My open baffle dipole with Beyma TPL-150

[StigErik]

Me and a friend started out with two large pine boards, and three hours later we had two working dipole speaker prototypes. The fastest speakers I've ever build!

We wanted to have a speaker that is true dipole, without using two tweeters like the Linkwitz Orion. The Beyma TPL-150 with the rear chamber and damper pad removed proved to be perfect! It extends more or less flat to 1.5 kHz. Removing the damper pad (which is glued to the rear side of the driver) produced a 5-6 dB peak around 9 kHz, but that was easy to get rid of with parametric EQ.

The midrange is the same Seas W22EX001 used in the Linkwitz Orion.

For the bottom octaves we are going to use a new 21" from Beyma, which has not yet arrived. This woofer named 21SW1600nd was presented at Prolight and Sound in Frankfurt a month or so ago, and production drivers are expected around the end of May. While we wait, a Beyma 18G50 acts as stand-in. The longthrow 21" will produce at least 6 dB more SPL than the 18G50. Not that the 18" is anyway close to bad... there is plenty of bass down to approx 30 Hz.

XO and EQ is run digitally in my computer. There's a long thread on this forum about that: "A how to for a PC XO". XO frequencies are 150 and 1500 Hz, 4th order slope.

[StigErik]

And here's an other picture...

[gainphile]

Seas W22 ... so much have been said about this driver being so great. ..... but they're $450 each in Australia. The better half will kill me if I spend $1000 on a pair of speakers and not even the box he..he..

How is it in Norway? Do you just show up at their factory?

[StigErik]

Unfortunately, Seas dont have a factory outlet.... well, they do... but you will have to buy at least one pallet of drivers. I dont need fifty W22 woofers right now...

The retail price here in Norway is a bit higher (would you believe that...) than in the US - NOK 1961, which is approx USD 300. At Madisound they are USD 230 or so.

The W22 is however the cheapest driver in this speaker. The TPL150 costs twice the W22, and the Beyma 21" is expected to be around 700 USD at least. Yes - its a lot of money in drivers, but there's no large cabinets to waste your money on...

[studiotech]

Looking good! I am considering something very similar to this myself. I was thinking of using the Raal 140, so no dipole highs, but after such great success with the TPL here Beyma tpl-150 + Lambda TD12M's I am considering making the Beyma open backed also. Since you have experience with both styles now, what can you say about the perceived differences? Also, do you have any direct experience with the Raals?

We have a dealership for Beyma, so would it be cost effective for you to purchase drivers from us and post them overseas to you? Just a thought. Will help you out if we can.

Greg

PS. Am I correct in thinking that you worked on the Griffin studio monitors?

[StigErik]

Yes - I designed the original Griffin Model 1, Model 2 and Model 2A.

After this little experiment, I would strongly recommend running the tweeter dipole as well, so you should look for somthing that can be modified for true dipole operation. Most AMT's will work, the B&G Neo3 and Neo8 (although the latter is not a tweeter) will also work, and they sound great too.

I think the TPL150 is absolutely wonderful when modified for dipole. I've tried it with the rear chamber also of course, but I feel that it sounds more open, stress-free and naturally detailed when the rear chamber comes off.

Have no experience with the RAAL ribbons.

I have a Beyma dealership as well, thats how I can get hold of not yet released drivers like the 21SW1600nd.

This experiment proved to be very successful, so we will of course develop it further! The current baffle is not very stiff, dead or heavy at all, being made of 18 mm pine board... Current ideas are 60 mm MDF for the lower frame (around the woofer), and maybe 30 mm MDF for the upper parts. We are also considering a very heavy base, possibly using a thick steel plate. I think we will keep the "wings" around the edge to further increase stiffness.

[Brett]

Is this intended to replace the Gjallarhorn?

[StigErik]

This is not a replacement for my big Gjallarhorn 418. I'm builing the prototype with a good friend of mine - it will be his main speakers. I will build a pair for myself also, just because its fun to have a pair.....

Today I have done some more measurements. The tweeter behaves very well, the only problem is a quite large peak at 1400 Hz, which correlates well to the baffle width at the tweeter height. Thats easy to control with parametric EQ, so I dont worry.

Whats worse, is a series of nulls and peaks in the midrange. There are nulls/dips at 300 and 900 Hz, and a rather large peak at 600. This also correlates well with the rear to front distance around the baffle. There is a bit too much of these peaks and dips, I would like to avoid that. Is the only solution a wider baffle? I did not see this that obvious when I measured the W22 before we cut of the edges to get the pyramid shaped baffle (we started with a rectangular board).

Any ideas regarding the midrange problems?

The bass works fine, much better than expected. There is however no point in adding EQ lower than 30 Hz, as we run out of cone excursion very very fast.

[StigErik]

Found the source for the 300 and 900 Hz dips...: the floor bounce...

The large 600 Hz peak remains, but thats easy to control with EQ. Actually its 500 Hz, and the cure is a 9 dB dip there. Yes 9 dB. The speaker is unlistenable without EQ here.

And now - with a pair of absorbers on the floor, it measures a lot better, and its starting to sound really good!

[pixelpusher]

StigErik,

Could you post your measurements of the TPL-150 with open back? It might work in my dipole design. Currently I use the BG Neo3pdr open backed but it is rather low in sensitivity and the PHL 1120 mid I'm using needs to be padded down quite a bit. I'd thought of using two Neo3's stacked in series but maybe the TPL-150 would be better. Do you think the TPL-150 sounds better than BG Neo3pdr?
Thanks