I have registered in the Dahlquist Yahoo Group and have read the patents. Jon Dahlquist had grasped the importance of timing errors in multi way dynamic speakers back then. He talked about common large baffle diffraction problems and need for coherent rise time between sources.
He also talked about the staggered baffle sizes. He used them to aid the lower extension of each driver's passband. But before they created diffraction effects in the upper range he said. He integrated their lowering extension effects into his crossover design. He must have had some rise time criterion that he measured when designing the relative distances of the sources in the Z plane.
To grasp all that back then, is brilliant! I can imagine him performing literally hundreds of trials with a square wave generator a microphone, and an oscilloscope!
Ok, today we would talk about impulse response, edge diffraction, radiation centers, cardioid mids etc.
We could probably find many mistakes, do it better with the benefit of hindsight, FFT, internet etc.
But the dynamic speaker ground that Jon Dahlquist broke back then, and his guts to bring it down to an actual product that is still considered excellent, and worthy of a restoration cult, is amazing!
To be fair to the audio patriarchs, he is the interpreter of Quad's founder Peter J. Walker in the voice coil domain IMHO. But that was no mean feat!
RESPECT!
He also talked about the staggered baffle sizes. He used them to aid the lower extension of each driver's passband. But before they created diffraction effects in the upper range he said. He integrated their lowering extension effects into his crossover design. He must have had some rise time criterion that he measured when designing the relative distances of the sources in the Z plane.
To grasp all that back then, is brilliant! I can imagine him performing literally hundreds of trials with a square wave generator a microphone, and an oscilloscope!
Ok, today we would talk about impulse response, edge diffraction, radiation centers, cardioid mids etc.
We could probably find many mistakes, do it better with the benefit of hindsight, FFT, internet etc.
But the dynamic speaker ground that Jon Dahlquist broke back then, and his guts to bring it down to an actual product that is still considered excellent, and worthy of a restoration cult, is amazing!
To be fair to the audio patriarchs, he is the interpreter of Quad's founder Peter J. Walker in the voice coil domain IMHO. But that was no mean feat!
RESPECT!
chrisb said:
I can envisage a clone, which does not change the polar behavior between woofers and mid-woofer, so to cancel the original DQ-10 sonic discontinuity there as Gordon Holt spotted. That 600Hz cross between omni and semi cardioid was most likely causing the problem. That is why Jon Dahlquist was at a loss, by researching just the bass alignment. Also, much better sensitivity and nice impedance can be done today. A four way is very doable with much better drivers in the detail department too. Alignment and diffraction control can be well attained with today's knowledge and digital tools. As for the crossover, I would keep it first order series.
In the early 1980's I fell in love with some DQ-10s but couldn't afford them as a student. So I think Salas' idea of recreating the sound as a DIYer is excellent.
I already have the 15" version of the Goldwood woofers Salas mentioned and think that the Eminence Alpha 15As are are a much better bang for the buck. So I'd argue for one Alpha 15A per side.
I tried modeling this using Martin King's OB worksheet for two drivers and the following parameters:
- the Alpha 15A and TangBand W6 789S Salas mentioned above
- a 32 x 22" baffle (no stand included),
- the Alpha 15A centered 10" up from the bottom
- mid-woofer directly above the Alpha off the baffle
- 1st order crossover at 200Hz as suggest by Salas
This is the result:
As the TangBand is off the baffle its response drops off earlier but the result is still very good. The modeled impulse response appears to be excellent. [Note that the model assumes no rear wall]. Adding a 6" stand flattens the response curve. Going up to a 12" stand actually flattens the woofer response but requires minor crossover changes.
I already have the 15" version of the Goldwood woofers Salas mentioned and think that the Eminence Alpha 15As are are a much better bang for the buck. So I'd argue for one Alpha 15A per side.
I tried modeling this using Martin King's OB worksheet for two drivers and the following parameters:
- the Alpha 15A and TangBand W6 789S Salas mentioned above
- a 32 x 22" baffle (no stand included),
- the Alpha 15A centered 10" up from the bottom
- mid-woofer directly above the Alpha off the baffle
- 1st order crossover at 200Hz as suggest by Salas
This is the result:
An externally hosted image should be here but it was not working when we last tested it.
As the TangBand is off the baffle its response drops off earlier but the result is still very good. The modeled impulse response appears to be excellent. [Note that the model assumes no rear wall]. Adding a 6" stand flattens the response curve. Going up to a 12" stand actually flattens the woofer response but requires minor crossover changes.
The Alpha crossed my mind too. Nothing against it. Your cross looks over 300Hz to me. Use a little individual baffle for the TB and cut at 200Hz. Try to de-bump the low mids.
Because there is going to be a rear wall, I suggest thick foam behind the Alpha, and thick felt behind the TB as per the original mid-woofer.
Slow transition to cardioid will be achieved. Excellent for music's power range in a small room. Get those bands acoustically right, and the original DQ-10 will be a poor relative to the brave brand new -sensitive- clone.
Because there is going to be a rear wall, I suggest thick foam behind the Alpha, and thick felt behind the TB as per the original mid-woofer.
Slow transition to cardioid will be achieved. Excellent for music's power range in a small room. Get those bands acoustically right, and the original DQ-10 will be a poor relative to the brave brand new -sensitive- clone.
Attachments
salas said:
Unfortunately the TangBand's low Qts (0.37) means that even in a 12 x 12" baffle the acoustic response will begin to naturally drop off higher. Adding the 200Hz 1-order electrical filter doesn't help... If you don't add the filter you'll hit the xmax limit of the TangBand with as little 3 watts. I'd actually go with a 300 Hz or higher 2-order (electrical) filter here or a higher cross with a 1st order filter.
I'd be worried that the cardiod response you're looking for will at least partially disappear as the materials won't uniformly reduce the rear output across the driver's frequencies. But I can't back this up with a reference or by modeling with Martin's worksheets.
Note that the modeled response with the rear wall set to be 1 m away is still very good:
An externally hosted image should be here but it was not working when we last tested it.
I have also thought that many improvements could be made on the DQ10, however it is also interesting to take a close look at the original design.
The woofer was from the large Advent. Interesting that Advent stopped making it with the masonite adapter some time around the mid 1970s. It seems that Dahlquist continued to have them made and used them in the DQ20 as can be seen in these samples dated 1987:
http://www.classicspeakerpages.net/IP.Board/index.php?s=&showtopic=2931&view=findpost&p=72351
The woofer was from the large Advent. Interesting that Advent stopped making it with the masonite adapter some time around the mid 1970s. It seems that Dahlquist continued to have them made and used them in the DQ20 as can be seen in these samples dated 1987:
http://www.classicspeakerpages.net/IP.Board/index.php?s=&showtopic=2931&view=findpost&p=72351
Dq-10 DIY clone
Hello Salas! I'm the new guy to DIY. I just got done with an experiment based on the Dahlquist dq-10 speakers. My uncle has had a pair of dq-10's since 1985 when I was 11 years old. They are some of the best sounding speakers I've ever heard. My Aunt (his wife) was Frank Zappa's secratary for Barking Pumpkin records during this time. So I got to hear all kinds of music through the dq-10's. I haven't found any close to me where I live, so I decided to try to come up with my own version of a Dq-10. I collect vintage audio equipment, so I have a lot of speakers and electronic to experiment with. So here is what I used: For the Advent woofer I used a JBL L-19 with the 8" woofer, for the midbass I used a JBL 52T 4.5" driver and disconnected the tweeter, for the midrange I used a Vifa 4" midbass driver, for the tweeter I used an old Polk audio 1" driver from a pair of monitor 7's that someone gave me years ago because they didn't like the sound of them, and for the supertweeter I actually bought an old pair of motorola's from a guy in Whales off of eBay that he took out of his dq-10's to do the ribbon upgrade. I changed the crossover values a little bit to get a better roll off between the drivers to get a full parametric sound achievement. If your interested in the schematic diagram let me now, I'll be glad to post it. I got amazing results and they sound incredible. I stuck with the same crossover concept as Jon D. but I modified it to my specs. The midbass and midrange are in boxes but the tweeters and supertweeters are open baffle. I will soon build an open baffle for the midbass and midrange soon. Oh yeah, I am using a marantz 2252 fully restored and upgraded with LED's and a cooling fan for the heat sinks.
Hello Salas! I'm the new guy to DIY. I just got done with an experiment based on the Dahlquist dq-10 speakers. My uncle has had a pair of dq-10's since 1985 when I was 11 years old. They are some of the best sounding speakers I've ever heard. My Aunt (his wife) was Frank Zappa's secratary for Barking Pumpkin records during this time. So I got to hear all kinds of music through the dq-10's. I haven't found any close to me where I live, so I decided to try to come up with my own version of a Dq-10. I collect vintage audio equipment, so I have a lot of speakers and electronic to experiment with. So here is what I used: For the Advent woofer I used a JBL L-19 with the 8" woofer, for the midbass I used a JBL 52T 4.5" driver and disconnected the tweeter, for the midrange I used a Vifa 4" midbass driver, for the tweeter I used an old Polk audio 1" driver from a pair of monitor 7's that someone gave me years ago because they didn't like the sound of them, and for the supertweeter I actually bought an old pair of motorola's from a guy in Whales off of eBay that he took out of his dq-10's to do the ribbon upgrade. I changed the crossover values a little bit to get a better roll off between the drivers to get a full parametric sound achievement. If your interested in the schematic diagram let me now, I'll be glad to post it. I got amazing results and they sound incredible. I stuck with the same crossover concept as Jon D. but I modified it to my specs. The midbass and midrange are in boxes but the tweeters and supertweeters are open baffle. I will soon build an open baffle for the midbass and midrange soon. Oh yeah, I am using a marantz 2252 fully restored and upgraded with LED's and a cooling fan for the heat sinks.