Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Multi-Way Conventional loudspeakers with crossovers

Beyond the Ariel
Beyond the Ariel
Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 23rd July 2015, 02:28 AM   #12941
Pano is offline Pano  United States
diyAudio Moderator
 
Pano's Avatar
 
Join Date: Oct 2004
Location: SW Florida
Beyond the Ariel
OK, I'll start a new thread tonight.
  Reply With Quote
Old 23rd July 2015, 05:39 AM   #12942
Pano is offline Pano  United States
diyAudio Moderator
 
Pano's Avatar
 
Join Date: Oct 2004
Location: SW Florida
Beyond the Ariel
Phantom center fix thread is here
Fixing the Stereo Phantom Center
  Reply With Quote
Old 23rd July 2015, 06:19 PM   #12943
Scott L is offline Scott L  United States
diyAudio Member
 
Join Date: Dec 2008
Location: Knoxville, TN
Beyond the Ariel
Default Crossover

Lynn/ Gary:

I'm sorry if this has already been covered, but what are you doing for a crossover between the Altec and the horn?

Last I remember, it was @ 700Hz, but what type/slope ?
  Reply With Quote
Old 24th July 2015, 03:30 AM   #12944
g3dahl is offline g3dahl  United States
diyAudio Member
 
g3dahl's Avatar
 
Join Date: Jun 2004
Location: Silverdale, WA
It hasn't been covered in any detail, really. The original goal was a nominal frequency of 700 Hz. We were aiming for a curve shape somewhere between Butterworth and Bessel. I began with cookbook part values and made measurements. With the woofer, the acoustical rolloff was way too high. I tweaked the values until everything added up to the nicest-looking combination of flat passband response and smooth rolloff. Measured crossover frequency ended up closer to 900 Hz. I found that I could get a much nicer response shape if I didn't try to move it lower so I left it alone.

The HF section was at first approached in a similar manner. The original crossover was created with the regular Radian 745-PB driver (ferrite magnet, aluminum diaphragm). It utilized a 3rd-order filter, Zobel, swamping resistor and Slagle auto former attenuator. Response was quite flat to 17 kHz, but I must admit that I wasn't really satisfied with the tonal character of the upper octaves.

When the 745NeoBe became available, I snapped up a pair in hopes that the beryllium diaphragm's performance advantages would result in a more pleasing result. Without adjustment to the crossover, results were mixed. Its lower range had somewhat less output, and there was a broad peak around 4-5 kHz. Above this, response rolled off in a quite linear manner. For some time, I used the RAAL Lazy Ribbon tweeter to cover the top end.

In offline conversations, Pierre and I compared notes about the integration of the compression driver and ribbon. Eventually, he reported optimum results using passive equalization of the 745 driver, sans the ribbon. He shared his schematic with me, but since our autoformers (and woofers) were different, the actual values of the parts didn't apply in my case. I took my speakers outdoors so I could make some clean measurements (at least when the neighbor's lawn mower wasn't running...) and worked to optimize the crossover as best I could.

After some reworking, I was really pleased with the results on the LF section. As for the HF section, Pierre's approach brought about remarkable results. After optimizing part values for my system, the simple circuit was able to simultaneously lift the LF droop in the CD's response, level the peak, and flatten the previously rolled-off HF region. Final response went smoothly to somewhere around 14 kHz or so. Yet I wan't quite satisfied with the way the LF and HF sections summed; there was a wrinkle in the response that I still believe could be ameliorated by raising the turnover frequency of the HF section.

I wasn't able to improve it with capacitors on hand, and didn't have inductors of suitable value for further experimentation, so I brought the system inside for a listen, not knowing what to expect. Wow, just wow. Even without the last top half-octave, the entire upper half of the frequency range sounded so beautifully natural, and the imaging was greatly improved. The drivers integrate with delightful seamlessness.

It wouldn't hurt my feelings if it gets better than this (!), but it exceeded my expectations for this stage of the project, and is definitely a plausible stopping point. As an amateur hobbyist speaker builder, I am frankly pretty darn proud to have reached this level of musical realism, with the help of Lynn, Pierre and others on this board. But there is most certainly room for improvement. I look forward to taking it further, and benefiting from the contributions of other DIY enthusiasts.

In its current form, the crossover is the best I have been able to do using the combination of drivers, enclosures and parts on hand. It is certainly not optimized for any other combination of drivers and enclosures. Anyone building this project should expect to make their own measurements and adjustments--it was most definitely needed in my case.

Click the image to open in full size.

Gary Dahl
__________________
"So many tubes, so little time..."
  Reply With Quote
Old 24th July 2015, 05:05 AM   #12945
swak is offline swak
diyAudio Member
 
Join Date: Jul 2003
Location: .
That is a very interesting report. I think the process points out the importance of a well-designed crossover to get the most out of the drivers and configuration.
What is the advantage of using the autoformer? Is it custom made or off the shelf?
It would be very interesting to see what the modifications from Pierre do to the CD’s response (modeled and measured).

Regarding subs for me the question is rather if 4 10”s in an average room (25-36m2) work better than two 15”s integrated with the main speakers. What is the max recommended low pass frequency for multiple subs (given certain room dimensions)?

S.
  Reply With Quote
Old 24th July 2015, 05:44 AM   #12946
g3dahl is offline g3dahl  United States
diyAudio Member
 
g3dahl's Avatar
 
Join Date: Jun 2004
Location: Silverdale, WA
The combination of the autoformer and the swamping resistor serves to isolate the crossover from the large impedance peak just below the horn's LF limit; in this case, about 350 Hz. With the swamping resistor in parallel with the the driver/autoformer combination, the crossover works into a relatively flat impedance. The large difference in sensitivity between the woofer and the driver calls for a significant amount of attenuation of the latter. To my ears, autoformer attenuation (of compression drivers) brings a smaller sonic penalty than that of resistors.

There are other ways to do this, especially with the equalization. Martin has reported excellent results using equalized crossover circuits without autoformers. I haven't tried any of this, but have an open mind.

Pierre's modification consists of the resistor-capacitor combination between the autoformer's output and the driver; in the original it was just a wire. He described to me the general effect of varying the values but left me to determine the optimum combination for my circuit. Sorry to say I didn't save the curves I made during the process; because of circumstances I had to work fast, and honestly I didn't expect my results that day to be as sonically successful as they were.

Undoubtedly someone with technical chops beyond mine will take this design to greater heights.

My autoformers (Slagle) were inadvertently made with a couple of missing taps, so I guess they're really not off the shelf. Luckily, the missing taps aren't of consequence to me.

Gary Dahl
__________________
"So many tubes, so little time..."
  Reply With Quote
Old 24th July 2015, 06:05 AM   #12947
Truetone is offline Truetone  Australia
diyAudio Member
 
Join Date: Apr 2008
When using LTSpice to simulate a cross over, how should one model the driver? I just measure the Z with a suitable meter at the nearest possible point and enter the value as an R in the model. (in my case it was 1000Hz). I'm a learner driver but it seems a really nice way to tune the xo, eg for Pierre's add on. But then the Z of the driver is probably different when this takes effect.
  Reply With Quote
Old 24th July 2015, 12:51 PM   #12948
Pano is offline Pano  United States
diyAudio Moderator
 
Pano's Avatar
 
Join Date: Oct 2004
Location: SW Florida
Beyond the Ariel
I suppose you could do it in Spice, but there are easier ways. Programs like XSim or PCD are meant for simulating passive crossovers. And they are free.

To get accurate results, you should have impedance measurements and FR measurements to enter into the sims. Doing that, the results are pretty accurate.
__________________
Take the Speaker Voltage Test!
  Reply With Quote
Old 24th July 2015, 02:16 PM   #12949
PierreQuiRoule is offline PierreQuiRoule  Canada
diyAudio Member
 
PierreQuiRoule's Avatar
 
Join Date: Feb 2004
Location: Ottawa
Beyond the Ariel
I modelled everything in LTSpice. Ultimately the accuracy is limited by my poor models of the horn and autoformer. Still, simulations were useful for understanding the effect of various changes to the crossover configuration.

Gary is a brave man purchasing the Radian Neo's as soon as they came out. I waited and bought mine only after reading his positive review.

Pierre
  Reply With Quote
Old 24th July 2015, 02:48 PM   #12950
g3dahl is offline g3dahl  United States
diyAudio Member
 
g3dahl's Avatar
 
Join Date: Jun 2004
Location: Silverdale, WA
Of course I made impedance curves and frequency response measurements of my drivers in their horns and enclosures before commencing development and optimization of the crossovers. But I have no skills in modeling, so all I could do was build and make measurements of the real thing, then make repeated cycles of further modifying and measuring.

Dave Slagle's LTSpice models of crossovers using autoformers with swamping resistors made a great resource when I was working on the original crossover. It would be interesting to see the effect our EQ in his model (I didn't ask). But at least it was very easy to see the effect of part value changes on the actual speaker's response.

Gary Dahl
__________________
"So many tubes, so little time..."
  Reply With Quote

Reply


Beyond the ArielHide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


New To Site? Need Help?

All times are GMT. The time now is 07:40 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.79%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio
Wiki