Dipole with BMS Accuton and B&G

[Martijn_H]

Hello, this is my first post on this board!

I am building a three way dipole with the BMS15N840, the Accuton C90 (173N-C90/6) and the B&G Neo3pdr.

The dipole is a semi U-frame for the woofers, so half cardioid, half 8 figure behavior.


crossovers will be: 250, 2kHz.

pict2168dy2.jpg Image is not available

My first question!

The woofers are very sensitive! I estimate +/- 102dB together. (97dB/1watt a piece) The mid unit is 93dB.

Dipole correction needs to be done untill 125Hz.

Is it possible to work with a very large coil of about 12mH to use all the sensitivity of the woofers to get very low and still be able to cross at 250 to the mid? I don't like to put resistors in serie or par. with the woofers.

Wenn I go 2,5 octaves lower than the 125 Hz I'll get to +/- 93dB.

Is there someone else doing something with dipole?

Greetings,

Martijn (the Netherlands)

[Dennis H]

Beautiful speakers, Martijn! A large coil is certainly a good option. Do you have crossover design software? When you get measurements of all the drivers, you should be able to model it and see how it looks. Don't forget to include the DC resistance of the coil. A little DCR can be helpful for shaping the bottom end as it raises the Q of the woofers.

[ThomasW]

Hello, this is my first post on this board!........
Is there someone else doing something with dipole?

Hi and welcome.... :T

Nice looking speakers...

Yes we do quite a bit with dipoles......

https://www.htguide.com/forum/showthread.php?t=23598&page=14&highlight=isiris

https://www.htguide.com/forum/showthread.php?t=7644&highlight=Arvo+Part

That's a pretty big miss-match in efficiency between your woofers and the midrange. Over the years I've found it's easier to design and setup dipoles using active electronics. Do you have access to something like a Behringer DCX2496? For prototyping they're are great...

[Martijn_H]

Dennis, I sometimes use Boxsim. I measure with Arta and speakerworkshop.

Thomas, Yes I do have a dxc, but I don't like the quality for long term use. When the cabinets are finished I'll use it to see how things work out. For a prototype very handy I guess!

My first plan was to go active for the woofers and passive for the mid and tweeter. After hearing a complete passive dipole I'd like to give it a try!

102vs93dB is not easy to handle, but it allows you to go very low with the dipole. The hardest thing is reaching from a very low 25 to the higher 250 Hz with the woofers, but I think it can be done. (I hope! :roll: ) They have enough headroom!

Response at different room locations (SPL is not calibrated and there was an airoplane at20Hz red line)

responswoofersindefframdd1.png Image not available

[ThomasW]

The hardest thing is reaching from a very low 25 to the higher 250 Hz with the woofers, but I think it can be done. (I hope! :roll: ) They have enough headroom!

Correct.......

My experience is that getting good output below ~50Hz with a dipole is very problematic unless one has a HUGE amount of Vd and uses TONS of EQ.

One experiment I did was to compare the Arvo Ultra to the floor mounted IB sub in my family room.





The Arvo Ultras had 8-12" drivers, the IB 4 of the same 12" drivers. At no time and regardless of the amount of EQ used could the dipoles provide the same bottom end as the IB.

[Martijn_H]

Lot's of VD or a baffle design that allows you to go loud! The dipole in the picture hasn't got a very wide baffle as far as I can see from the picture and will problably have troubles in the very lows. Tell me when I am wrong, I can't see the back. My dipolebaffle gives me 104dB output a piece without 6dB vloorgain at 35Hz, which I think is more then ebough. Especially the U frame helps me with that, and don't forget the 14mm xmax of the BMS ofcourse. I'don't want to say that the dipole in the picture is not good for lows, but it definitively has a different design philosophy for radiation patern and LF output. It will be more symmetric when front compared to back(in the case it has no wings), in the contrast with my (short) U-frame wich is more cardioid and because of that more powerfull in the lows but les symmetric. Both have their good and bad sides I guess.. When I compare my dipoles with my BMS18N850 clossed subwoofers, they won't win SPL contests, but that isn't the goal with dipole.. They are just better than boxes, but with less SPL! :W :twisted:

So output isn't the problem. When I try to take the dipole correction from 30 to 250Hz, I'm afraid that above 125 Hz to much energy is getting lost what makes xo at 250 difficult. That is because correction is only needed till 125Hz.

Maybe tomorrow I'll try an xo for the woofers! When things don't work out passive I'll hop over to active..

All experience with dipoledesigning is welcome! :wink:

Greetings,

Martijn

[pixelpusher]

Nice Baffles!
I built a dipole system using the B&G Neo3pdr, PHL 1120 midrange and four Lambda DP15 on each side. When using the Neo3pdr with no back cup it has a peak around 10k. After EQ the Neo3pdr is down to about 85db spl. The mid and woofers will need to be padded down quite a bit.

[Martijn_H]

Nice Baffles!
I built a dipole system using the B&G Neo3pdr, PHL 1120 midrange and four Lambda DP15 on each side. When using the Neo3pdr with no back cup it has a peak around 10k. After EQ the Neo3pdr is down to about 85db spl. The mid and woofers will need to be padded down quite a bit.

Did you use a 0,09/0,1mH coil in serie with the dipole B&G? It is possible to get the curve very nice without the peaks! 0,65mH and someting like 8,5uF and offcourse the correction with the small coil.

edit: I think you mean the peak at 15kHz. That might need a small notch or something..

[Dennis H]

Tell me when I am wrong, I can't see the back. My dipolebaffle gives me 104dB output a piece without 6dB vloorgain at 35Hz, which I think is more then ebough. Especially the U frame helps me with that, and don't forget the 14mm xmax of the BMS ofcourse.

Okay, I'll be the one to tell you you're wrong. Thomas's speakers have deeper U-baffles than yours and his drivers move much more air. The BMS is only 7mm Xmax 1-way (the standard way it's measured) compared to several time that for Thomas's drivers. I think you're WAY overestimating your SPL at 35 Hz. Have you actually measured that or is it a calculation? Remember you can't measure a dipole nearfield. You need to back the mic far enough away so the backwave can blend. When you do that, the bass dies, like 15-20dB.

You have Speaker Workshop so you can answer all your questions with its crossover simulator once you have some good measurements from ARTA at 1-2 meters.

[pixelpusher]

Did you use a 0,09/0,1mH coil in serie with the dipole B&G? It is possible to get the curve very nice without the peaks! 0,65mH and someting like 8,5uF and offcourse the correction with the small coil.

edit: I think you mean the peak at 15kHz. That might need a small notch or something..

Yes, I checked my files. I had a 8db peak at 14kHz.
-pp

[Martijn_H]

Okay, I'll be the one to tell you you're wrong. Thomas's speakers have deeper U-baffles than yours and his drivers move much more air. The BMS is only 7mm Xmax 1-way (the standard way it's measured) compared to several time that for Thomas's drivers. I think you're WAY overestimating your SPL at 35 Hz. Have you actually measured that or is it a calculation? Remember you can't measure a dipole nearfield. You need to back the mic far enough away so the backwave can blend. When you do that, the bass dies, like 15-20dB.

You have Speaker Workshop so you can answer all your questions with its crossover simulator once you have some good measurements from ARTA at 1-2 meters.

Sorry, I forgot to mention the measuringmethod: mic 100cm and dipole 200cm of the wall.

What do you define as xmax Dennis?

I find it very hard to say exactly what lineair xmax is. I don't think any driver stays very lineair at several mm's..

BMS15N840
Voice coil winding depth 38 mm
Magnet gap depth 10 mm

(38-10)/2=14mm one way!

There is also a version with 11mm xmax.

According to this method the Seas L26 has 7mm xmax.

http://www.seas.no/index.php?option=com_content&task=view&id=116&Itemid=138

My baffle is max 52cm wide and 40 deep..

Again, I am not very familiar with the dipoldesigns mentioned above. But a lack of bass underneath 50Hz sounds strange to me. Maybe at stadiumloud levels?

SPL is calculated with a linkwitz SPL sheet for dipole and monopole and this sheet often underestimates the output.

Greetings,

Martijn

[Dennis H]

Oops, my mistake. I was looking at the 830.

[JimS]

Hi Martijn - here's a shot of the u baffles on the arvos

they go plenty deep on music for me - I've got PE RS315HFs on the bottom end which were rated at 14mm xmax (somehow it's 12.3 in the new PE catalog, so either a typo or they've downrated them a bit) I believe Thomas' are TC Sounds which have quite a bit more . . .

for movies I use a sub for the under 50hz stuff

I'll post detailed impressions once I've had a chance to live with them a bit more and listen to some music, but I think I can safely say I'll not be going back to boxes anytime soon :W

[JimS]

BTW - excellent work :T

[Martijn_H]

Thank you! Sounds like the Arvo is a succes!

No more boxes he? I had that feeling when I comperad my testbaffle to my clossed system. Dipole misses the distortion of the enclosure.

First filtertesting for the woofer. Well got it from 33 to 200Hz. The peak at 70Hz is my room. Impedanz might be a problem, which is getting bellow 3ohm at 50Hz. So active filtering is still in the game.

respons12mh300uf10ohmdd1.png Image not available

[cjd]

Impedanz might be a problem, which is getting bellow 3ohm at 50Hz.

What's wrong with <3ohm?! Yeesh, you'd think everyone has sissy amplifiers! :B

I love the baffles, whether they work perfectly or no.

C

[Paul Ebert]

I've got PE RS315HFs on the bottom end which were rated at 14mm xmax (somehow it's 12.3 in the new PE catalog, so either a typo or they've downrated them a bit)

I believe it's the eight ohm version that is rated at 12.3.

Wish I could afford those BMS woofers. They look to be excellent.

[Martijn_H]

pict2325sw6.jpg Image is not available

The new Mundorf AMT tweeter. The Neo3 was also very good, but misses the looks.

pict2234cj5.jpg Image is not available


Sound is very good. I've used a dcx for filtering. Passive high/mid and active low will be next!

[AJINFLA]

Very nice. Keep posting the updates.

cheers,

AJ

[JohnL]

Hello Martin,

I love that speaker design, the woodwork is beautiful. :T I think I'd like to try something similar to that for my next design. Could you explain how you built the curved side walls of the U part of the baffle?

John

[Martijn_H]

Hello John!

The design is a little on the edge for some people though!

The curved U frame is made with 'bend'mdf. It has special rills so it can be bend. Another option is to take 8 pieces of 4mm mdf, that will do also. The complete story is posted here: http://www.zelfbouwaudio.nl/forum/viewtopic.php?t=4025 (you can see that I copied C90 measurements from this board :B ) Sorry for the language, but the pictures make up for it!

I have also an update to report!

I am about to get me a DEQX pre: http://www.deqx.com/PDC26P-Preamp.html

I'll use it with a 4 channel UcD180 HG and a NAD216THX on the woofers

[Martijn_H]

The dipoles are almost finished. I've made some changes to the frame to handle resonances better.

[Jed]

Most impressive. How do you like those Accuton Mids? Is the Heil tweeter easy to work with? I used to have ER4s and they were a pain to voice properly.

[augerpro]

Can that Mundorf tweeter be run dipole like the original AMT?

Good looking baffle ;x(

[Martijn_H]

Augerpro, It is perfect for dipole use!

Jed, I use it with a 128dB xo (digital) passive you could go 4th order at >2kHz, so you don't need a notch on 1.4kHz. You have to correct the raising frequency response in dipole use with a coil.

The mids are nasty to handle. In dipole I needed a wide notch in the midband. With different xo points and maybe a shallow lp filter on the woofer you might compensate the extra midenergie I guess, so a notch might not be needed. Digital filtering is not fully comparable with passive.. Sound is good from 200Hz-2kHz, but only when the balance is 100%.

[Jed]

The mids are nasty to handle. In dipole I needed a wide notch in the midband.

Your comments of that driver remind me of the C79s I recently sold. Luckily your accutons have more sensitivity than the C79s, so using a notch right in the midrange doesn't lower the system response all that much to get a flat response.

[Martijn_H]

The are finished! 8)

Amps are now 2*3way UcD HG Balanced
Filtering: 210 and 1800 Hz 120/168dB fase lineair

Sound is perfect for me! :T

[mikela]

The are finished! 8)

Amps are now 2*3way UcD HG Balanced
Filtering: 210 and 1800 Hz 120/168dB fase lineair

Sound is perfect for me! :T

Martijn_H,

How do you like the Hypex amp? I have 4 UcD400s just waiting for me to put them together.

Mike

[Martijn_H]

I like the Hypex amps a lot! Mostly because they are very efficient. I don't want a pair of heaters in my room.

They are very stable too! At louder levels they can control the woofers very well! (parallel connected ~3.5ohm)

Good luck with yours!