Gauges for inductor components in 2-way speaker with impedance correction

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
A

alexclaber

Member
Joined 2005
I have designed the following 2-way crossover for a woofer+mid bass guitar cab:

Series Inductor on Woofer: 1.9mH (2nd order Butterworth)
Parallel Cap on Woofer: 22uF

Series Cap on Midrange: 22uF (2nd order Butterworth)
Parallel Inductor on Mid: 2mH

Series RC Parallel to Midrange: 10uF and 8.2ohms (Zobel)
Series RL Parallel to Midrange: 0.6mH and 3.3ohms (Resonance peak correction)

I expect the speaker to be used with amps up to 1000W though that kind of power is very unlikely to be sustained due to the dynamic nature of bass guitar and live music. Maybe the cab will see 50-100W average with peaks up to 1000W+.

I'm intending to use a 15 gauge P-core plus discs inductor in series with the woofer but I'm don't know how low a gauge I can safely use on the two shunt inductors on the midrange. Any answers?

Thanks,

Alex
 
P

pjpoes

Member
Joined 2005
For that kind of power you will want to stay with 15-12 guage. I would stay away from standard iron core because of core saturation, but would consider something like a pcore, Erse Super Q, or the C-core toroidal inductors. All of which will not saturate as easily and can handle considerable power. None are rated for 1000 watts, but as you say, the speakers will likely never see 1000 watts sustained, so it isn't a big issue. I would mount the crossovers externally though, as it will make repairs quicker in the event that an inductor does get fried. It also allows air to circulate around the components better.
 
S

sreten

R.I.P.
Joined 2003
Hi,

All those parallel components on the midrange do sound like a good
idea, especially the RL with R=3.3R. The Zobel on the midrange is
not needed, whilst omitting it for the bass unit means the "textbook"
2nd order c/o function will not be accurate for it.


There is a lot to be said for a simple 1st order series crossover
with just 2 components for this application - as it is resilient
to driver impedance variation, cheap and does not waste power.

Values for 1khz / 8ohm are 1.3mH and 20uF.


:)/sreten.
 
A

alexclaber

Member
Joined 2005
sreten said:
There is a lot to be said for a simple 1st order series crossover
with just 2 components for this application - as it is resilient
to driver impedance variation, cheap and does not waste power.

Values for 1khz / 8ohm are 1.3mH and 20uF.
Click to expand...

Thanks, that's very interesting. Have done some experimenting with this and it seems to work rather well in theory...

Is there anything in particular I should be aware of with such a 1st order series design?

Alex
 
A

alexclaber

Member
Joined 2005
sreten said:
Hi,

It depends on the intended c/o point and the details of the two drivers.

:)/sreten.
Click to expand...

The drivers in question are the Eminence 3015LF 15" woofer and the Beyma 605Nd 6.5" midrange. The crossover point needs to be high enough for the midrange to be safe from overexcursion/overpowering but low enough for the off-axis respons to be relatively smooth. With a 24dB active filter you can cross as low as 500Hz without problems but with a shallower slope ~1kHz seems wiser.

Alex
 
S

sreten

R.I.P.
Joined 2003
Hi,

With ~ 6dB slopes the midrange will still increase in excursion until
it hits its its bass roll-off, in other words it should be tuned to a Qts
of at least 0.7, or possibly somewhat higher, up to 0.95 (1dB ripple)
to maximise the frequency where excursion will begin decreasing.

I'm not sure what the best arrangement is for those two drivers,
or that they are a particulary good combination, the Eminence
has issues around 1.5 kHz that sort of implies it will mess up
the midrange of a separate unit if a simple c/o is used.

The Deltalite for example looks easier to crossover.

:)/sreten.
 
badman

badman

Member
Joined 2003
pothaudio.com
alexclaber said:


The drivers in question are the Eminence 3015LF 15" woofer and the Beyma 605Nd 6.5" midrange. The crossover point needs to be high enough for the midrange to be safe from overexcursion/overpowering but low enough for the off-axis respons to be relatively smooth. With a 24dB active filter you can cross as low as 500Hz without problems but with a shallower slope ~1kHz seems wiser.

Alex
Click to expand...


Jeeze man! That's a heck of a high crossover point for a 6.5", high efficiency prostyle or not. The 15" will be well into breakup at 1kHz, and will have beaming and off axis lobing like crazy. Even the .5mm Xmax Audax PR17OMO has recommended XO to a midbass transducer at around 500 (or lower) with a 12dB slope.

Headroom is a lovely thing to have, but I think you're taking the power handling requirements very extremely.
 
A

alexclaber

Member
Joined 2005
sreten said:
Hi,

With ~ 6dB slopes the midrange will still increase in excursion until
it hits its its bass roll-off, in other words it should be tuned to a Qts
of at least 0.7, or possibly somewhat higher, up to 0.95 (1dB ripple)
to maximise the frequency where excursion will begin decreasing.
Click to expand...

Good point. I've reduced the midrange enclosure size to raise the Qts and raise the bass roll-off which gives usefully more output higher up too.

sreten said:
I'm not sure what the best arrangement is for those two drivers,
or that they are a particulary good combination, the Eminence
has issues around 1.5 kHz that sort of implies it will mess up
the midrange of a separate unit if a simple c/o is used.

The Deltalite for example looks easier to crossover.
Click to expand...

Right, I'll make sure the output is reduced more before the break-up peak. The 3015LF is outstanding when it comes to moving air - huge Vd, which means it can keep up with a whole array of generic 10" bass guitar woofers, so it's worth having to deal with crossover challenges.

badman said:
Jeeze man! That's a heck of a high crossover point for a 6.5", high efficiency prostyle or not. The 15" will be well into breakup at 1kHz, and will have beaming and off axis lobing like crazy. Even the .5mm Xmax Audax PR17OMO has recommended XO to a midbass transducer at around 500 (or lower) with a 12dB slope.

Headroom is a lovely thing to have, but I think you're taking the power handling requirements very extremely.
Click to expand...

Ok, thanks, advice taken on board!

I'm now experimenting with a conventional 2nd order crossover with full impedance correction. Due to the output of the midrange diminishing as it goes lower it seems very difficult to get the woofer/mid output to acoustically cross lower than 1kHz even though for this version of the crossover I have the woofer lowpass at ~800Hz and the mid highpass at ~500Hz.

Alex
 
A

alexclaber

Member
Joined 2005
Done another redesign, this time experimenting with values to get the response I want with minimal impedance compensation. Have managed to do without woofer Zobel and mid LCR, however I have used a mid Zobel as without it I get a nasty peak in response around 1.2kHz.

Components are as follows:

Woofer: 2mH 56uF
Mid: 39uF 0.8mH
Mid Zobel: 10uF 7.5Ohm

Intending to use a pair of 15R 20W resistors in parallel for the Zobel, 15 gauge P-core for woofer, 18 gauge P-core for mid (both are 0.15R).

Acoustic output of 15" and 6.5" now crosses at ~700Hz and the filter slopes are pretty flat on the far side of each filter, mid highpass F3 @ 530Hz, woofer lowpass F3 @ 730Hz, and slopes are close to 12dB/octave without any anomalies.

How does that sound?

Thanks!

Alex
 
A

alexclaber

Member
Joined 2005
sreten said:
where do you get 39uF with 0.8mH from ?
39uF seems far too large for a sensible filter Q.
Click to expand...

Trial and error using the passive crossover designer spreadsheet. I used the theoretical values as a starting point and then experimented with how increasing or decreasing each component value changed the filter response. After a long time going round and round in ever decreasing circles I found that this gave the best response but was still sullied by the peak at 1.2kHz but then found that the peak could be flattened by the Zobel.

Would I be correct in assuming that I've ended up with a very high filter Q which the Zobel then damps?

Is such a Q likely to have negative effects upon transient response?

Alex
 
A

alexclaber

Member
Joined 2005
sreten said:
Hi,

Well theoretical values for 500Hz and 6R give C around 40uF.
Values for L are ~ 6mH Q=0.5, ~3mH = 0.7 and ~ 1.5 for ~ 1.0.

i.e. the higher the value of L the more damped the filter.

Try omitting the zobel and increasing L.
Click to expand...

Thanks, that's very useful, makes sense. Have now tried doing that and although the response is better than without the Zobel and with a small inductor it still isn't as good as the previous iteration with the Zobel, due to a peak again happening at the filter knee. Could the high Qms (9.68) of the midrange be causing this peak?

Alex
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom