After finally putting a measurement setup for Speaker Workshop together, the result surprises me. Please look at the graphs below.
The Vifa TC11SG49-08 is surposed to have a resonance frequency in free air of 60hz. I measured 74-75hz:
And the simulated response in a 3.5L ported box using Vifas specs showed gave me a Fb of approx 68hz. The measurements show an Fb of 57hz!
Vifa's response graph and specs can be found here.
Can anyone confirm my measurements?
The Vifa TC11SG49-08 is surposed to have a resonance frequency in free air of 60hz. I measured 74-75hz:
An externally hosted image should be here but it was not working when we last tested it.
And the simulated response in a 3.5L ported box using Vifas specs showed gave me a Fb of approx 68hz. The measurements show an Fb of 57hz!
An externally hosted image should be here but it was not working when we last tested it.
Vifa's response graph and specs can be found here.
Can anyone confirm my measurements?
Ahh, so here were the curves that I called for in the other thread.
Fb is the position of the "valley" in the impedance curve. It is *a bit* hard to read the frequency from the graph, but you may be right. Now Fb is *not* affected by the driver. It is only affected by the box helmholtz resonance. So, you are left with two possibilities, the port dimensions are wrong or the acoustic volume of the box is wrong. You say that Vifa has suggested the dimensions. Is the port circular/rectangular? This may change the amount of co-oscillating air a bit. Do you use stuffing inside the box? That can make the box appear larger. Could you describe these details, and possibly also a photo?
The high Fs is suspicious, though. I'd suspect the drivers are faulty in some way, unless you have problems in your measurement setup. Are the impedances measured with sine sweeps? Could they be out of sync in some way?
Fb is the position of the "valley" in the impedance curve. It is *a bit* hard to read the frequency from the graph, but you may be right. Now Fb is *not* affected by the driver. It is only affected by the box helmholtz resonance. So, you are left with two possibilities, the port dimensions are wrong or the acoustic volume of the box is wrong. You say that Vifa has suggested the dimensions. Is the port circular/rectangular? This may change the amount of co-oscillating air a bit. Do you use stuffing inside the box? That can make the box appear larger. Could you describe these details, and possibly also a photo?
The high Fs is suspicious, though. I'd suspect the drivers are faulty in some way, unless you have problems in your measurement setup. Are the impedances measured with sine sweeps? Could they be out of sync in some way?
Two things about the resonant frequency: first, manufacturer's specs are often not representative. Second, a lot of things can shift fs around, like mechanical break-in and temperature. And I assume you have the driver well clamped?
If you haven't broken in the driver yet, run a low frequency sine wave through it at a sufficient level to cause excursion through around half of Xmax. Leave it running overnight. Let the driver relax for a while, then remeasure.
Finally, the absolute value of fs isn't as important as the interrelationship with the other T-S parameters.
If you haven't broken in the driver yet, run a low frequency sine wave through it at a sufficient level to cause excursion through around half of Xmax. Leave it running overnight. Let the driver relax for a while, then remeasure.
Finally, the absolute value of fs isn't as important as the interrelationship with the other T-S parameters.
Well, that will depend on what value you get for Qts and Vas, too. As long as fs/Qts isn't too different than spec, nor is fs2Vas, you can still work with the driver.
In Dickason's LDC, he gives the example of two nominally identical drivers with an fs spread even bigger (relatively), driver 1 having fs = 31.5, Qts = 0.45, Vas = 2.97 cu ft; driver 2 having fs = 38 Hz, Qts = 0.54, Vas = 1.92 cu ft. For this example, with a target Qtc = 1.0, the optimum box volume is 0.75 cu ft for driver 1 and 0.79 cu ft for driver 2- not very different. More importantly, the spread in fc is minimal, 69.9 and 70.4 Hz respectively.
You can plug these parameters into the vented box alignment tables and come up with similarly similar results. For example, in a BB4 alignment with QL = 7, driver 1 gives an f3 of 31.3 Hz, driver 2 gives an f3 of 33 Hz. Not too different, I'd say.
So, yes, that variation in fs may be too great, but it may not be, depending on the other measured T-S parameters and the willingness to adjust box volume and tuning. And that difference may be illusory if the driver was not broken in or clamped down well.
In Dickason's LDC, he gives the example of two nominally identical drivers with an fs spread even bigger (relatively), driver 1 having fs = 31.5, Qts = 0.45, Vas = 2.97 cu ft; driver 2 having fs = 38 Hz, Qts = 0.54, Vas = 1.92 cu ft. For this example, with a target Qtc = 1.0, the optimum box volume is 0.75 cu ft for driver 1 and 0.79 cu ft for driver 2- not very different. More importantly, the spread in fc is minimal, 69.9 and 70.4 Hz respectively.
You can plug these parameters into the vented box alignment tables and come up with similarly similar results. For example, in a BB4 alignment with QL = 7, driver 1 gives an f3 of 31.3 Hz, driver 2 gives an f3 of 33 Hz. Not too different, I'd say.
So, yes, that variation in fs may be too great, but it may not be, depending on the other measured T-S parameters and the willingness to adjust box volume and tuning. And that difference may be illusory if the driver was not broken in or clamped down well.
Svante: The measured Fb is not the thing that bugs me the most. The boxes may be slighty larger than originally simulated due to several things, such as no crossover taking up space and a fairly big margin. The port is probably a bit smaller than it's supposed to be. It's not a problem - I can always adjust the port length. The box is stuffed, but I was of the impression that stuffing in a vented box doesn't alter the aparent size significantly.
No problem here, really.
Svante and SY: The free air response still puzzles me. The setup is rather improvised but no matter how the driver is mounted - downfiring, upfiring etc. - or not mouted at all, the impedance varies within a couple of hz at the resonance point. I don't think clamping the driver will change the result as much as 14hz considering what written above.
I cannot post pictures of the box/setup, sorry. I don't have a digital camera.
EDIT: BTW, the drivers are breaked in. They speakers have been used for a couple of months, but I have not until now had the opertunity to measure the drivers' electrical properties nor the acoustics response.
No problem here, really.
Svante and SY: The free air response still puzzles me. The setup is rather improvised but no matter how the driver is mounted - downfiring, upfiring etc. - or not mouted at all, the impedance varies within a couple of hz at the resonance point. I don't think clamping the driver will change the result as much as 14hz considering what written above.
I cannot post pictures of the box/setup, sorry. I don't have a digital camera.
EDIT: BTW, the drivers are breaked in. They speakers have been used for a couple of months, but I have not until now had the opertunity to measure the drivers' electrical properties nor the acoustics response.
Thanls for clarifying. The one point I harp on again and again is that one must always measure the actual drivers on hand and not rely on datasheets. Your experience here is a shining example of why my obsession is not unreasonable.
Nonetheless, I'd still move on and get Vas and Qts to see if the driver will still give the same results as the spec sheet driver for your chosen alignment.
Nonetheless, I'd still move on and get Vas and Qts to see if the driver will still give the same results as the spec sheet driver for your chosen alignment.
Yes, and I will do that shortly.... whenever I get around to installing the usual office stuff.... just reinstalled WindowsXP... Claudio Negro has a very nice spread sheet for calculating driver parameters based on electrical measurements in Speaker Workshop.
Thank you both for your help.
Thank you both for your help.
Effects of absorving mat
Hello,
here http://home.hccnet.nl/ine.dick/p9.html you can see some results of the influence of the amount of absorption in the box, closed and vented.
Claudio
Hello,
here http://home.hccnet.nl/ine.dick/p9.html you can see some results of the influence of the amount of absorption in the box, closed and vented.
Claudio
- Status
- Not open for further replies.