To connect a 1.5 way system together it is traditionally done with just 1 inductor in series with one of the drivers. With this setup there is a chance of lobing with the overlap of frequencies.
What would the outcome of turning it into a 2nd order filter and connecting a cap across the terminals of that driver?
I have not had a chance to play around with this on my setup... I have a 1.5 way with 2x alpair 10p drivers. width of cabinet is 9.5" and I am currently using a 3mH 1.38ohm air inductor.
Can anyone point me in the right direction as of size of cap and how to calculate it?
I did see a crossover simulation software somewhere....
Thank you all for any help!!
What would the outcome of turning it into a 2nd order filter and connecting a cap across the terminals of that driver?
I have not had a chance to play around with this on my setup... I have a 1.5 way with 2x alpair 10p drivers. width of cabinet is 9.5" and I am currently using a 3mH 1.38ohm air inductor.
Can anyone point me in the right direction as of size of cap and how to calculate it?
I did see a crossover simulation software somewhere....
Thank you all for any help!!
Blast, hit the wrong button. 
Rewriting the essence swiftly, if all you are doing is using an identical wideband driver to provide LF support in a 1.5 system, you can technically use a higher order filter, although generally there is little advantage to doing so in such designs. Time alignment in the sense being referred to above primarily (primarily) applies to multiways with a greater driver offset and matching high - low pass acoustic slopes to achieve good phase tracking through the transition band, so a 2nd order electrical can be used, although depending on the design details you may need to invert the electrical polarity of one unit. You may also find it necessary to use an RC Zobel to flatten the rising VC impedance as 2nd order & > electrical filters exhibit more deviation in behaviour than a 1st order electrical, where the rate of rolloff simply reduces compared to what was intended as the VC impedance increase.
Edit (hopefully I won't hit the wrong button this time
):
Re whether identical drivers are time aligned, they don't have the deviation different types of drive unit (e.g. a woofer & a tweeter) usually will have in terms of the VC, but they are not coincident, so one is still somewhat offset from the listening position relative to the other. If you measure the distance from your listening position to each driver, unless there is a physical offset, one will always be further away, how much depending on the design.
Rewriting the essence swiftly, if all you are doing is using an identical wideband driver to provide LF support in a 1.5 system, you can technically use a higher order filter, although generally there is little advantage to doing so in such designs. Time alignment in the sense being referred to above primarily (primarily) applies to multiways with a greater driver offset and matching high - low pass acoustic slopes to achieve good phase tracking through the transition band, so a 2nd order electrical can be used, although depending on the design details you may need to invert the electrical polarity of one unit. You may also find it necessary to use an RC Zobel to flatten the rising VC impedance as 2nd order & > electrical filters exhibit more deviation in behaviour than a 1st order electrical, where the rate of rolloff simply reduces compared to what was intended as the VC impedance increase.
Edit (hopefully I won't hit the wrong button this time
): Re whether identical drivers are time aligned, they don't have the deviation different types of drive unit (e.g. a woofer & a tweeter) usually will have in terms of the VC, but they are not coincident, so one is still somewhat offset from the listening position relative to the other. If you measure the distance from your listening position to each driver, unless there is a physical offset, one will always be further away, how much depending on the design.
Last edited:
Thanks for your replies!!
So generally..... a wider front baffle would be better as it would decrease the frequency of baffle step. This in turn would reduce lobing between the 2 drivers.
So I think I will go from my 9.5" wide baffle to a 14".
Would you recommend a Zobel on the .5 driver?
And I would be correct in saying that the cap for the 2nd order filter would be across the terminals of the .5 driver? hence using a Zobel would lessen the interaction the cap has with the rising vc impedance.
is this correct?
So generally..... a wider front baffle would be better as it would decrease the frequency of baffle step. This in turn would reduce lobing between the 2 drivers.
So I think I will go from my 9.5" wide baffle to a 14".
Would you recommend a Zobel on the .5 driver?
And I would be correct in saying that the cap for the 2nd order filter would be across the terminals of the .5 driver? hence using a Zobel would lessen the interaction the cap has with the rising vc impedance.
is this correct?
You'll be able to figure out if a zobel is needed when you run the crossover simulations.
Typically a low XO frequency with a low inductance voice coil will not need a zobel, so the requirement will be derived from your driver and design choice.
I've occasionally stumbled upon designs where the zobel is elegantly worked into the shunt leg, eliminating the need for multiple components.
Typically a low XO frequency with a low inductance voice coil will not need a zobel, so the requirement will be derived from your driver and design choice.
I've occasionally stumbled upon designs where the zobel is elegantly worked into the shunt leg, eliminating the need for multiple components.
its for an Alpair 10p.
inductance is 50uH
see:
Markaudio Alpair-10P Paper Cone 6" Full Range: Madisound Speaker Components
thanks
inductance is 50uH
see:
Markaudio Alpair-10P Paper Cone 6" Full Range: Madisound Speaker Components
thanks
Given that it is a 1.5 way and that the XO is low enuff to use the extra driver as BSC, as long as they are on the same baffle (and close together) they are already time aligned. At least before the XO is introduced.
How a 2nd order filter affects things is dependent on many things and your question cannot be answered with out real worls measuring.
We have rarely found need for BSC (it is often overused) and would wire a 2nd driver in series with a big cap across it. This way you get no extra gain, but 4x less excursion in the bass. With a 1.5 way in parallel w an inductor you almost always end up with more compensation than needed (6dB) and too much bass. Or we use the 2nd driver in such a way that the 2nd driver is essentially acoustically rolled off (ie a castle-style “bipole”.
Baffle Diffraction Step (some parts might be a bit out-of-date). There are a number of other articles that follow the intro.
dave
I should add these to the conversation:
http://www.planet10-hifi.com/downloads/Dual-Driver-Wiring.pdf
and i might as well post this one too (4 drivers):
http://www.planet10-hifi.com/downloads/Quad-Driver-Wiring.pdf
dave
http://www.planet10-hifi.com/downloads/Dual-Driver-Wiring.pdf
and i might as well post this one too (4 drivers):
http://www.planet10-hifi.com/downloads/Quad-Driver-Wiring.pdf
dave
Using a 1st order approximation the 1st will have a BS(-3) at 480 Hz, the 2nd at 325 Hz. BSC filter should be between 0.707 to 1 times BS(-3). In our WAWs the lower numbers seem to work better. And do have the drivers such that their C-C is less than a quarter wavelength at teh filter point.
That is very dependent on the specific driver. What ar eyou using?
dave
I have tried both the series connection with the cap and also parallel with the inductor... to my ears I preferred the inductor and paralleled. Cannot tell you why though.
Maybe the caps I used were not of high quality (would it really matter?).. or I hadn't dialed in the correct cap value .. but I had dialed in correct inductor value.
Maybe it looks like I need to revisit it.
All I know I can tell there is some combing effects going on with my setup... actually now that I think of it... it doesn't happen when I listen just to one side (L or R) but when I listen in stereo... maybe that's normal. or an artifact of room reflections etc.
Maybe the caps I used were not of high quality (would it really matter?).. or I hadn't dialed in the correct cap value .. but I had dialed in correct inductor value.
Maybe it looks like I need to revisit it.
All I know I can tell there is some combing effects going on with my setup... actually now that I think of it... it doesn't happen when I listen just to one side (L or R) but when I listen in stereo... maybe that's normal. or an artifact of room reflections etc.
Maybe, one of those situations were you have to try and see how things work out.
Can you biamp them — you'd need 4 channels the same, and then a series R, shunt cap in fron tof the 2nd amp. No Zobel needed.
dave
Either you have them puleld way out in a fairly large roonm, or like more bass than flat (adding BSC may flatten the on-axis, but by the same amount gives too much room response). I do not fault anyone for their personal preferences.
I would not use anything less than a poly cap, how good depending on how high your budget is. We typically use the cheap Solen polys (many other brands coming out of the same factory). A bipoler elco will get in the way.
This size poly is approaching beer can size.
Given that with the cap you have no extra gain, the actual point of the filter is way less critical.
Maybe it looks like I need to revisit it.
That suggests that it is a stern thing, not a 2 driver thing.
dave
The highest filter you would want to use would turn over at about 480 Hz. The inductor would need to be that or lower, but its value is driven (mostly) by the cabinet width.
dave
looking at your calculations I think my inductor for the parallel connection is spot on.
For the capacitor looks like a 80uF would work for the 9.5" wide baffle. Looked at my caps I have and I was using a 60uF.
Wrapping my head around what a 60uF vs a 80uF cap would do as far as baffle step and frequency response.
For the capacitor looks like a 80uF would work for the 9.5" wide baffle. Looked at my caps I have and I was using a 60uF.
Wrapping my head around what a 60uF vs a 80uF cap would do as far as baffle step and frequency response.
Series conenction has no effect on FR, just switches impedance around, and means you need a low output impedance amp (most of them). It does mean an easier job for the amp at the expense of having a fourth the amount of power in the LF.
60 bvrs 80 moves the filter point down by ~ a 60/80 = 3/4 multiplier. As long as the cap is big enuff to move the filter below the c-c 1/4 wavelength you should be fine,
dave
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.