Dennis,
I understand what you are saying and your last set of plots make sense. But your first set of plots do not make sense. The overlayed SPL curves cannot be produced by just switching magnets and then applying EQ, it is an apples and oranges comparison.
I understand what you are saying and your last set of plots make sense. But your first set of plots do not make sense. The overlayed SPL curves cannot be produced by just switching magnets and then applying EQ, it is an apples and oranges comparison.
But that's exactly how I made those curves. The only difference between the (important) T/S parameters of the drivers is one had twice the Bl (1/4 the Qes) of the other.
Qes = 2*pi*Fs*Mms*Re/Bl^2
I used a Linkwitz transform to increase the Q to match the other driver and adjusted the power until the SPL was about the same. I could have used a simple parametric band filter to raise the Q since Fs wasn't changing but the LT calculated the right value for it automatically.
Something does not look right to me, the SPL and power curves in your first post just don't make sense.
High Qts has an advantage of less EQ if your XO is passive. However, in such low frequencies there are a lot to compromise if doing it passively.
With active EQ/XO, as far as design complexity is concerned, it does not matter if it is higher Qts or lower Qts.
Note that both S. Linkwitz and John K use low Qts drivers for H/U-frames down to 40Hz and they have great success. Below 40-50Hz I would use monopoles.
Regards,
Bill
With active EQ/XO, as far as design complexity is concerned, it does not matter if it is higher Qts or lower Qts.
Note that both S. Linkwitz and John K use low Qts drivers for H/U-frames down to 40Hz and they have great success. Below 40-50Hz I would use monopoles.
Regards,
Bill
I don't know what to say, I only posted what the sims show and it makes sense to me. 🙂 The low-Q (twice the magnet strength) driver is 6dB more efficient at very high and very low frequencies so the nominal voltage before EQ is reduced 6dB. The EQ needed to turn Qes=.35 into Qes=1.4 is about 10dB at Fs so you're supplying about 4dB net more voltage to the low-Q driver at Fs. (Edit: and you can see that on the second set of curves -- it needs about 4dB boost at 38Hz.) But the low-Q driver's impedance is much higher at Fs so that increased voltage doesn't result in more power (volts x amps) from the amp. At very high and low frequencies, the low-Q driver is only using 1/4 the power (-6dB).
Last edited:
Looking at the drivers you have simmed
Fs around 40hz ?
What kind of woofers is that ?
Your low Q woofers shows what looks like peaking around 40hz, I wonder why ?
Looks much like a poor high Q BR tuning with sloppy bass
If so, I wouldnt use as reference
Maybe a peaked bass is just exstra output fore free, but I would want that to happen
Fs around 40hz ?
What kind of woofers is that ?
Your low Q woofers shows what looks like peaking around 40hz, I wonder why ?
Looks much like a poor high Q BR tuning with sloppy bass
If so, I wouldnt use as reference
Maybe a peaked bass is just exstra output fore free, but I would want that to happen
Last edited:
You've got it backwards. The yellow lines with the peaking are the Goldwood Twister linked with Qes=1.4 (Qts a bit less). The one with the gradual roll-off is the same driver with twice the magnet strength (Qes=.35). They are both modeled on an infinite baffle (actually a sealed 10000 liter box with Qa=10000 for those who want to try it in WinISD Pro) so there are no box effects modeled.
That peaking you noticed is what the high-Q guys advocate to avoid dipole EQ. I'm just trying to show that you can do the same thing, only better, with a low-Q driver plus EQ if that's a response that floats your boat.
Ahh, I see, thanks
Then I dont understand your EQing
Seems you need a little more EQ on the low Q woofers, and a little less or different EQ on the high Q woofer
Maybe it will change the difference a bit, maybe not that much, but still
Anyway, why use fixed linkwitz grainer EQ?
I thought todays standard and future was digital😛
I would like to see what happens if you shaped the response a bit more
Last edited:
I wasn't trying to shape the response for an OB. Just showing that you can match the response of a high-Q driver with a low-Q driver plus EQ. With an OB, you'd measure the real response of the driver on the baffle and go from there with the EQ to get things flat. That's what the OP (Steve) did with 'Bob' and his digital EQ.
http://www.doddsy.net/steve6_009.htm
There is one advantage with using higher Q drivers and a digital EQ such as the DCX2496. The DCX likes to have the input as hot as possible, and this combined with the amount of gain needed to EQ the woofers can cause the output gain to overload.
Driver Q is not relevant. What's important is the target Q and Fs of the system. LT and SLP can shape the response.
However High-Q drivers are generally less expensive due to smaller magnets. Hence I like drivers with Q around 0.5 to 0.7, compared to the heavy and large 0.2-0.3 Q drivers.
However High-Q drivers are generally less expensive due to smaller magnets. Hence I like drivers with Q around 0.5 to 0.7, compared to the heavy and large 0.2-0.3 Q drivers.
- Status
- Not open for further replies.