It was trial and error
I tried with other attenuations and frequencies but the best sound is with these.
It seems to me a lot of coincidence that woofer resonance frequency is the same of the room.
-> http://medleysmusings.com/kef-q100-drive-unit-testing/
KEF Q100
It was trial and error. And exactly 65Hz.
I tried with other attenuations and frequencies but the best sound is with these.
It seems to me a lot of coincidence that woofer resonance frequency is the same of the room.
-> http://medleysmusings.com/kef-q100-drive-unit-testing/
An externally hosted image should be here but it was not working when we last tested it.
KEF Q100
It was trial and error. And exactly 65Hz.
Last edited:
They're not glued on Q200. They use exactly the same rubber socket that the speaker grille uses but there are six to hold the ring firmly so it is better in fact to use two tools, opposite each other. Back in the day, I used to have tools to disassemble early gen smartphones that woukd be ideal.
So what on earth has 65 Hz free-air resonant frequency of the driver to do with anything, when the Q100 vented box resonant frequency is about 55 Hz? How 55 Hz vented box resonance is connected with the supposed 65 Hz standing wave in your room?
I've come across reflex speakers deliberately designed not to null the impedance peak by splitting it into two smaller peaks, but instead the designer has tried to emphasize bass from a small driver by tuning for a bigger peak with reduced power handling. I don't know if that is the case for Q100 but I think my Q200 does not behave exactly according to Kefs very basic info. Curiosity about this is why I started this thread.
Another thing I'm wondering about is whether Kef put the dropping resistor before or after the tweeter cap on Q100. I guess before but I found that component value calculation and selection are simpler for Q200 if it is after.
Last edited:
The inductor does not give a consistent high frequency roll off if it meets a rising impedance. By using a zobel, the inductor sees an impedance that can be close to flat so the high frequency roll off is consistent. Or at least can be. In theory 🙂 However, if the bass driver roll off is carefully designed so it can match the tweeters increasing output, then a zobel probably isnt needed. This seems to the case with the Q100 but my Q200 has a zobel in the original XO, even though it seems to be the same driver. I guess this the cabinet playing its part, or is perhaps cos the Q200 is a more complex 3 way crossover when Q100 is a simple 2 way.
Last edited:
Indeed. What can be made with this kind of circuit can also be made without, with the same results. Impedance correction used between driver and crossover is a means to an end.
On the other hand a second order low pass filter can also use series RC shunting the driver, they look the same.
If anybody would like to fiddle around some, here are the impedance numbers for both (raw, no crossover) drivers in the Q100. (Foam plug inserted for the woofer measurement.)
FYI, the stock crossover is:
Single inductor on the woofer, 0.6mH, 0.55 DCR. (Normal polarity.)
Series RC on the tweeter, 0.68 ohms, 3.9uF. (Reversed polarity.)
Have fun. 🙂
Dave.
FYI, the stock crossover is:
Single inductor on the woofer, 0.6mH, 0.55 DCR. (Normal polarity.)
Series RC on the tweeter, 0.68 ohms, 3.9uF. (Reversed polarity.)
Have fun. 🙂
Dave.
Cool - thanks for the help. 🙂 The not-a-zobel is 8.2R and 10uF btw.
So how about the effects of a 2-litre sealed enclosure on a driver designed for a larger reflex load? Would seem like a bad idea .... so I guessed SP1587 in Q100 would not have the same suspension etc as SP1587.2 in Q200... same driver or different? Unfortunately, I haven't got around to getting a sharc or dsp etc so I have only limited tools for testing - DMM, ancient scope, old ears.
Yes, we can.
R and C are in series with each other and in series with tweeter, so which is the first or second (or third) doesn't matter.
Thanks ! Very interesting - looks like resonant frequency is 73hz at 32r ?
Last edited:
Got it thanks. I'm trying to raise the impedance of the tweeter to a constant 4R and drop the output by 3dB so I'm trying out a T-pad. But since Davey shared accurate impedance, I have that a little wrong. Again. 🙂
Hi,
This is the Q200 circuit for the mid &tweeter. The woofer circuit is independent.
Tweeter => 0.5R + 4.3uF in series.
Mid => 2.2R + 100uF in series, 6R + 0.95mH air-core to negative, and 10uF electro + 8.2R to negative.
The drivers continue to loosen up as Maty Tinman says, however, I found the nicest sound so far with 5R+3.4uF in series on the tweeter and a hand-wound inductor about 0.45mH and no zobel on the mid.
The drivers continue to loosen up as Tinman has said. I've lost track of break-in hours but it's "a lot". 🙂
I am puzzled that the cap is larger (4.3 vs 3.9uF) on the Q200 vs Q100 but the crossover frequency is stated as higher 2.8 vs 2.5Khz).
Last edited:
The cutoff frequency first order depends on the driver impedance plus the series resistor, which might explain the size discrepancy.
The intention of the mid crossover is a little unclear, but the series RC shunting the driver has values similar to impedance mitigation. If that were the case, a second order filter has some immunity to impedance variations and doesn't need it as much, especially considering it looks like the top end that is being operated on. On the other hand the filter is fairly well damped making it more like first order.
The RC might act in conjunction with the small series resistance (2r2) to shelve the top end a little. This might not be too unusual considering the reduction in acoustic power at the top end due to cone size.
One thing DIYers don't worry about as much is budget. Maybe it would be better fleshed out. There's a reason to choose the rolloff for each end of each way.
The intention of the mid crossover is a little unclear, but the series RC shunting the driver has values similar to impedance mitigation. If that were the case, a second order filter has some immunity to impedance variations and doesn't need it as much, especially considering it looks like the top end that is being operated on. On the other hand the filter is fairly well damped making it more like first order.
The RC might act in conjunction with the small series resistance (2r2) to shelve the top end a little. This might not be too unusual considering the reduction in acoustic power at the top end due to cone size.
One thing DIYers don't worry about as much is budget. Maybe it would be better fleshed out. There's a reason to choose the rolloff for each end of each way.
- Status
- Not open for further replies.