diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Multi-Way (http://www.diyaudio.com/forums/multi-way/)
-   -   Starting an Active Open Baffle design. (http://www.diyaudio.com/forums/multi-way/238002-starting-active-open-baffle-design.html)

DrDyna 21st June 2013 03:27 PM

Starting an Active Open Baffle design.
 
1 Attachment(s)
Well, hello folks.

It seems that I have this sickness that makes me want to try various types of designs, getting to know them, getting bored with them, and trying something else for fun.

This time, I'm going to try open baffle. I saw a design by James Trexx called "The Gate" and the impressive looks, coupled with his listening impressions have baited me into trying something like this, WMTMW. I've got spare drivers hanging around and a MiniDSP (soon to be two) so I can pretty much try anything. Open Baffle looks intriguing, mostly because it seems extremely easy to build (I can probably rip out the two main H Frames on a table saw in the same amount of time it takes me to pencil mark where the internal pieces go in a folded horn :P)

So, now we're onto driver selection. I have a supply of Peerless 831727s (4 of em) that have been gathering dust for a while and I'd like to do something with them. So, I whipped up a rather crude H-Frame 11"w 22"h 12"d and mounted them in there push-pull. Measured, they seem to be fairly well behaved from 60 hz to around 600 hz, so I'm guessing they'll work fairly well tasked with 60 - 300hz or so. The design I'm working on wouldn't have a push-pull H frame, but rather a single driver bottom and single driver top.

For a midrange driver, I'm considering something from Dayton RS, perhaps RS100s or RS150s.

For the tweeter, I'm prepared to cannibalize the B&G Neo3 PDRWs that are currently in my speakers.

I've received some input from a couple of members already from another thread that show some concern with the WMTMW layout, I.E vertical issues with the Neo3, and also some issues with it not being able to keep up with flanking H-Frame midwoofers. Perhaps I can overcome the sensitivity issue by simply lowering the gain on the midwoofers and using the extra headroom to eek out some extra low end extension with EQ.

I'm open to any and all suggestions / warnings / kicks in the chuff. Here's a preliminary sketch I did yesterday, borrowed heavily from James Trexx's Gate, and the center baffle borrowed from 2Pi's Aristoteles.

Fire your cannons!

CharlieLaub 21st June 2013 04:01 PM

Quote:

Originally Posted by DrDyna (Post 3534079)
Well, hello folks.

It seems that I have this sickness that makes me want to try various types of designs, getting to know them, getting bored with them, and trying something else for fun.

This time, I'm going to try open baffle. I saw a design by James Trexx called "The Gate" and the impressive looks, coupled with his listening impressions have baited me into trying something like this, WMTMW. I've got spare drivers hanging around and a MiniDSP (soon to be two) so I can pretty much try anything. Open Baffle looks intriguing, mostly because it seems extremely easy to build (I can probably rip out the two main H Frames on a table saw in the same amount of time it takes me to pencil mark where the internal pieces go in a folded horn :P)

So, now we're onto driver selection. I have a supply of Peerless 831727s (4 of em) that have been gathering dust for a while and I'd like to do something with them. So, I whipped up a rather crude H-Frame 11"w 22"h 12"d and mounted them in there push-pull. Measured, they seem to be fairly well behaved from 60 hz to around 600 hz, so I'm guessing they'll work fairly well tasked with 60 - 300hz or so. The design I'm working on wouldn't have a push-pull H frame, but rather a single driver bottom and single driver top.

For a midrange driver, I'm considering something from Dayton RS, perhaps RS100s or RS150s.

For the tweeter, I'm prepared to cannibalize the B&G Neo3 PDRWs that are currently in my speakers.

I've received some input from a couple of members already from another thread that show some concern with the WMTMW layout, I.E vertical issues with the Neo3, and also some issues with it not being able to keep up with flanking H-Frame midwoofers. Perhaps I can overcome the sensitivity issue by simply lowering the gain on the midwoofers and using the extra headroom to eek out some extra low end extension with EQ.

I'm open to any and all suggestions / warnings / kicks in the chuff. Here's a preliminary sketch I did yesterday, borrowed heavily from James Trexx's Gate, and the center baffle borrowed from 2Pi's Aristoteles.

Fire your cannons!

Here are my thoughts, from a recent endeavor involving active OB :

WMTMW layout is a potentially bad idea because of vibrations, especially from the top section, which will not be stiffly coupled to the floor because the middle is open. Move all woofers to the bottom.

Use a 4in-5in diameter true midrange driver, of as high a sensitivity as possible. For instance, in my prototype I used a pair of Seas MP12 drivers that I had around, and this worked pretty well. After extensive searching and modeling it seems the best driver out there for this purpose is the Seas 10F.

Do not use two midranges - this will definitely result in undesirable interference, because you can not cross over to the B&G until 3-4kHz IIRC.

Change to a 4-way system. You will need a driver to fill in below the small diameter midrange, which has to be crossed over somewhere around 600-700Hz, and the woofers, which can only operate up to about 100-150Hz because of the TL resonance of the H-frame.

Suggest U-frame loading for the woofers instead of H-frame. U frame has better sensitivity for given physical "size/length" and a better dispersion pattern.

You will need to EQ the woofers substantially. Sugges that you do NOT do this in the MiniDSP (except during the development phase when you are trying out different EQs) - keep in mind that the max peak output voltage of the MiniDSP is NOT very high (2V pk-pk?). The high amount of gain that is needed to extend the woofer response means that all other levels must be depressed in order to prevent internal digital levels greater than 0dB and very high and obvious distortion, etc. Instead, design and build an analog circuit for this purpose using an LT + a shelving filter.

Consider using a low noise, fixed gain stage with 10dB-20dB of gain on each MiniDSP output to better match its output to amplifier input sensitivity, or use amplifiers that have 0.707Vrms or less input voltage to reach max output power. You will appreciate the extra headroom!

-Charlie

DrDyna 21st June 2013 05:09 PM

Thanks Charlie!

A couple of additional questions, if I may.

You suggest against H frame loading, preferring U frame instead. Are there still issues with the TL resonances using the U frame that would prevent their usage past 150 hz?

As far as the output of the MiniDSP, I'm planning on using my spare HTR as an amplifier, it has multichannel inputs and offers around 100 wpc, along with a fairly substantial amount of gain, as it's really just a 7 channel integrated amplifier when used like that.

Lastly, will I need to EQ the woofers substatially? In the cheezy prototype H-Frame (and I'm guessing the U frame wouldn't be terribly different) they go down to about 90 hz without any EQ at all, and the drop off below that doesn't seem very bad. In either case, I've got a pair of OD-ML-TL subs that I can run all the way up to about 120 hz if I needed to.

CharlieLaub 21st June 2013 06:04 PM

Quote:

Originally Posted by DrDyna (Post 3534186)
Thanks Charlie!

A couple of additional questions, if I may.

You suggest against H frame loading, preferring U frame instead. Are there still issues with the TL resonances using the U frame that would prevent their usage past 150 hz?

Yes, both U- and H-frame have the line resonance. One main reason for using a U- or H- frame woofer is that you get near constant-directivity, but that is true ONLY when used below the first (lowest frequency) resonance of the short "transmission line" that is formed by the geometry. If you make the depth of the frame long enough to get good bass response, this also lowers the first resonant frequency, so there is a tradeoff. Do not expect to use the U- or H- frame up to 500 Hz, for instance. If you need to, then just use an open baffle, which doesn't have these restrictions.

Quote:

Originally Posted by DrDyna (Post 3534186)
As far as the output of the MiniDSP, I'm planning on using my spare HTR as an amplifier, it has multichannel inputs and offers around 100 wpc, along with a fairly substantial amount of gain, as it's really just a 7 channel integrated amplifier when used like that.

Look up the input sensitivity in the manual. The power capability means nothing if you can not use it because the MiniDSP doesn't generate enough output voltage. Don't forget that you have to reduce all the levels in the system so that the highest output ever needed (eg at full "volume") is no higher than about 0dB digitally.

For instance, I used an Adcom GFA-2535 which has 4 channels. I bridged two channels, and spec'd power is 200W into 8 ohms. But the input sensivity for "full power" for the bridged configuration is 1.78V RMS! The MiniDSP CAN NOT reach this output voltage! So I could not "use" all 200W. Worse, the amp has a couple of dB of headroom, but you need to supply an even higher input voltage to push the amp into clipping. In short I could not use the full capabilities of the amp, which was kind of disappointing. Another 10+dB of gain would have been useful...

Quote:

Originally Posted by DrDyna (Post 3534186)
Lastly, will I need to EQ the woofers substatially? In the cheezy prototype H-Frame (and I'm guessing the U frame wouldn't be terribly different) they go down to about 90 hz without any EQ at all, and the drop off below that doesn't seem very bad. In either case, I've got a pair of OD-ML-TL subs that I can run all the way up to about 120 hz if I needed to.

If you use a subwoofer, you won't need to EQ the woofers that much. But then again, because of the line resonance it doesn't make sense to use a U- or H-frame plus a subwoofer. You should change to OB for the woofers instead. If you do use a U- or H-frame, then you should operate the woofers between the line resonance (e.g. about 150 Hz) and the lowest frequency that balances bass extension, SPL capability, power requirement, and driver max excursion limitations.

Here are some links to info on U- and H-frame:
U and H Frames - Quarter Wavelength Loudspeaker Design (Martin King)
DIY-dipole-1 (Music and Design)
H-U frame woofers (Linkwitz Labs)
A comparison of OB, H-, M- and U-frame dipole

In the Linkwitz Labs link, for instance, the line resonance that Siegfried is modeling is at 143Hz (see his plots) and the woofer output needs to be already down at this point, so the crossover point should not be higher than about 100 Hz.

-Charlie

DrDyna 21st June 2013 06:47 PM

Thanks for explaining the line resonances, that makes a lot more sense. Most of those links were already greyed out from my research yesterday and today, but it would appear I should read them more carefully...my mind wanders sometOH LOOK THERES A CHICKEN.

I guess this greatly simplifies my idea, now I'm thinking just one straight open baffle, no H or U frames, crossed over to the sub around 100 hz or so, I'm guessing that'll be where f3 happens on a flat baffle with those woofers.

I'll make some sawdust here shortly and see what it gets me, at least I can get 3 or 4 design iterations out of a single piece of MDF :)

Thanks again!

Rudolf 21st June 2013 08:57 PM

2 Attachment(s)
I simulated two Visaton W250 drivers in a H-frame (red) and in OB (blue) to give a rough comparison without any EQ applied:
Attachment 355549
The H frame is 9 cm deep on both sides. If you don't intend to use the dipole woofers below 60 Hz, a shallow H-frame will not have much advantage over a simple OB of same cross area. But a shallow H-frame will be usable up to at least 300 Hz too. Note that the frame resonance is only about 4 dB.

Doubling the depth of the H-frame would give another boost below 40 Hz, but reduce the upper range to about 200 Hz:
Attachment 355551

Rudolf

DrDyna 21st June 2013 09:07 PM

Cool, thanks Rudolf. I started cutting a test baffle just to measure the 831757 to see how usable it is and if I can get my subs crossed over high enough to meet them. Right now my baffle looks a lot like Gainphile's s19. I have to fiddle with the driver cutout a bit, so it might be tomorrow before I get the drivers mounted up. If its a complete disaster, I won't bother measuring it, but I'm anticipating that with a little eq, I should be able to get some halfway decent results.

DrDyna 21st June 2013 10:37 PM

1 Attachment(s)
Here's what the 831757 does in the quickie baffle I made, about what I expected I suppose.

What do you guys think, worth EQing and messing with?

Rudolf 22nd June 2013 09:42 AM

Quote:

Originally Posted by DrDyna (Post 3532997)
As far as sonic advantages, I'm really just borrowing from two designs that I've seen, James Trexx's and the aristoteles. Its a grand expirement. What I'm looking for in having the upper mid bass h frame is really just "sonic size" and the visual impact.

Quote:

Originally Posted by DrDyna (Post 3533079)
By sonic size, really what I mean is just lighting the whole wall up with sound, not necessarily shaking the house. I've always had trouble with sonic size, where it always really feels like you're "listening down", so I was going for symmetry and lots of baffle area and having that top H frame being pretty much against the ceiling.

I transfered the above posts from the other thread because I'd like to discuss the "sonic size" and the "wall of sound" in more depth.

There is nothing in the literature indicating that you can raise the vertical impression of < 200 Hz bass by lifting a bass source. Bass below 200 Hz will be dominated by room modes. And the excitation of those modes doesn't rely very much on the floor or ceiling position of the drivers.
Nevertheless I don't doubt that the visual impact of the upper woofer can feed the illusion of raised bass. Ventriloquism is a strong persuader ;).

Raising the midrange is a different story. Having a second midrange driver above the tweeter can help to lift the acoustic scene.

How about "lighting the whole wall up with sound"? Dipoles allow for a really impressive horizontal sound panorama. It depends on the interaction of the radiation pattern and the room and is not a "feature" of the speaker alone. Lighting the wall from bottom to top is a matter of increasing ASW. You can have the proverbial "wall of sound", but only at the expense of widened individual phantom sources. Did I mention that increasing ASW is very much a function of your room treatment (or lack of it) too?

Rudolf

puppet 22nd June 2013 03:23 PM

1 Attachment(s)
My OB attempts have evolved into something like this recently. It's active 4-way and sonically is the best thing I've tried so far. An H or U frame can help boost the lower end but the resonance(s) aren't for me ... after trying it without. I have been using 3rd order bessel elect filters with spread frequencies to knock down the peaks at the various xo points. Simple enough to do.

The 15" driver has a 6db shelf w/5db boost starting @ 320hz .... his HP is 144hz or so. The 12" lower mid is rolled off @ 80hz .... varying this can allow a petty large contribution to the low end or not depending on the room or one's listening tastes.

Sorry that I don't have proper data to offer, however, I think it's worth a look see Doc.


All times are GMT. The time now is 11:48 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2