2.5 way with series woofers?

[depthsounder]

I'm cooking up a design with dual woofers, but the woofers I have on hand are low impedance (3.5 Re). I'm considering a 2.5 way, but obviously cannot parallel the woofers in the standard fashion. Would a shunt capacitor after the first woofer in a series configuration work?

-David

[Wolf]

I'm cooking up a design with dual woofers, but the woofers I have on hand are low impedance (3.5 Re). I'm considering a 2.5 way, but obviously cannot parallel the woofers in the standard fashion. Would a shunt capacitor after the first woofer in a series configuration work?

-David

Normally, what you filter the first component in a series connection will also affect the second woofer in a like manner.
Later,
Wolf

[romanbednarek]

I'm cooking up a design with dual woofers, but the woofers I have on hand are low impedance (3.5 Re). I'm considering a 2.5 way, but obviously cannot parallel the woofers in the standard fashion. Would a shunt capacitor after the first woofer in a series configuration work?

-David

Yes, there is a way to put two drivers in series and have a 2.5 way configuration similar to how you suggested. I've never built one personally but I've designed a few via simulation in the past.

The trick to this method is to "bypass" the bass around the 0.5 woofer by using a capacitor in parallel with that woofer only as you suggested. Sometimes using a series combiation of a capacitor and a resistor in the "bypass" circuit allows you to tweak things a bit further to get a better response (think of the capacitor as a device that has very low impedance at high frequencies and acts as a short across the speaker. Adding the resistor in series with the capacitor allows the bypass shunt to approach the value of the resistor at high frequencies rather than a short and allows some of the signal to still get to the 0.5 woofer).

The best way to design this type of crossover is to simulate it and experiment with different values until things look decent. I'm pretty sure that Jeff B. is familiar with this topology as well because I seem to recall discussing this with him on the forum a long time ago.

[depthsounder]

I'm cranking on it in speaker workshop, and it does seem to work. This may end up even less expensive than the parallel method, capacitors are less costly than inductors.

I wish I could have this design ready for DIY Atlanta, but there's no way I could build the cabinets in a month. I'm under heavy scrutiny from the wife until I finish the kitchen remodel.

-David

[Pete Schumacher ®]

Yes, there is a way to put two drivers in series and have a 2.5 way configuration similar to how you suggested. I've never built one personally but I've designed a few via simulation in the past.

The trick to this method is to "bypass" the bass around the 0.5 woofer by using a capacitor in parallel with that woofer only as you suggested. Sometimes using a series combiation of a capacitor and a resistor in the "bypass" circuit allows you to tweak things a bit further to get a better response (think of the capacitor as a device that has very low impedance at high frequencies and acts as a short across the speaker. Adding the resistor in series with the capacitor allows the bypass shunt to approach the value of the resistor at high frequencies rather than a short and allows some of the signal to still get to the 0.5 woofer).

The best way to design this type of crossover is to simulate it and experiment with different values until things look decent. I'm pretty sure that Jeff B. is familiar with this topology as well because I seem to recall discussing this with him on the forum a long time ago.

It won't work with series drivers for one reason, you don't get the sensitivity increase with drivers in series.

2.5-way requires that the drivers in question be in parallel, so that you have the 6dB sensitivity increase at the lowest frequencies.

I'd go ahead and add a couple more woofers and do the MTMWW 2.5-way, as I did with the Schumakubin. You keep the impedance above 4 Ohm overall that way.

[romanbednarek]

It won't work with series drivers for one reason, you don't get the sensitivity increase with drivers in series.

2.5-way requires that the drivers in question be in parallel, so that you have the 6dB sensitivity increase at the lowest frequencies.

I'd go ahead and add a couple more woofers and do the MTMWW 2.5-way, as I did with the Schumakubin. You keep the impedance above 4 Ohm overall that way.

No Pete, you are missing some key factors of how this method works as compared to a parallel 2.5 way design when you consider the sensitivity gain of two drivers in parallel versus what happens with two drivers in series. With two drivers in series the key is that you are able to allow the 0.5 woofer to only fill in the baffle step loss region.

It has been a while but I'm certain that I've helped people who had 4 ohm woofers use this approach. The advantage is an overall impedance higher than 4 ohms without any gains in sensitivity but without any losses due to BSC that would normally be required for a two way design with the same drivers (I think that's right but I'm not totally certain from memory).

Rather than try to explain things and make mistakes doing everything in my head, I ask that you take one of your existing simulations for a two way design and model it as a dual woofer series connection with the cap in parallel with the lower woofer to see how things work in this configuration.

[depthsounder]

I see what you mean there Pete. I just modeled it in SW and can somewhat confirm what you are saying. The first woofer will see an increase in sensitivity in the upper frequencies as the impedance in the shunt circuit drops. Even though the lower woofer is not receiving the current, the upper woofer makes up the difference. The result is no net difference in the upper frequencies.

I guess I'll stick with a MT design on these.

-David

[Pete Schumacher ®]

No Pete, you are missing some key factors of how this method works as compared to a parallel 2.5 way design when you consider the sensitivity gain of two drivers in parallel versus what happens with two drivers in series. With two drivers in series the key is that you are able to allow the 0.5 woofer to only fill in the baffle step loss region.

It has been a while but I'm certain that I've helped people who had 4 ohm woofers use this approach. The advantage is an overall impedance higher than 4 ohms without any gains in sensitivity but without any losses due to BSC that would normally be required for a two way design with the same drivers (I think that's right but I'm not totally certain from memory).

Rather than try to explain things and make mistakes doing everything in my head, I ask that you take one of your existing simulations for a two way design and model it as a dual woofer series connection with the cap in parallel with the lower woofer to see how things work in this configuration.

Roman, I know that I've got it right already.

Look at the overall sensitivity when the drivers are in series. It is no better than a single driver. You need to have a gain in sensitivity if 2.5-way is to work as intended. Now, if you add BSC to the XO, you could have a pseudo 2.5-way by shunting the ".5" woofer and letting only the upper woofer handle the mids.

[MSaturn]

Two drivers in series can indeed produce a small amount of gain, but not for BSC purposes.

[romanbednarek]

Roman, I know that I've got it right already.

Look at the overall sensitivity when the drivers are in series. It is no better than a single driver. You need to have a gain in sensitivity if 2.5-way is to work as intended. Now, if you add BSC to the XO, you could have a pseudo 2.5-way by shunting the ".5" woofer and letting only the upper woofer handle the mids.

Yeah, I think I understand what you are saying. My memory of this configuration isn't so great and I misinterpreted the results but I guess that you still arrive at a configuration similar to a 2.5 way design in the way that both woofers cover the lowest octaves and only the upper woofer covers the midrange for the most part. I suppose that the only advantage of this configuration is a higher overall impedance (relative to a 2 way design with a single 4 ohm woofer) unless I'm still misinterpreting the actual performance.

I guess this version of a 2.5 way design could be seen as an alternative to a 2 way MTM using two 4 ohm drivers in series, but please correct me if I'm wrong.

[Pete Schumacher ®]

Yeah, I think I understand what you are saying. My memory of this configuration isn't so great and I misinterpreted the results but I guess that you still arrive at a configuration similar to a 2.5 way design in the way that both woofers cover the lowest octaves and only the upper woofer covers the midrange for the most part. I suppose that the only advantage of this configuration is a higher overall impedance (relative to a 2 way design with a single 4 ohm woofer) unless I'm still misinterpreting the actual performance.

I guess this version of a 2.5 way design could be seen as an alternative to a 2 way MTM using two 4 ohm drivers in series, but please correct me if I'm wrong.

One other benefit over a TM is that you have double the radiating area for bass, and that's never a bad thing.

[Pete Schumacher ®]

Two drivers in series can indeed produce a small amount of gain, but not for BSC purposes.

Not sure what you mean by small amount of gain.

Without touching the gain control on the amp, adding a second driver in series will halve the power, but double the Sd, keeping the radiated SPL the same.

You gain efficiency, but not sensitivity.

[Wolf]

Not sure what you mean by small amount of gain.

Without touching the gain control on the amp, adding a second driver in series will halve the power, but double the Sd, keeping the radiated SPL the same.

You gain efficiency, but not sensitivity.

You read my mind, Ditto!
Wolf

[Paul W]

Here's a 2.5 MMTMM from 2003 using a series cap shunt to manage the vertical radiation pattern (not baffle step compensation).



http://www.geocities.com/pnwright3/T...TOR_photo.html

[depthsounder]

These are a super budget design ($10 woofer on sale, $7 tweeter) for my brother in law, so I think I'll stick with a MT rather than buying four more woofers at the non sale price.

This will be an interesting design with a peerless SDS woofer and a vifa BC25 tweeter (buyouts). His birthday is in December so don't hold your breath for pics!

The SDS seem to have a natural rise in the low end, and he'll likely have these up against a wall, so I don't think I'll need alot of BSC anyways.

Thanks for all the contributions, I'm sure we all benefited from the discussion.


By the way Paul those are pretty cool. I'd like to experiment with controlling vertical directivity one of these days. Your use of the shunt capacitor to have the midrange come from just the center speakers is pretty ingenious and is exactly the effect we observed here.

-David