I have recently finished my W3-871s pictured below. As yet they have no filtering at all. I am very impressed with the imaging but they are too bright. I would like some advice on how to tame the highs preferably without ruining the imaging.
As you can see they are in a closed spherical enclosure which is about 3 litres.
Regards BT
As you can see they are in a closed spherical enclosure which is about 3 litres.
Regards BT
Attachments
Hi,
read up on baffle step compensation and the
various filter designs published for the W3-871s.
http://sound.westhost.com/bafflestep.htm
/sreten.
read up on baffle step compensation and the
various filter designs published for the W3-871s.
http://sound.westhost.com/bafflestep.htm
/sreten.Zaph audio has a notch filter designed for that driver. It really needs one, its got some awful peaks in the high range.
Don't fear the filter, and check out: http://www.zaphaudio.com/archives.html
The W3-871 design is the third one down, I think, download it and take a look at the filter, it might be the ticket.
I don't think you'll need a baffle step compensation, because of the spherical enclosure.
peace,
sam
Don't fear the filter, and check out: http://www.zaphaudio.com/archives.html
The W3-871 design is the third one down, I think, download it and take a look at the filter, it might be the ticket.
I don't think you'll need a baffle step compensation, because of the spherical enclosure.
peace,
sam
I've had a quick look at the 2 links supplied as response to my question.
1) What is the baffle width of a sphere?
2) On the Zaph Audio pages there are 2 notch filter for differing enclosures. I assume possibly incorrectly, that the capacitor is the component which takes care of baffle step. If so, how do I calculate what it should be for me?
Regards BT
1) What is the baffle width of a sphere?
2) On the Zaph Audio pages there are 2 notch filter for differing enclosures. I assume possibly incorrectly, that the capacitor is the component which takes care of baffle step. If so, how do I calculate what it should be for me?
Regards BT
toseland said:
Probably the baffle width can be the diameter of the driver initially. In reality it also depends on the driver FR, so you might end up with an optimum baffle somewhere smaller than the diameter of the sphere.
The inductor and the resistor is mostly what takes care of the baffle step.
How did you make the sphere? It looks nice.
soonqsc I didn't make the spheres I bought them in a department store in the UK.
John from Zaph Audio uses the circuit attached. (This is for a 7 inch baffle. For the 5.75 inch baffle only C1 changes to 3.3mfd)
My sphere is about 8 2/3 inches so C1 would need to be larger than the 8.2mfd used in the 7 inch example.
However changing C1 will also change W0 and Q.
I just need to work out if is enough to just change C or whether
L will need to be modified and to what values?
John from Zaph Audio uses the circuit attached. (This is for a 7 inch baffle. For the 5.75 inch baffle only C1 changes to 3.3mfd)
My sphere is about 8 2/3 inches so C1 would need to be larger than the 8.2mfd used in the 7 inch example.
However changing C1 will also change W0 and Q.
I just need to work out if is enough to just change C or whether
L will need to be modified and to what values?
Attachments
Zaph mentions that the W3-871S doesn't need that much compensation due to its natural frequency response. I assume the notch values chosen account for this.
You can't guestimate your notch values based on Zaph's design since yours is quite different. You will probably need to measure the FR to design an accurate notch.
You can't guestimate your notch values based on Zaph's design since yours is quite different. You will probably need to measure the FR to design an accurate notch.
Hi, I basically suspect, as Vikash said, that 0.7mh maybe too much. There is the danger of overexcursion or a thick sound, or both.
But basically we dont know what is at play in your system as a whole. Can you take measurements of both sat responses and impedance? I will be happy to simulate your needs in Leap or Soundeasy.
But basically we dont know what is at play in your system as a whole. Can you take measurements of both sat responses and impedance? I will be happy to simulate your needs in Leap or Soundeasy.
What more can I tell you. Basically I will have a very similar volume to Vikash, a sealed enclosures like Vikash and the same driver as Vikash. Only my enclosures are spheres of 22cm (8.66 inches) external diameter.
These are driven with a gainclone much like Vikash's. At present
there is nothing between the amp and drivers but cable.
The subs are connected across the speaker terminals in parallel with the satellites.
The Subs can be tune for phase, applitude and frequency.
That about all the information I have.
regards BT
These are driven with a gainclone much like Vikash's. At present
there is nothing between the amp and drivers but cable.
The subs are connected across the speaker terminals in parallel with the satellites.
The Subs can be tune for phase, applitude and frequency.
That about all the information I have.
regards BT
Hi,
zaph's filters are designed to reduce a small hump is frequency
response and reduce the hump caused by diffraction/baffle step.
A sphere should not have the hump, just a smooth transition.
The low diffraction box has a 1.5dB hump, the 5.75" box is not shown.
The filters are low Q and implement baffle step for lower
frequencies and extend the treble for higher frequencies.
They only differ in the centre frequency, by my calculations,
2.3Khz and 3Khz, but this doesn't include the drivers inductance.
Because the sphere doesn't have a hump I'd reduce R to 5.6R.
Might sound like a cop-out but keep 0.7mH and use 4.7uF.
You should be very near what you need, and have a lot better sound.
As your using subs, a 100uF series capacitor is going to be better than nothing.
(Bypassed with a similar quality / value capacitor to the above)
edit: I don't suspect 0.7mH is too much, R controls the amount
of baffle step and consequent effective treble boost/extension.
/sreten.
zaph's filters are designed to reduce a small hump is frequency
response and reduce the hump caused by diffraction/baffle step.
A sphere should not have the hump, just a smooth transition.
The low diffraction box has a 1.5dB hump, the 5.75" box is not shown.
The filters are low Q and implement baffle step for lower
frequencies and extend the treble for higher frequencies.
They only differ in the centre frequency, by my calculations,
2.3Khz and 3Khz, but this doesn't include the drivers inductance.
Because the sphere doesn't have a hump I'd reduce R to 5.6R.
Might sound like a cop-out but keep 0.7mH and use 4.7uF.
You should be very near what you need, and have a lot better sound.
As your using subs, a 100uF series capacitor is going to be better than nothing.
(Bypassed with a similar quality / value capacitor to the above)
edit: I don't suspect 0.7mH is too much, R controls the amount
of baffle step and consequent effective treble boost/extension.
/sreten.
toseland said:
Hi,
as detailed (see B3S) on Zaph's site a series capacitor is not
particularly effective. If amplifier load is not an issue then things
can be improved by connecting a 10watt 15R resistor to ground
after the 100uF capacitor, smoothing the load it sees.
Also note the speakers should point sraight ahead, not toed in.
/sreten.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.