The Reflexos: A simple, elegant OB / VB, passive / active hybrid speaker.

[jonasz]

Re: The Reflexos: A simple, elegant OB / VB, passive / active hybrid speaker.

fjhuerta: I think you eventually should try some H-baffles, dipole bass is almost as addictive as dipole midrange! :D

These Forsman dipoles (the one to the left) use only two 8" Seas per side and puts out more than enough bass for "normal" listening.
I'm sure RS225 or similar drivers would work fine.

[fjhuerta]

Re: The Reflexos: A simple, elegant OB / VB, passive / active hybrid speaker.

Raising the crossover to 4-600hz will ruin much of the dipole magic.

Exactly.

I wasn't sure a pair of RS-150 would handle anything lower than 200 Hz. I was quite wrong. At 120 Hz, the boxed woofers work harder than the dipole units most of the time. I could drop the frequency even lower if I wanted; the issue is, the baffle is the limit here. There's a point where I'd have to add quite a bit of boost to reach a 4th order target, and I don't want to force things that much. I'll live with 120 Hz for a while and then try and go to 100 Hz or lower.

[fjhuerta]

Re: The Reflexos: A simple, elegant OB / VB, passive / active hybrid speaker.

fjhuerta: I think you eventually should try some H-baffles, dipole bass is almost as addictive as dipole midrange! :D

These Forsman dipoles (the one to the left) use only two 8" Seas per side and puts out more than enough bass for "normal" listening.
I'm sure RS225 or similar drivers would work fine.

Hmm. I have 4 RS225 drivers. ;) It's quite possible my next project will be a dipole bass unit.

Honestly - I like these speakers so much I may just disassemble my huge "monkey coffins" (the ones on the photos, with 2 RS270s, 2 W4-1337s and 1 SB-28 neo) and use the RS270s for the bass modules...

[6thplanet]

Re: The Reflexos: A simple, elegant OB / VB, passive / active hybrid speaker.

You don't nessessaraly need active DSP/EQ to have decient H-frame bass...I proved that with my Open Invit8tions build. Try it, you'll like it.

[fjhuerta]

Re: The Reflexos: A simple, elegant OB / VB, passive / active hybrid speaker.

Do you have a thread for your build? I was considering an H-frame or a V-shaped frame, like the new Linkwitz monitor.

[Mayhem13]

Re: The Reflexos: A simple, elegant OB / VB, passive / active hybrid speaker.

Nice Job and writeup!

[fjhuerta]

Re: The Reflexos: A simple, elegant OB / VB, passive / active hybrid speaker.

Update.

I suspect there'll be a few of these coming.

Fine tuning.

The eye and the ear interpret a frequency response reading in vastly different ways. That's what I've been learning lately.

At first, I thought the power response looked fine. I took an average RTA measurement around my listening seat and it seemed OK. Classical and acoustic music still sounded great, but pop and rock were thin and gutless.

It turns out there was still a lot to do.

The crossover between the mids and highs works correctly. As it is a passive circuit, I'm glad I didn't have to do any changes to it. And thanks to the MiniDSP, I could do stuff I could only dream of with passive components - on the fly. Delays, EQ, shelfs... I had a field day with those.

When I took the RTA measurements, I noticed there was an apparent excess of energy centered around 5 KHz, which emphasized sibilants. I also saw a bit of a hump at 800 Hz. Nothing else caught my eye.

That is, until I put the power response and RTA together and saw that.... there was no excess energy at 5 KHz, but rather a big dip at 3 KHz. And the hump at 800 Hz was there, but was really small. Yep. My mistake was in the dipole correction EQ.

It's just like the glass is full / empty thing. There wasn't too much treble, but rather, bass was a bit down. 2~2.5dB down, at most. The woofer level was mostly correct, so when I looked at the "big picture" I never realized 100 Hz - 600 Hz were the issue.

[fjhuerta]

Re: The Reflexos: A simple, elegant OB / VB, passive / active hybrid speaker.

Still more fine tuning...

It's incredible how difficult it really is to build a dipole speaker. I guess it's only when you really build it problems seem apparent.

So far, I had to raise the FR from 2-4KHz by 3 DB, drop 4-5KHz by 2 DB... the Neo3 are really really really difficult to work with. They sound incredible when properly EQ'd, but they do require lots of TLC.

I've revised the crossover point and raised it to 170 HZ, since the little RS150's were having trouble keeping up with the vented box. Now everything breaks up at more or less the same power level.

Another thing. Measured straight at 1.5 M, the speakers are as flat as the Neo3 allow me, with just a hint of a BBC dip.



But at 3 M, I get humps at 500 Hz and 200 Hz or so. I'd assume that's where the dipole panel starts adding non-coherently. It makes the speakers sound a tiny bit deep - nothing bad, and only with certain recordings. And it's somewhat position -dependant, so the sweet spot is the only place where this happens.





I had to EQ a bit the bass response on the bass box. Nothing bad.

One thing's for sure. Having the MiniDSP is a blessing. I'd have had to rebuild the passive crossover countless times to do what the MiniDSP does in 5 seconds. I love it, it's transparent, its... great.

As for the real test - music listening - what can I say. I get lost in the music with these speakers. Once properly positioned and EQ'd, they are completely different than a box speaker, but in new ways. Now their tonal qualities are the equivalent of a box - since they are properly EQd. But they image completely different, outside their panels, inside, front and back. The midrange is, for the lack of a better word, effortless. I still don't know whether I prefer dipolar tweeters, though - they are more open, but they somehow make imaging a bit more diffuse than I like - I wouldn't know how to explain it, but perhaps I should try damping the Neos rear wave, and see if I like it better.

They are also very efficient. Active EQ is a blessing for small amps. They run barely warm to the touch at high SPLs.

Still - I'll tweak these speakers for a long time to come. I'm at that stage where 1 dB makes a lot of difference. I'll probably read more on SL's experiences with dipoles and apply his techniques on my speakers.

[greywarden]

Great writeup. I have to agree - I love my MiniDSP too! My second pair of amps should be in tomorrow and I'll be ordering a second MiniDSP here shortly although mine isn't so transparent, I have it powered by a phone charger, but when the second one comes in, I'll be finding a cleaner power supply for the both of them.

[charlielaub]

Re: The Reflexos: A simple, elegant OB / VB, passive / active hybrid speaker.

I thought of some questions to ask and didn't see these covered in the thread:

How have you been making frequency response measurements? Are you doing something other than RTA measurements? Is your microphone calibrated?

How did you develop the crossover? Did you do any modeling, or it is it done via a trial and error feedback loop using the RTA and then tweaking the MiniDSP settings?

Which MiniDSP are you using and with which plug-in?

-Charlie

[fjhuerta]

Re: The Reflexos: A simple, elegant OB / VB, passive / active hybrid speaker.

Great writeup. I have to agree - I love my MiniDSP too! My second pair of amps should be in tomorrow and I'll be ordering a second MiniDSP here shortly although mine isn't so transparent, I have it powered by a phone charger, but when the second one comes in, I'll be finding a cleaner power supply for the both of them.

I powered it with a nice big generic Radio Shack wall wart, and it's dead silent (it's a balanced model).

[fjhuerta]

Re: The Reflexos: A simple, elegant OB / VB, passive / active hybrid speaker.

I thought of some questions to ask and didn't see these covered in the thread:

How have you been making frequency response measurements?

When I do a > 1.5M measurement, I use SynRTA @ 1/48 at my listening position, tweeter height, and smooth measurements a bit with LSPCad.

When I do a 1 ~ 1.5M measurement, I use JustMLS, tweeter height. Most of the time, I take the speakers out and use a 50ms gate, and smooth them a bit, too.

Are you doing something other than RTA measurements? Is your microphone calibrated?

Yeah, RTA & MLS. Yes, my mics were calibrated at Cross Spectrum Labs. I sometimes use an ECM-8000, other times the PE equivalent.

How did you develop the crossover? Did you do any modeling, or it is it done via a trial and error feedback loop using the RTA and then tweaking the MiniDSP settings?

I started with your tools, but I couldn't make the advanced plug-in behave as expected. I followed the process, but ... I could never find what was wrong. So, what I did was use the Edge to calculate the dipole response, create the EQ curve using standard L-R filters on the MiniDSP, low and high shelfs, and start from there. It was pretty much an iterative process - measure, modify curves, measure, compare to standard L-R filters, correct, measure, compare, etc. For time delay, I once again used JustMLS in 2 channel mode and measured the time of flight, then calculated the delay - measured using JustMLS, and then inverted polarity to see if I was on the right track.

Which MiniDSP are you using and with which plug-in?

I started with the Advanced. When I couldn't use the tools, I had to switch to the standard and begin trying different L-R filters / shelf filters.
[/quote]


Hello Charlie, I added the answers inline to your question. Thanks for your post!

Javier