modifying Nelson Pass 4333 crossover mod for bi-amp

[David24x7]

Hi All,
When I built my 4333 project this summer, I decided to bi-amp using a Rane AC22 active crossover to split the signal for low and mid/hi. I used the Nelson Pass L300/4333 mod (http://www.firstwatt.com/pdf/art_l300.pdf) to build a passive crossover for the mid/hi and just decided to eliminate the low frequency circuit.

Now I want to experiment with crossover frequency but I'm afraid that the passive crossover still has a low pass filter for 800hz for the mids. If I use my active crossover to change to a 500Hz crossover point, for instance, will that leave me with a gap between 500Hz and 800Hz? If so, how would I modify the crossover circuit to allow me to actively change the crossover point between low and mid?

Does that make sense? I assume that the low frequency circuit only has a hi-pass filter and the high frequency circuit would only have a low-pass filter but the mid-range would have both low and hi-pass filters.

FWIW - the bass does sound thinner when I crossover at 500Hz. I'm using a tubed dynaco ST-70 (35wpc) for the mid/hi end and a solid state Sansui AU-7500 (32wpc) for the bottom end (until I get my Crown D150A Series II back from the shop).

You can see the Nelson Pass Crossover schematic here: http://www.firstwatt.com/pdf/art_l300.pdf

Thanks in advance for your advice.

[ivica]

Hi All,
When I built my 4333 project this summer, I decided to bi-amp using a Rane AC22 active crossover to split the signal for low and mid/hi. I used the Nelson Pass L300/4333 mod (http://www.firstwatt.com/pdf/art_l300.pdf) to build a passive crossover for the mid/hi and just decided to eliminate the low frequency circuit.

Now I want to experiment with crossover frequency but I'm afraid that the passive crossover still has a low pass filter for 800hz for the mids. If I use my active crossover to change to a 500Hz crossover point, for instance, will that leave me with a gap between 500Hz and 800Hz? If so, how would I modify the crossover circuit to allow me to actively change the crossover point between low and mid?

Does that make sense? I assume that the low frequency circuit only has a hi-pass filter and the high frequency circuit would only have a low-pass filter but the mid-range would have both low and hi-pass filters.

FWIW - the bass does sound thinner when I crossover at 500Hz. I'm using a tubed dynaco ST-70 (35wpc) for the mid/hi end and a solid state Sansui AU-7500 (32wpc) for the bottom end (until I get my Crown D150A Series II back from the shop).

You can see the Nelson Pass Crossover schematic here: http://www.firstwatt.com/pdf/art_l300.pdf

Thanks in advance for your advice.
HI David,
I would not suggets you to use 500Hz crossover for 2420 &2312
Regards
Ivica

[David24x7]

HI David,
I would not suggets you to use 500Hz crossover for 2420 &2312
Regards
Ivica

Fair enough. I know that the x-over point as designed is 800Hz and it does sound better that way (now).

But do you know if I properly designed my passive crossover to take advantage of an active crossover with variable xover frequencies? Or did I inherently implement an 800Hz low pass filter to my LE85 mids no matter what my amp is feeding these drivers?

[Don C]

If you add an external crossover for bi-amping, you will have two filters at the same frequency, resulting in a loss of output at and near the crossover frequency. It will sound awful. Instead, you would want to remove the hi-pass elements from the crossover network that are in the midrange circuit. These would be capacitors in series with the driver, you would replace those with wire. Or inductors wired in parallel with the driver would serve a similar high pass function, those would be removed and the circuit left open there.
The trouble here is that your midrange driver will be susceptible to damage in case the settings on the active crossover are set wrong. As an option the values of those high pass components could be altered to lower the crossover frequency at the midrange, so that there is no overlap with the external crossover. That has the advantage of leaving some protection in place so that your driver is less likely to be damaged. JBL followed this practice in the XPL200, and a similar circuit is found on the new M2s.
You may find, having done all of this, that the values used in the original circuit are not the ones that electrical theory would have called for, the circuit having been developed by ear. And thus that the passive crossover matches the driver components better and sounds better than the active crossover.
It would be completely understandable if you decided that this might be more complex than you would have thought and would decide to leave things alone.

[ivica]

Fair enough. I know that the x-over point as designed is 800Hz and it does sound better that way (now).

But do you know if I properly designed my passive crossover to take advantage of an active crossover with variable xover frequencies? Or did I inherently implement an 800Hz low pass filter to my LE85 mids no matter what my amp is feeding these drivers?

Hi David,

As i have remembered driver diaphragm displacement rises (for the same sound level -pressure ) when the applied frequency lowr down, say for two times less frequency would involve four times larger diaphragm displacement, so if want o comare 500 Hz vs 800Hz on 500Hz diafragm displacemen would be (8/5)^2= 2.56 times larger, not to mention that LF and MF frequency range would "cover" very large amout of sound energy (lewel), so not to mention that on 500Hz woud be larger diaphragm displacemen, but the amout of reproduced sound energy would rise tward lower frequency.
I would suggest to apply protective capacitor (about 33 to 47 uF) connected in-series with the driver if using active crossover, to protect the driver from lower frequency.
Using passive networks such capacitors are there in the network construction.
Using active networks it is possible to introduce time compensation between the drvers.
Regards
Ivica

[speakerdave]

In the 4333A the biamp switch leaves two things from the passive crossover in place: The high pass network complete, and the zobel across the woofer. Therefor all that was needed in the active crossover was an 800Hz low pass. N. Pass's low pass looks very similar to the original, so maybe the best approach would be to follow the same method, i.e. maybe his network does not change the crossover frequency very much.

If you try to fly without the passive high pass, how are you going to determine your EQ settings for the frequency compensation that is in that network? The same question applies even more so if you change the crossover frequency. Also, it has been shown that there are body resonances that come into play with one-inch compression drivers coming down out of the high hundreds. I like them crossed over at about about 1200. Plus, I second Ivica on 500Hz with that horn. Not made for it.

You did well to build a speaker system that was well-engineered. Now you are wandering out into a pond with an unknown bottom. If you are sensing a slight something missing in the midrange of that speaker, I sympathize fully, since I experienced the same thing. My answer was a four-way.

[Mannermusic]

In the 4333A the biamp switch leaves two things from the passive crossover in place: The high pass network complete, and the zobel across the woofer. Therefor all that was needed in the active crossover was an 800Hz low pass. N. Pass's low pass looks very similar to the original, so maybe the best approach would be to follow the same method, i.e. maybe his network does not change the crossover frequency very much.

If you try to fly without the passive high pass, how are you going to determine your EQ settings for the frequency compensation that is in that network? The same question applies even more so if you change the crossover frequency. Also, it has been shown that there are body resonances that come into play with one-inch compression drivers coming down out of the high hundreds. I like them crossed over at about about 1200. Plus, I second Ivica on 500Hz with that horn. Not made for it.

You did well to build a speaker system that was well-engineered. Now you are wandering out into a pond with an unknown bottom. If you are sensing a slight something missing in the midrange of that speaker, I sympathize fully, since I experienced the same thing. My answer was a four-way.

Last two sentences: +1.

[David24x7]

If you add an external crossover for bi-amping, you will have two filters at the same frequency, resulting in a loss of output at and near the crossover frequency. It will sound awful. Instead, you would want to remove the hi-pass elements from the crossover network that are in the midrange circuit. These would be capacitors in series with the driver, you would replace those with wire. Or inductors wired in parallel with the driver would serve a similar high pass function, those would be removed and the circuit left open there.
The trouble here is that your midrange driver will be susceptible to damage in case the settings on the active crossover are set wrong. As an option the values of those high pass components could be altered to lower the crossover frequency at the midrange, so that there is no overlap with the external crossover. That has the advantage of leaving some protection in place so that your driver is less likely to be damaged. JBL followed this practice in the XPL200, and a similar circuit is found on the new M2s.
You may find, having done all of this, that the values used in the original circuit are not the ones that electrical theory would have called for, the circuit having been developed by ear. And thus that the passive crossover matches the driver components better and sounds better than the active crossover.
It would be completely understandable if you decided that this might be more complex than you would have thought and would decide to leave things alone.

Thanks Don,
That's what I was concerned about - even if I keep the crossover point at 800Hz (which seems to get universal agreement). I'll have to experiment with bypassing some of the filters that you mentioned to see if there is any difference. Thx for the tips.

Based on the page 6 schematic (http://www.firstwatt.com/pdf/art_l300.pdf), are you suggesting to remove both capacitors (22uF and 2.2uF) and both inductors (1.0MH and 0.2MH) - just leaving the two resistors and the lpads? I'm having trouble translating the theory into the schematic.

[Don C]

No.
I don't know the exact function of all of these parts. The function may be more complex than I understand, I'm not a crossover designer myself. So I'd want to run some sort of simulation to see what the response would be if they were removed. But just considering some very basic electronics that I do understand, a cap in parallel with the driver such as how C4 is wired, or an inductor in series with the driver, such as how L3 is connected, would generally serve as low pass filters. So they likely would be part of the crossover that works between the midrange and tweeter, not between the woofer and midrange.
So you would probably want to leave those alone. I was also trying to suggest gently that not knowing the function of these parts before beginning, is a sign that it might be best not to attempt any changes at all.


Thanks Don,
That's what I was concerned about - even if I keep the crossover point at 800Hz (which seems to get universal agreement). I'll have to experiment with bypassing some of the filters that you mentioned to see if there is any difference. Thx for the tips.

Based on the page 6 schematic (http://www.firstwatt.com/pdf/art_l300.pdf), are you suggesting to remove both capacitors (22uF and 2.2uF) and both inductors (1.0MH and 0.2MH) - just leaving the two resistors and the lpads? I'm having trouble translating the theory into the schematic.

[tonymontana168]

Hello
What I am using is the nelson pass passive xo for mid le85 and 077 tweeter and an electronic crossover for the 136a for sub. MY electronic crossover is a tdm sledgehammer 24cx-2 which is a 24db slope I believe and I crossed it at 1200hz which is what I am supposed to do I guess??? right? but......the nelson I believe is a 12db slope. Is that a problem for me?? do I need a new electronic crossover? I have a pass labs preamp aleph L which has 2 Outputs and I use the second one for the highs which goes to the amp and then to the passive xo for high!
It sounds pretty good but I think theres something wrong. I have a cheap 100$ calibrating mic which I take mesurements and it always makes a hump at the crossing point. Is that normal? Any advise would help! The mid bass just doesn't seam full if I make the frequency flat but now with this 100$ mic and r.e.w is it even accurate? I see that my sub gets signal all the way too 8khz! Isnt that bad? 82615

[ivica]

Hello
What I am using is the nelson pass passive xo for mid le85 and 077 tweeter and an electronic crossover for the 136a for sub. MY electronic crossover is a tdm sledgehammer 24cx-2 which is a 24db slope I believe and I crossed it at 1200hz which is what I am supposed to do I guess??? right? but......the nelson I believe is a 12db slope. Is that a problem for me?? do I need a new electronic crossover? I have a pass labs preamp aleph L which has 2 Outputs and I use the second one for the highs which goes to the amp and then to the passive xo for high!
It sounds pretty good but I think theres something wrong. I have a cheap 100$ calibrating mic which I take mesurements and it always makes a hump at the crossing point. Is that normal? Any advise would help! The mid bass just doesn't seam full if I make the frequency flat but now with this 100$ mic and r.e.w is it even accurate? I see that my sub gets signal all the way too 8khz! Isnt that bad?

Hi Tonymontana,

Passive networks have more complex responses then standard active types. If ONLY looking at the LC values of the LF and HF parts it seems to me that LF is round 750Hz and HF is round 1000Hz. But driver response itself and ITS impedance would introduce more complex response to the whole system then expected. Time delay differences introduced by driver placements, and network phase FR response can make another problem, especially round crossover region. So all that have to be aware off when applying active crossover instead of passive crossover.
I would suggest to do Your measurements , at least , 2 or 3 meters from the speakers, far away of larger surface such as floor, walls or ceiling , I would suggest to take the speakers outside the room , laying on the back box side, and put the microphone about 2 or 3 m above, while it would "look at" the 2308 lenses or 2405 horn, and do the proper measurements of the FR response. In such measurements, You would remove the influences of the sound waves reflections, and can get real speakers FR response.

regards
ivica