Precide Heil clone with better dynamics and MT blending
For the last ten years, two Precide models (pictures) have been my main speakers.
I have also used a REL sub for bass extension.
I absolutely love the dipole AMT tweeter: clean and dynamic HF and big soundstage.
However, I don't like the midbass dynamics and mid-tweeter blending.
Drivers are physically very close, and yet I can still hear the two different sources.
Since there aren't many similar commercial speakers, I have been trying to DIY a better design.
First I built fully open baffled prototypes. I could not get enough bass passively, so I bought a miniDSP. With active, I found that most drivers can't handle the EQ needed for full OB. Now I gave up on full OB.
My next project is to revisit Precide design and try to improve the midbass dynamics.
Tweeter will be exactly same configuration: Precide AMT or B&G Neo8s, dipole and naked (no baffle).
MT crossover will also remain around 1000Hz (active).
I will use a bigger, more dynamic and more efficient midbass driver,
preferably in a 30-40L sealed box (Precide cabinets are ported).
I will then highpass the midbass around 80-100Hz and add a large PA style subwoofer in a seperate BR cabinet (single 18" or dual 15").
Midbass driver can be 45 degrees tilted as in Precides, or direct radiating. I am open to suggestions.
I do not know much about boxed speaker design, please help me with this build.
First question is which midbass driver?
I am thinking something in the range of 8-12".
I already have a few vintage Seas alnico drivers in hand.
8" 21TV-GW BLACKCONES
8" 21TV-EW Vintagehifi.dk • Vis emne - Seas 21TV-EW Thiele-Small och Frekvens
10" 25TV-EW Vintagehifi.dk • Vis emne - Seas 25TV-EW TS och frekvensmätning
I really like the 21TV-GW for simple acoustic music, but it does not handle more complex passages well.
Shall I go for a more robust pro driver this time?
They say good things about the dynamics of Beyma 12P80nd and Liberty-8.
How about FaitalPro 12PR300?
18 Sound 10NMB420 or 8NMB420 (nice plots)?
For 30-40L sealed from 1000Hz down to 80-100Hz: what TS parameters to go for? Fs? Qts? mms? .
Does highest possible efficiency and Qms give best dynamics?
I have DSP EQ and efficiency matching does not matter because it is all active. I will also try to do a decent round-over starting from driver edge, so consider minimal baffle-step.
For the last ten years, two Precide models (pictures) have been my main speakers.
I have also used a REL sub for bass extension.
I absolutely love the dipole AMT tweeter: clean and dynamic HF and big soundstage.
However, I don't like the midbass dynamics and mid-tweeter blending.
Drivers are physically very close, and yet I can still hear the two different sources.
Since there aren't many similar commercial speakers, I have been trying to DIY a better design.
First I built fully open baffled prototypes. I could not get enough bass passively, so I bought a miniDSP. With active, I found that most drivers can't handle the EQ needed for full OB. Now I gave up on full OB.
My next project is to revisit Precide design and try to improve the midbass dynamics.
Tweeter will be exactly same configuration: Precide AMT or B&G Neo8s, dipole and naked (no baffle).
MT crossover will also remain around 1000Hz (active).
I will use a bigger, more dynamic and more efficient midbass driver,
preferably in a 30-40L sealed box (Precide cabinets are ported).
I will then highpass the midbass around 80-100Hz and add a large PA style subwoofer in a seperate BR cabinet (single 18" or dual 15").
Midbass driver can be 45 degrees tilted as in Precides, or direct radiating. I am open to suggestions.
I do not know much about boxed speaker design, please help me with this build.
First question is which midbass driver?
I am thinking something in the range of 8-12".
I already have a few vintage Seas alnico drivers in hand.
8" 21TV-GW BLACKCONES
8" 21TV-EW Vintagehifi.dk • Vis emne - Seas 21TV-EW Thiele-Small och Frekvens
10" 25TV-EW Vintagehifi.dk • Vis emne - Seas 25TV-EW TS och frekvensmätning
I really like the 21TV-GW for simple acoustic music, but it does not handle more complex passages well.
Shall I go for a more robust pro driver this time?
They say good things about the dynamics of Beyma 12P80nd and Liberty-8.
How about FaitalPro 12PR300?
18 Sound 10NMB420 or 8NMB420 (nice plots)?
For 30-40L sealed from 1000Hz down to 80-100Hz: what TS parameters to go for? Fs? Qts? mms? .
Does highest possible efficiency and Qms give best dynamics?
I have DSP EQ and efficiency matching does not matter because it is all active. I will also try to do a decent round-over starting from driver edge, so consider minimal baffle-step.
Attachments
Last edited:
I could be wrong, but I don't think it will work-out for you. 
Some people are just more sensitive to low vertical dispersion.
You can however *alter* the dispersion pattern, but it will alter the freq. response and overall efficiency of the Heil-type or Neo 8.
For altering vertical directivity:
Utilize the "foam" method of the larger RAAL's. google "ppi foam" and purchase a range of samples/ppi over 30 ppi that you can cut into those foam wedges. (..the higher the rating the more dense the foam, the greater the attenuation.)
Try the technique not only on the front, but also the front and back.. to see if one method is more preferable. (..I'm assuming you have measurement capability here with the minidsp.)
As to the driver, the more efficient the better. 8 db of sound pressure loss in a radial format is common (depending on baffle diffraction), and much more so if it's in the range (higher freq.) where the driver becomes more directive. (..which is why the Precide designs you've shown aren't full radials, but are rather slanted toward the listener somewhat.)
Low mass is desired, as is a high Qts for efficiency with respect to the lower mid-band and upper bass:
215 RTF64 SUPRAVOX
Some people are just more sensitive to low vertical dispersion.
You can however *alter* the dispersion pattern, but it will alter the freq. response and overall efficiency of the Heil-type or Neo 8.
For altering vertical directivity:
Utilize the "foam" method of the larger RAAL's. google "ppi foam" and purchase a range of samples/ppi over 30 ppi that you can cut into those foam wedges. (..the higher the rating the more dense the foam, the greater the attenuation.)
Try the technique not only on the front, but also the front and back.. to see if one method is more preferable. (..I'm assuming you have measurement capability here with the minidsp.)
As to the driver, the more efficient the better. 8 db of sound pressure loss in a radial format is common (depending on baffle diffraction), and much more so if it's in the range (higher freq.) where the driver becomes more directive. (..which is why the Precide designs you've shown aren't full radials, but are rather slanted toward the listener somewhat.)
Low mass is desired, as is a high Qts for efficiency with respect to the lower mid-band and upper bass:
215 RTF64 SUPRAVOX
Last edited:
Thank you ScottG; I was hoping that you would contribute.
I have already done some experiments with foam and felt on neo8s and it has been promising. The Precide AMT is much more compact, so there is less of a vertical dispersion problem. Overall, I am rather confident / comfortable above 1000Hz.
Now 100 to 1000Hz in 30-40L sealed enclosure...
So you think ideally I should be looking at high eff, highish Qts, and low mms.
The Supravox you have suggested is very similar to my SEAS 21TV-GW. Only difference is Supravox can probably handle a bit more power. I have listened to SEAS driver in its original sealed bookshelf cabinet and it is indeed very pleasant as long as it is not pushed hard.
So perhaps I should try the Seas I have first.
I had a 18 Sound 8MB400 in hand briefly. Nicely built robust PA driver. I listened to it naked and there was a lot of ringing. Vintage Seas is no way near robust, but it does not have any such nastiness when you listen to it naked. Why would that be? Rising response? Cone profile?
I have already done some experiments with foam and felt on neo8s and it has been promising. The Precide AMT is much more compact, so there is less of a vertical dispersion problem. Overall, I am rather confident / comfortable above 1000Hz.
Now 100 to 1000Hz in 30-40L sealed enclosure...
So you think ideally I should be looking at high eff, highish Qts, and low mms.
The Supravox you have suggested is very similar to my SEAS 21TV-GW. Only difference is Supravox can probably handle a bit more power. I have listened to SEAS driver in its original sealed bookshelf cabinet and it is indeed very pleasant as long as it is not pushed hard.
So perhaps I should try the Seas I have first.
I had a 18 Sound 8MB400 in hand briefly. Nicely built robust PA driver. I listened to it naked and there was a lot of ringing. Vintage Seas is no way near robust, but it does not have any such nastiness when you listen to it naked. Why would that be? Rising response? Cone profile?
If the Seas is similarly efficient with low mass - then sure. (..I'm not "sold" on any brand, it's more about what the individual driver will do for your for the particular application.)
A driver's break-up resonance is "set" by the structure of the diaphragm (it's size, thickness and shape), which is then damped by the diaphragm itself (internal loss of the material) followed by the surround and any compliant surface treatment. The 18Sound is a hard material straight cone with a fairly wide diameter VC (2").
As a radial though, it might not be a problem at all.
Pressure loss and combing effects when listened to further off-axis - tend to entirely mitigate this issue. In fact most of the time you are struggling to get more pressure so that you can crossover to a higher freq. to keep the tweeter from struggling.
Test your Seas driver at various angles with several cardboard-box prototypes to see what the resulting freq. response is like, and what you'll need to crossover to the tweeter. Of course also *listen* to the various angles as well.
The real master of this is Dan Neubecker - who is a forum member here and on other forums like HTGuide forum and Parts Express forum. After you've done of bit of prototyping - try to contact him and get his thoughts on the design direction.
Last edited:
Pictured speakers are intrinsically poor design; At low frequency they are effectively monopole, at crossover frequency they are briefly cardioid, and then continue to become dipole. In short polar pattern is a mess. This is very good reason why few commercial designs exist.
Your own observations of hearing speakers as two different sources with poor blending is exactly what any commercial designer would encounter. Do what they would: abandon it. Whole design is showcase for tweeter. Tweeter appears to have plenty of potential, but not as pictured.
Ten years is much too long.
Consistent feature across all top ranked speaker designs for domestic listening is smooth, wide, and uniform radiation pattern.
Your own observations of hearing speakers as two different sources with poor blending is exactly what any commercial designer would encounter. Do what they would: abandon it. Whole design is showcase for tweeter. Tweeter appears to have plenty of potential, but not as pictured.
Ten years is much too long.
Consistent feature across all top ranked speaker designs for domestic listening is smooth, wide, and uniform radiation pattern.
Lets revisit the dipole mid option one more time (narrow or no baffle).
More specifically 200-1000Hz.
I can think of 3 options:
1. Cone driver, low eff, low Fs, high Xmax, Qts ~0.3, lots of EQ
2. Cone driver, high eff, high Fs, low Xmax, Qts ~0.7, less EQ
3. Neo10, very high Qts, least EQ
I have tried option 1 and 2 and found them very limiting.
I am willing to give Neo10 a try too.
More specifically 200-1000Hz.
I can think of 3 options:
1. Cone driver, low eff, low Fs, high Xmax, Qts ~0.3, lots of EQ
2. Cone driver, high eff, high Fs, low Xmax, Qts ~0.7, less EQ
3. Neo10, very high Qts, least EQ
I have tried option 1 and 2 and found them very limiting.
I am willing to give Neo10 a try too.
Let's assume I use Precide dipole AMT tweeter naked for over 1000-1500Hz,
and a decent dynamic separate bass section.
Which is the better option to fill in between:
Option 1:
Naked Neo10, EQed flat, and steep crossover around 200Hz
Option 2:
8-12" cone in a 30-40L closed box, and crossover around 100Hz
and a decent dynamic separate bass section.
Which is the better option to fill in between:
Option 1:
Naked Neo10, EQed flat, and steep crossover around 200Hz
Option 2:
8-12" cone in a 30-40L closed box, and crossover around 100Hz
High distortion with #1. (..largely due to the low crossover freq. in combination with the pressure loss from not having a baffle.) Now if it were on/in a baffle of reasonable dimensions and not played to loudly - then yes. Model baffles in the Edge:
Tolvan Data
Option 2 will be tough on the AMT tweeter that low, but with an 8" driver (allowing a slightly higher crossover freq. for the AMT) that has good off-axis performance and is reasonably efficient, sure (..not 10" or 12"). You could sort of "suspend" the AMT over the box a'la Ascendo System M-S.
Considering the use of minidsp.. with 8" drivers:
http://www.parts-express.com/pe/showdetl.cfm?partnumber=295-345
or,
and likely triple the cost:
http://www.enviee.de/bass/index.html
Last edited:
ScottG, thanks for the Neo10 advice.
It is easy to test no baffle set-ups.
I will try Neo8s and a 12" driver combination first.
I would like to use Beyma SM112.
But I came across a pair of used Faital Pro 12PR300 for half the retail price and went for it.
I will see how it works covering 100-800Hz nude.
It is easy to test no baffle set-ups.
I will try Neo8s and a 12" driver combination first.
I would like to use Beyma SM112.
But I came across a pair of used Faital Pro 12PR300 for half the retail price and went for it.
I will see how it works covering 100-800Hz nude.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.