New SB Acoustics Reference design published (writeup is still a work in progress)
Design was started (and drivers obtained) before Zaph's SB 12.3 design was published, but once we saw Zaph's design it was obvious that comparisons would be inevitable.
At one point we had a full passive crossover, but the 7.5mH/400uF capacitor in the resonant peak filter for the SB15 "midranges" can get really expensive if high quality components are used. We decided active is a better way to go.
Some of the details of this design are specific to the club environment (40" tweeter height and larger cabinets to compensate for the relatively high output impedance of our clubs' Tube amps.)
Evolving writeup here at the Audio Society of Minnesota:
https://sites.google.com/site/audiosocietyofminnesota/Home/diy-projects/sb-acoustics-reference
Feel free to share any opinions or questions.
Design was started (and drivers obtained) before Zaph's SB 12.3 design was published, but once we saw Zaph's design it was obvious that comparisons would be inevitable.
At one point we had a full passive crossover, but the 7.5mH/400uF capacitor in the resonant peak filter for the SB15 "midranges" can get really expensive if high quality components are used. We decided active is a better way to go.
Some of the details of this design are specific to the club environment (40" tweeter height and larger cabinets to compensate for the relatively high output impedance of our clubs' Tube amps.)
Evolving writeup here at the Audio Society of Minnesota:
https://sites.google.com/site/audiosocietyofminnesota/Home/diy-projects/sb-acoustics-reference
Feel free to share any opinions or questions.
you are right, 400 uF looks like a bad idea, not because of parts quality,
rather causing a near dead short on the impedance graph, if I understood
correctly. You were referring to the passive low pass filter of the woofers?
The resonant peak filter was a series notch 7.5mH. 400uF and 4 ohms (minus inductor DCR) in parallel with the SB15 woofers - so not a dead short by any means. I was speaking of passive "highpass" of the SB15 5" midwoofers, which have very tall impedance peaks - shown are measurements with socks loosely plugging ports, so more of a quasi- aperiodic than strictly sealed, which should actually reduce the impedance peaks a bit.
So, you were trying to null the impedance resonant
frequency peak of the midwoofers in the box ?
Yes. If you don't use a peak filter, a passive highpass doesn't work very well Measured impedance was quite flat using only the peak filter, so it was designed properly It is a COMPLETELY moot point since the final design didn't use it...
Ron E,
I am affraid you are not right about nulling impedance peak as
a necessity. You can do whatever passive high pass filter you desire
without building RLC circuit for nulling impedance peaks. I have
been there and done that. RLC has this unwanted effect that it makes
amplifier waste more power for nothing. This kind of circuits are
normally placed in parallel to amplifier output and there they alter
impedance as wanted but do not change FR response.
Wish you success !
I am affraid you are not right about nulling impedance peak as
a necessity. You can do whatever passive high pass filter you desire
without building RLC circuit for nulling impedance peaks. I have
been there and done that. RLC has this unwanted effect that it makes
amplifier waste more power for nothing. This kind of circuits are
normally placed in parallel to amplifier output and there they alter
impedance as wanted but do not change FR response.
Wish you success !
Zaph also does it in his SB big three way and shows what nulling the impedance peak does to the frequency response.
In general nulling an impedance peak will have no effect on the frequency response, if and only if, there is no filter trying to act on the driver in the frequency range that the impedance nuller is going to have its effect on. If there is a filter, such as a high pass, then it has a considerable effect and it is absolutely necessary for the correct function of said filter.
In general nulling an impedance peak will have no effect on the frequency response, if and only if, there is no filter trying to act on the driver in the frequency range that the impedance nuller is going to have its effect on. If there is a filter, such as a high pass, then it has a considerable effect and it is absolutely necessary for the correct function of said filter.
pete, its on madisound page and link to zaph's
The Madisound Speaker Store
Zaph|Audio - SB12.3 3-Way Tower
The Madisound Speaker Store
Zaph|Audio - SB12.3 3-Way Tower
Ron E,
I am affraid you are not right about nulling impedance peak as
a necessity. You can do whatever passive high pass filter you desire
without building RLC circuit for nulling impedance peaks. I have
been there and done that. RLC has this unwanted effect that it makes
amplifier waste more power for nothing. This kind of circuits are
normally placed in parallel to amplifier output and there they alter
impedance as wanted but do not change FR response.
Wish you success !
Hi,
You are completely wrong regarding this particular case.
A passive x/o that can handle such mid peaks is more
complicated than simply nulling the driver peak, and
has nothing to do with wasting amplifier power.
rgds, sreten.
Zaph uses 7mH and 600uF for the trap. The obvious way to get rid
of most of the expensive x/o components is to biamp with passive
line level crossovers, bass to mid, very doable :
However if the mid/treble amplifier has significant impedance,
it complicates things regarding the mid units impedance peak.
Last edited:
Does the high pass filter work better with or without an impedance peak nulling ?
Yes. If you don't use a peak filter, a passive
highpass doesn't work very well
Attachments
Thanks Henry,
so it appears that where the Z nulling is done in Zaph & Ron's designs is a bit different from how it's done in the Elsinore? (schematic post 1673):
http://www.diyaudio.com/forums/multi-way/97043-elsinore-project-thread-168.html
so it appears that where the Z nulling is done in Zaph & Ron's designs is a bit different from how it's done in the Elsinore? (schematic post 1673):
http://www.diyaudio.com/forums/multi-way/97043-elsinore-project-thread-168.html
pics with no explanation
Here's one for those who don't understand these circuits. After this, I'll be happy to educate you on how and why these are needed, but if you want to continue to hijack this thread and argue needlessly about resonant peak filters, please start a new thread.
Attached are pics with circuit and the resultant "midrange" frequency response, the SB15 is being crossed over highpass within an octave or so from its resonant peak - One pic is with the filter, the other with the series resistor raised to a very high value to take it out of circuit.
As you can plainly see, the impedance peak filter makes the crossover work properly. Without it you get peaking.
Attachments
Thanks Henry,
so it appears that where the Z nulling is done in Zaph & Ron's designs is a bit different from how it's done in the Elsinore? (schematic post 1673):
A resonant peak filter acts over a narrow range. Those two filters in the schematic you linked to seem to (I did not read beyond that page) filter two peaks rather than one, and they aren't there for the crossover, just to make the input impedance flatter. That is almost always a waste of money, unless you just want flat impedance for some reason Tube amps tend to do better with flatter (and higher) impedance.
Last edited:
After this, I'll be happy to educate you on
how and why these are needed,
Ron,
I was very clear with pictures I posted on what I meant by saying
what I've said. Now that you have clearly showed where you have
placed your RLC circuit, there is no misunderstanding.
Ron,
I was very clear with pictures I posted on what I meant by saying
what I've said. Now that you have clearly showed where you have
placed your RLC circuit, there is no misunderstanding.
Your pics only make it clear that you didn't understand what I was saying and were ramming your position home regardless of need. A person who doesn't understand crossover design might imagine that your pics prove that I was doing something unnecessary.
In other words, you were doggedly defending a position you didn't hold, based on information you didn't have. If you were unclear on anything I had written, you could have just consulted the write up, which makes it a bit clearer, or asked me a question.
Last edited:
Equalizing the impedance of the speaker from a parallel network before the crossover doesn't make much sense with voltage source amps. It is, however, necessary for correct shaping of some XOs when applied AFTER the crossover to normalize the driver, not the system, impedance.
Ron, any more details on the woofer box size/tuning?
cheers,
Pete McK
Pete, as I recall its about 100 liters tuned to 25 Hz for the pair of 8"ers. They go plenty deep, to put it mildly. Very tight and tuneful as well. I'll try to get some more details on the bass cabinets put up on the ASM webpage (www.audiomn.org) over the weekend. I found some of my original "ASM baffle" measurements of the individual drivers and a nearfield with port sum that I did on the 8"ers.
I will probably start a new thread here about these speakers, as this thread has been completely derailed by people who want to argue about a filter (absolutely necessary in a passive highpass ~80-150Hz with the 5"ers) that wasn't even used in the final (active) design...
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Loudspeakers
- Multi-Way
- SB Acoustics reference 3-way by Audio Society of Minnesota