john k... said:
I am not advocating adding a series resistance and am focusing on just the low Qts driver + EQ versus high Qts driver trade-off. I am trying to identify what advantage or disadvantage does low Qts + EQ bring to the table if we assume the same mechanical driver construction, meaning different magnets to produce the high and low Qts values, and the same baffle size.
My method has been to use large diameter high Qts woofers and adjust the low frequency response with the baffle style/geometry. Clearly that is different from John K and Linkwitz OB systems were EQ is used to shape the bass response. I am not concluding one is better then the other, I am just trying to put together a list of advantages and disadvantages for each so accurate trade-offs can be made. Maybe the only way to do this is to listen to one next to the other, I don't know.
Hi Martin, Happy to see you here ! I will make my TL soon 😉
I take time to understand this. According to my experiment, a low Qts give you a faster driver. Faster mean for me a faster peak response to a pulse (the beginning to the maximum of the first peak of the impulse response ). For example : QTS=0.3 to QTS=0.6, you divide the time of by 2 : 130uS to 282uS for my 10" driver.
If you make active equalization, it's a more dynamic bass, better bass my opinion than a high QTS driver. And it's easier to marry the woofer with the midrange. I think the group delay is more linear with the low Qts.
If you make a passive equalization, it's better to use the higher QTS driver. I am totally agree with you.
In the midrange, I suppose low displacement, the driver is enough fast. I never notice any problem in particular.
Rodolphe.
jerome69 said:
Yes, this is also my opinion. "fast" may not be the best word to use, but that's what one spontaneously think.
In some setups a higher XO point is desirable (ex. my case) and it is easier to blend well a low Qt woofer.
Lastly, high quality speakers have usually low Qt, esp. in the smaller 12" and 10" size. I took two months to choose suitable woofers for my project and i think (xo in the works still) i have found one.
MJK said:
Hi Martin,
As I said, nothing wrong with using Hi Qts drivers if that works. There are different approaches to eqing each type depending on what the desired response is to be See my page on Gradient woofer EQ. One of the key factors is the final roll off and associated group delay of the woofer. Using a low Q woofer places one pole of the response at very low frequency and that allow maintaining a 2nd order acoustic roll off the the dipole woofer. Frankly, to day I favor a minimum 3rd order roll off since it is the minimum required to prevent the driver excursion for continuing to increase as the frequency drops below the system cut off frequency.
john k... said:
John,
Are you sure that plot is correct? It took me a few days to put my finger on what was bothering me, but as the voice coil inductance rises should the driver's SPL roll-off? I would think that the current through the voice coil would be dropping, assuming a voltage source, just like a first order crossover coil would impact a woofer's SPL output. Maybe I am thick, but something does not look right to me.
Martin,
I think you are right about this simulation. But more appropiate than putting resistors in front of loudspeakers to achieve bassresponse would be to put them into H-frame baffles. That will extend bass response as well as EQ response levels. So if the element is fit with appropiate parameters, like decent X-max and not to low Qts , this would be a good way to go.
/Erling
I think you are right about this simulation. But more appropiate than putting resistors in front of loudspeakers to achieve bassresponse would be to put them into H-frame baffles. That will extend bass response as well as EQ response levels. So if the element is fit with appropiate parameters, like decent X-max and not to low Qts , this would be a good way to go.
/Erling
MJK said:
No, no. You are not thick. The sim omits the effect Le on SPL. That was intensional so that the effect of the series resistance would be isolated from the roll off due to Le. All the curves would start to roll off at higher frequency due to Le if I included it in the SPL. Say around 1k Hz. Same for the dipole result.
The reason I usually omit the effect of Le on SPL is because while Le rolls off the response the effect of directionality tends to boost the response on axis and depending on the design a driver may remain nominally flat on axis well beyond where Le would be expected to roll the driver off.
skorpion said:
Putting the driver in an H frame will not EQ the response. It will shift the dipole peak frequency and change the sensitivity but the 6dB roll off will still be present. Additionally, depending on the relative moving mass of the driver and the air mass in the H frame the driver in the H frame may show a shift in fs and Qts as well. This effect is more significant with drivers with low moving mass. With heavy drivers, like the Peerless XLS and XXLS the air mass load makes little difference.
john k,
No not by itself, but with a combination of chosen crossover frequency and the H-extended bass-response a pretty good EQ could be achieved without any electrical intervation. But at the price of some sensitivity loss.
/Erling
No not by itself, but with a combination of chosen crossover frequency and the H-extended bass-response a pretty good EQ could be achieved without any electrical intervation. But at the price of some sensitivity loss.
/Erling
If you put a driver in an H frame the response is going to roll off nominally 6dB/octave relative to the driver's infinite baffle response. Additionally, the 1/4 wave resonance of the H frame will actually increase the slope of the roll off.
My argument against passive eq is with respect to the frequency dependent sensitivity of the dipole. You can do it passively or actively. The shape of the eq will remain the same but doing it passively throws away head room. Let's say the woofer system has a cut off of 30 Hz and it is desired to use the woofer to 120. The sensitivity of the woofer, due to the dipole roll off, will be on the order of 12dB greater at 120 Hz than at 30 Hz. Why would one want to throw away that 12dB increase in sensitivity using passive eq? Let's say the sensitivity of the woofer at 30 Hz is 80 dB/2.83v. If eq'ed passively and the woofer was asked to reproduce 30 Hz and 120 Hz at 80 dB the amplifier would need a peak voltage swing of +/- 8 volts (V = 4 sin(w1 x T) + 4 sin(w2 x T) will have peaks of +/- 8 volts). The actively eq'ed system would require only +/- 4.25v peak. (Note: for the sake of argument I am assuming the woofer has flat IB response).
My argument against passive eq is with respect to the frequency dependent sensitivity of the dipole. You can do it passively or actively. The shape of the eq will remain the same but doing it passively throws away head room. Let's say the woofer system has a cut off of 30 Hz and it is desired to use the woofer to 120. The sensitivity of the woofer, due to the dipole roll off, will be on the order of 12dB greater at 120 Hz than at 30 Hz. Why would one want to throw away that 12dB increase in sensitivity using passive eq? Let's say the sensitivity of the woofer at 30 Hz is 80 dB/2.83v. If eq'ed passively and the woofer was asked to reproduce 30 Hz and 120 Hz at 80 dB the amplifier would need a peak voltage swing of +/- 8 volts (V = 4 sin(w1 x T) + 4 sin(w2 x T) will have peaks of +/- 8 volts). The actively eq'ed system would require only +/- 4.25v peak. (Note: for the sake of argument I am assuming the woofer has flat IB response).
- Status
- Not open for further replies.