well from the voice coil position.
Try to measure impedance with the voicecoil blocked and see if its a motion problem.
You need to measure frequency response onaxis and from other angles too to check cone related impedances.
Or CSD would help.
Try to measure impedance with the voicecoil blocked and see if its a motion problem.
You need to measure frequency response onaxis and from other angles too to check cone related impedances.
Or CSD would help.
Just ordered a mic so going to to try this next week. thx
Question, If the impedance rises there is less current through the coil so in my opinion less force => less Db's
But in real world it is the opposite in the high freq. range. Where am i wrong? In the Fs range the mechanical resonance will make the cone will move harder so the eddy currents will also go higher and these are opposite to the signal so the resistance rises also.
Is there someone who can explain the high suppression of the low resonance freq. of my speaker (see some posts back for the plot)
What i did: I have one voicecoil winding but double wounded in parallel to get 4ohm and to have the height of the winding as low as possible.
If i connect one winding to the amp and the other shorted i get that result.
Oke it act in some way like a shorten ring (you can see in the upper freq. range) but Fs that get so low impedance i don't get it. First i thought it is the extra damping (like if you use a alu former) it is the effect of having the former in a sort of thick oil you can feel it very well when moving the former by hand. But if i shorten the second winding i don't feel any difference so it is not damping (it goes very loose). Does someone have some explonation on this?
Thx,
Hi Jeff,
if impedance rises then less current goes through voice coil then less dBs is not quite true. Only if you consider the electrical side but if you add mechanical and acoustical side it really depends on what freq region you are. In Fs range cone moves extremely easy, that is why many cones are broken at and below resonance.
About the second voice coil in parallel, Fs doesn't change. Only Qes changes. Please revisit my post here:
http://www.diyaudio.com/forums/mult...iy-field-coil-loudspeaker-26.html#post3512249
if impedance rises then less current goes through voice coil then less dBs is not quite true. Only if you consider the electrical side but if you add mechanical and acoustical side it really depends on what freq region you are. In Fs range cone moves extremely easy, that is why many cones are broken at and below resonance.
About the second voice coil in parallel, Fs doesn't change. Only Qes changes. Please revisit my post here:
http://www.diyaudio.com/forums/mult...iy-field-coil-loudspeaker-26.html#post3512249
Hi Jef,
IMO it does not make a difference - given the amplifier has very small output impedance, which usually holds with most solid state designs - due to Qes, whether you connect
- both voice coils in parallel to the amplifier's terminals or
- one coil to the amplifier's terminals and shortcut the other one.
The back EMF dampening the membrane's motion if released from a certain excursion will be the same:
The amp ideally acts like a shortcut at the connected coil, and the other coil is (actually) shortcut too. Effectively you have both coils shortcut now when watching the membrane settling.
A (measurable, tangible) difference in back EMF would occur, if you compare the 2nd coil open circuited vs. shortcut, while keeping the 1st coil connected to the amp in both cases.
Just the voltage sensitivity is lower, when the 2nd coil is shortcut instead of being (in parallel) connected to the amplifier, as only half of the driving force B x L x I applies to the voice coil due to impedance at the amplifiers terminals doubling and current I being halved.
B.t.w also conecting both coils in series to the amplifier will result in the same Qes. This is basically because L x I - which accounts for the back EMF - keeps the same when you move the membrane from outside to act like a generator.
Last edited:
Here is the new measurement Le at 1KHz. I do use a shorting ring in the magnetic system.
The rise at 250Hz is due to the spider I can remove it just by placing my finger very lightly on it, the rise at 1 and 2.5 KHz seems to be coursed by the surroundings... but I am not 100% sure on that one.
Jef where did you buy your surrondings, I am looking for some that is softer than the ones I use now.
The rise at 250Hz is due to the spider I can remove it just by placing my finger very lightly on it, the rise at 1 and 2.5 KHz seems to be coursed by the surroundings... but I am not 100% sure on that one.
Jef where did you buy your surrondings, I am looking for some that is softer than the ones I use now.
Attachments
Thx LineArray for the explonation.
Frank, here you can find my surround.
Alles om uw luidsprekers te repareren, foamranden, rubber randen, membranen, tweeters enzovoorts
first i got a foam one of stiffness class 1 but it was to stiff.
Now i have ordered a rubber surround of class 1 and those is very loose. (the measurement is with the rubber surround)
Frank, here you can find my surround.
Alles om uw luidsprekers te repareren, foamranden, rubber randen, membranen, tweeters enzovoorts
first i got a foam one of stiffness class 1 but it was to stiff.
Now i have ordered a rubber surround of class 1 and those is very loose. (the measurement is with the rubber surround)
Hi Hentai,
maybe, but could you point out which of the statements in
http://www.diyaudio.com/forums/mult...iy-field-coil-loudspeaker-53.html#post3856192
you think are wrong ?
Where is an Re influence changing a certain statement ?
Last edited:
Hi Oliver,
when you say its same having both coils in parallel and connected to amp with having one coil parallel and the other shorted.
I mean, yes from the voice coil's terminals perspective it is true. It will see low impedance there. But from system's perspective its not the same as impedance curves show clearly. And this is because Re changes. Its like you will have 1 voice coil with thicker wire.
when you say its same having both coils in parallel and connected to amp with having one coil parallel and the other shorted.
I mean, yes from the voice coil's terminals perspective it is true. It will see low impedance there. But from system's perspective its not the same as impedance curves show clearly. And this is because Re changes. Its like you will have 1 voice coil with thicker wire.
i think if you connect both windings to amp you don't have the impedance correction on the higher frequenties (like a shorten ring) so if you shorten the second winding you get a more flat response over the whole range. Did some listening tests today and this is audible. I'm going to try to make a 3 way system where the sub is connected both windings with a mid that is flat as possible (second winding shorted) and a tweeter. A normal A-class amp will have 60-100ohm output impedance so shorted is 10 times less. Correct me if i'm wrong but that is a great difference.
No not yet I am making new cones and voice coils. I forgot that the voice coil former had to bee wider, when I started to winde inside and outside..It was to tight a fitt to my taste.
By the way Le seems wrong @ 1KHz, 0.628mH translate to 78.5 ohm @ 20KHz??
yes this is an effect very specific to loudspeaker inductance. Some call it a semi-inductance as it doesnt rise as a normal inductor. There are a few models that take this effect in consideration. Some add a parallel resistor to Le some add another inductor with a parallel resistor all in series with Le and Re.
Hi Hentai,
sure, maybe i was too sloppy in not mentioning that. Naturally even voltage sensitivity changes. But you would agree, that damping by EMF should be the same, when the driver is settling from a specific excursion ?
I was looking at the driver's terminal perspective and the damping by EMF, since "no change in damping" was Jef's observation, when moving the membrane.
Hi Jef,
yes at higher frequencies there must be a difference, strange i was thinking of that this morning ...
At frequencies getting higher the connected coil and the shortened 2nd coil make up a transformer having the secondary shortened. The magnetic field built up by the connected coil is reduced, thus Lvc is reduced.
It seems, a double voice coil has many options in how to use it, in order to influence the performance of a driver.
But i guess shortening rings at the voice coil and a copper cap at the magnet's center pole should be more effective in reducing Lvc and also cause less mass to be moved with the voice coil. From a moved mass point of view, a shortened 2nd voice coil has too much copper to carry around for that particular job to do IMO.
Last edited:
I have now tested Hentai's and my spiders, with all of mine I got some peaks and bumps Hentai's had smaller peaks and bumps but the compliance was to low. I then tested some “standard” spiders and it was very flat but it to had a low compliance, then I soften it up by slitting some tracks in it and that worked. So my choice is now the modified “standard” spiders.
I also used some softer foam surroundings for this test.(thank for the link Jef)
Here is the result with the modified “standard” spider.
I also used some softer foam surroundings for this test.(thank for the link Jef)
Here is the result with the modified “standard” spider.
