AudioCircle

Industry Circles => GR Research => Topic started by: Folsom on 4 Jul 2018, 12:08 am

Title: Forum developed GR-R WAW speaker
Post by: Folsom on 4 Jul 2018, 12:08 am
Hey everyone.

I've been wanting to get started on something like this, but I don't have the wood shops available that I did to me any longer. I thought I'd pass on the concept and see if maybe some people on the forum would be into it. I know Danny is rather busy, too, so it may be up to us for a moment to expand on some possible projects.

What I've been wanting to do is make a capable speaker that is a WAW, woofer assisted wideband. Maybe that's the right acronym, I'm not even sure! The point is to use a fullrange driver with a bass driver.

What I'm proposing to do differently is use Bipole drivers for bass because of how they act. Even small drivers in bipole can produce powerful bass, surprising bass. I've heard 4" drivers absolutely pound before in Bipole. What is unique is that is doesn't have baffle step loss because both drivers are opposite firing in phase. What also is unique is the vibrations for the cabinet cancel really well - spikes and such don't do a lot for them so you can just use some felt pads or extend the base in some fun way with woof. The bass spreads well in a room.

The other component is I want to really make it sound surprisingly natural. So The fullrange will use an aperiodic vent. The bass drivers will sound very natural also with aperiodic vent, but they could also be converted to a transmission line if someone is skilled enough. A TL line will put out very impressive low bass.



Now onto what it looks like. I drew a simple box to show you. I would line the midrange with no-rez and stuff it decently. The stuffing is good because in this case the driver can use extra dampening. The ape vent should help keep the resonance of the internal chamber down to keep the sound natural. The bass with the ape vent will be very natural sounding.


(https://www.audiocircle.com/image.php?id=182122)


The reason the bass drivers (M165) are side firing is that the LGK needs as narrow of a baffle as it can get. At this point the narrowest is approximately 6". All edges should be rounded over, and felt would be pretty trick in a patch around the LGK. The LGK could be offset to one side some.

The crossover isn't completely decided yet. But it's looking at Danny's baffle compensation should work (I took a guess at the values), along with a 60mf Solen bypassed by a 6uf Gen1 Sonicap in series with the LGK & BSC - and my own little change following that. The M165's will need a 2nd order around 300hz. Keeping the LGK slope a little more gentle is intentional because the M165's should be setup to be on a 3 way switch to adjust for bass. As is at 89.9db they will be too sensitive for the nominal w/ BSC 85db LGK, so the M165s will have to be attenuated down to 85db, but in small rooms the gain may still be too much. For that reason a switch could be used to drop it down -3 and -6db. The details aren't fully settled on that, yet. I'm not certain if a 300hz first order will roll the LGK off better than in simulation yet or not, for example.

The FR response I'm seeing has a little bit of a dip at 4khz, but is actually surprisingly smooth otherwise. It looks like the baffle size at 6" wide may smooth the higher end of the LGK. Granted it may be a minor blurring, but that may not be a problem at all for the type of listener who wants smooth, natural sound, with expensive sounding bass.

Right now It's looking like the box would do well at 42" H x 6" W x 12" D w/ a 6" x 6" chamber for LGK on top. This gives around 1 cubic foot for the woofers. With them adjusted to 85db and a touch of room gain, you should be getting appreciable bass into the 40's, if not some into the 30's albeit quieter than a real sub.  The LGK would end up about ear level for recliners and such. The ape vents need 4.3" cut outs, so this is right at the max with .75" depth walls.

I thought about placing the M165's towards the bottom but I'm worried too much floor loading won't sound as natural.

Anyways, this project will get my support to work on the electrical side. I suspect that while Danny doesn't have time to build boxes, were someone near him to bring one to measure he might have some say so in crossovers (probably will have some say so period).

Whelp, I guess now we see where this goes, if anywhere...
Title: Re: Forum developed GR-R FAST speaker
Post by: gab on 4 Jul 2018, 02:44 pm
try the free book  :D

http://grailloudspeakers.com/

http://www.enjoythemusic.com/magazine/manufacture/0418/Grail_Loudspeakers.htm

https://forum.audiogon.com/discussions/grail-t5-speakers
Title: Re: Forum developed GR-R FAST speaker
Post by: Oscillate on 4 Jul 2018, 03:05 pm
Two things that Danny did with his LGKs might be useful here...

 1) Placing a top firing fullrange driver very close to the front firing fullrange to add ambience and output.
The response was particularly sensitive to placement (proximity) of the top foring full range driver.

2) Using a like just 1 resistor or 1 capacitor to shunt only the higher range frequencies to the second full
range driver, while the (first) full range driver played the full range.
Title: Re: Forum developed GR-R FAST speaker
Post by: Peter J on 4 Jul 2018, 04:16 pm
I'd be up for building a prototype or two, and perhaps drawing in SketchUp. Two constraints would be 1. I'm in Idaho which is a little ways from Texas  2. I may not be able to get to it straight away.

It might make a good teething project for big CNC router heading my way soon.
Title: Re: Forum developed GR-R FAST speaker
Post by: mlundy57 on 4 Jul 2018, 04:36 pm
It might make a good teething project for big CNC router heading my way soon.

 :green: That is also green with envy  :lol:

Mike
Title: Re: Forum developed GR-R FAST speaker
Post by: Danny Richie on 4 Jul 2018, 04:53 pm
If you guys power the lower woofers with a separate plate amp, with a variable phase control, crossover, and gain, then it makes the whole thing really easy to pull off.

Also, adding another full range driver to the rear covering only high frequency ranges really makes an improvement in imaging and sound stage layering.

Basically a Skinny 6 design but with powered lower drivers....

http://gr-research.com/skinny6s.aspx
Title: Re: Forum developed GR-R FAST speaker
Post by: Folsom on 4 Jul 2018, 05:26 pm
The Grail probably has a dead spot in frequencies. At the crossover frequencies the woofers will be nearly beaming straight out. BUT it looks like it uses bipole so that's probably the magic people are finding, or at least in the bass and a different sound in the higher registers.

Peter I'm in Spokane so maybe I could come visit... We could do a bit of measuring.

Using a plate amp means it has to be mounted on the side. Not the most attractive, and I think the naturalness of the sound is better suited to people's regular amps... (largely because they'll be playing much higher than subs) But I don't see any reason why an optional GR-T3 rear mounted below the Ape vent couldn't be a decent option. Yet of coarse this can go any direction. I'm just going to argue for a nice speaker that is accessible for people that don't have impressive custom dedicated rooms and such, something that can be a bit more casual while still impressing anyone. 

Title: Re: Forum developed GR-R FAST speaker
Post by: Danny Richie on 4 Jul 2018, 05:55 pm
I tried a side firing M-130 with the LGK driver. My issue with it was the phase relationship. It worked out really well when the side firing woofer was pushed forward of the LGK driver instead of behind it slightly with the centered side mounting. But I couldn't figure out a way to make it look right with the side firing woofer centered nearly 8" ahead of the LGK driver.

The other thing I realized was that the crossover parts for a really low crossover point made the values really big and in the end cost more than crossing the M-130 to a tweeter at 2kHz or so. And I felt like crossing the M-130 to a tweeter actually sounded better and had higher sensitivity.

The advantage of the Skinny 6 came from the thin baffle and next to nothing surface reflections. And with the rear mounted driver it really excelled in imaging and sound stage layering. And they sound big. They sound so big and layered that it is hard to believe they are so small. Also the more limited off axis response of the 3' driver make them more ideal for rooms with limited to no room treatment. I'm loving them in my small home theater system.
Title: Re: Forum developed GR-R FAST speaker
Post by: Folsom on 4 Jul 2018, 06:58 pm
Danny, what did you try to cross the LGK and M130 at?

There will be some big caps. I'll be spec'ing Solen caps bypassed by Sonicaps.

Some of the goals for this, that I've got in mind, are a bit different than the spatial ones.
Title: Re: Forum developed GR-R FAST speaker
Post by: Danny Richie on 4 Jul 2018, 11:03 pm
Danny, what did you try to cross the LGK and M130 at?

There will be some big caps. I'll be spec'ing Solen caps bypassed by Sonicaps.

Some of the goals for this, that I've got in mind, are a bit different than the spatial ones.

I tried crossing them close to 200Hz.
Title: Re: Forum developed GR-R FAST speaker
Post by: Folsom on 4 Jul 2018, 11:59 pm
Ah, that would be harder to get phase the lower you go if I am not mistaken. I suspect there is a possibility to raise the crossover point if phase remains a harder target.

I think it may be ok to settle on phase that maybe could be slightly better but sounds right. I tune subs by ear for phase.

If the phase remains tought I also suspect that moving the woofers down and reversing polairty might be an option. No doubt there may be some expermintation. It might work best to start with removable side panels.
Title: Re: Forum developed GR-R FAST speaker
Post by: jparkhur on 5 Jul 2018, 12:22 am
Ill continue the theme.  Why don’t you put the 165 on front baffle and at a 45 like the v2 subs.
Title: Re: Forum developed GR-R FAST speaker
Post by: Folsom on 5 Jul 2018, 12:32 am
There's two 165's per box. They need to be polar opposite. So I cannot follow how that would work.
Title: Re: Forum developed GR-R FAST speaker
Post by: Jonathon Janusz on 5 Jul 2018, 01:56 am
I think(?) JP is suggesting to open baffle the pair of M165 to get a dipole radiation pattern if that is what you're after, in this case building the section with the M165 like the bigger V2 that has the 12" Eminence drivers.

I know it is going way of script, but would a pair (triple? quad?) of M165 in an angled open baffle to keep it narrow, with a LGK up top, forward firing and itself in an open baffle work?  Would this basically be a tiny V2?  [insert Dr. Evil voice] I will call it... Mini-V...  :lol:
Title: Re: Forum developed GR-R FAST speaker
Post by: Danny Richie on 5 Jul 2018, 02:10 am
Actually keeping the phase right is easier the lower they cross.
Title: Re: Forum developed GR-R FAST speaker
Post by: Folsom on 5 Jul 2018, 04:22 am
I think(?) JP is suggesting to open baffle the pair of M165 to get a dipole radiation pattern if that is what you're after, in this case building the section with the M165 like the bigger V2 that has the 12" Eminence drivers.

I know it is going way of script, but would a pair (triple? quad?) of M165 in an angled open baffle to keep it narrow, with a LGK up top, forward firing and itself in an open baffle work?  Would this basically be a tiny V2?  [insert Dr. Evil voice] I will call it... Mini-V...  :lol:

None of that is applicable, or desirable.

Dipole is not the intention. Bipole is the intention because of what it does for bass and box vibrations. Open baffle would have no bass, and require a lot more space to sound ok. And I don't think open baffle is as good as ape enclosures - at least inside a house.
Title: Re: Forum developed GR-R FAST speaker
Post by: Folsom on 5 Jul 2018, 04:24 am
Actually keeping the phase right is easier the lower they cross.

I dunno, you were moving a driver 8" on top of some other distance with the m130. If the crossover was higher it would take less movement.... so I guess it depends on definition of easy?
Title: Re: Forum developed GR-R FAST speaker
Post by: Danny Richie on 5 Jul 2018, 02:12 pm
The lower wavelengths are longer. So the same amount of distance offset is less of a phase rotation than if the crossover point is higher.

And while bipole loading does cancel moving force, the box vibrations are still the same. Pressure changes in the box are still the same and side walls are excited exactly the same regardless of which way the woofers are facing.
Title: Re: Forum developed GR-R FAST speaker
Post by: Folsom on 5 Jul 2018, 06:10 pm
I think you should build a box and see how much less vibration there is... I've played with multiple bipoles.
Title: Re: Forum developed GR-R FAST speaker
Post by: Danny Richie on 5 Jul 2018, 06:43 pm
I think you should build a box and see how much less vibration there is... I've played with multiple bipoles.

I've built some on the past. The four sides flex or resonant based on bracing and wall thickness. And they are excited by pressure changes in the box. Whether the woofers are on the same side or apposing doesn't change anything. Positive and negative pressure changes are the same.
Title: Re: Forum developed GR-R FAST speaker
Post by: Folsom on 5 Jul 2018, 06:49 pm
Yes of coarse when the drivers play out or in at the same time you'll get changes in pressure. But the amount of mechanical vibration that is canceled is impressive. Normal speakers built with average boxes you can feel a lot of vibration. The transfer of energy into the box through the mounting is very high. Linkwitz did a page on it somewhere. For those not using hardwoods or really thick plywood it's a great way to keep the cost of the box reasonible while greatly reducing vibrations. Compared to 3/4mdf in a non bipole the reduction is very impressive. (on 3/4mdf in for example a bookshelf you can really feel a lot of vibration)

If you're just saying to use no-rez, I think that's a no brainer.
Title: Re: Forum developed GR-R FAST speaker
Post by: planet10 on 5 Jul 2018, 07:29 pm
What I've been wanting to do is make a capable speaker that is a FAST, Fullrange assisted sub tech. Maybe that's the right acronym, I'm not even sure! The point is to use a fullrange driver with a bass driver.

There is a movement to get people to start using WAW (Woofer Assisted Wideband) instead of FAST which is almost universally disliked.

The concept that you are proposing is very workable and exactly s we have done with Tysen V2.

(http://www.planet10-hifi.com/boxpix/tysenV2-passive.jpg)

This specific implementaion uses the Fostex FF85wk and a pair of Silver Flute W14. The push-push loading of the bass drivers goes a long way to reducing box-load and dramatically reduces the possibity of exciting potential box resonances.

The proximity of the drivers allows one to keep the XO above the 1/4 wavelength of the centre-to-centre distance. We do this with the typically 250 Hz PLLXO (also works with the 4th order XOs in my partners Onkyo HT receiver). A passive XO was also developed, it turned out to be about 450 Hz. From these XO points one may find that typical plate-amps do not have the range to XO high enuff. Use of a larger midTweeter might allow for that to work, the quality of the XO in the plate amp maybe a limiting factor.

dave
Title: Re: Forum developed GR-R FAST speaker
Post by: Folsom on 5 Jul 2018, 10:10 pm
So it appears I have mis-named it. Thanks for chiming in Dave.
Title: Re: Forum developed GR-R FAST speaker
Post by: Danny Richie on 5 Jul 2018, 10:35 pm
Quote
The push-push loading of the bass drivers goes a long way to reducing box-load and dramatically reduces the possibity of exciting potential box resonances.

I am not sure why you guys think that this mounting reduces box load. The load the box sees is exactly the same no matter where the woofers are mounted. Box resonances or the potential for box resonances remain unchanged.

If you want to control box resonances then brace it well and add No Rez. Problem solved.

Quote
A passive XO was also developed, it turned out to be about 450 Hz.

That's well up into the heart of the mid-range. that's not a good place for a phase shift of chopping up the vocal region into dis-similar drivers.

Quote
From these XO points one may find that typical plate-amps do not have the range to XO high enuff. Use of a larger midTweeter might allow for that to work, the quality of the XO in the plate amp maybe a limiting factor.

The plate amps are not a problem if crossing below 200Hz. The size of passive crossover parts for the low crossover point are more of an issue if the order of the crossover gets high. Really large cap values or iron core inductors are much more harmful.
Title: Re: Forum developed GR-R FAST speaker
Post by: Folsom on 5 Jul 2018, 11:01 pm
The biploars I have played with have minimal vibration. My X-LS is the exact recommended box w/ bracing and no-rez still vibrate a lot. In these cases it is just my finger accelerometer judging, but it is very easy to tell. The M165 is more of a beast than people give it credit for, I's say.

I cannot speak for what bipoler setups you have tried Danny. And who said anything about not using bracing?

450hz isn't "ideal" but 2khz~ for tweeters isnt either, we do it anyway.
Title: Re: Forum developed GR-R FAST speaker
Post by: planet10 on 5 Jul 2018, 11:03 pm
I am not sure why you guys think that this mounting reduces box load. The load the box sees is exactly the same no matter where the woofers are mounted. Box resonances or the potential for box resonances remain unchanged.

You make the assumption here that most of the energy that can excite box resonances comes from the changes in air pressure inside the box. It is very easy to show that this is not the case and much more energy comes from the direct connection of the driver basket to the box. Push-push done properly kills some (guess) 90% of the driver vibration thru active cancelation, and even more so at the lower frequencies that are more likely to excite the box in a negative manner..

Quote
If you want to control box resonances then brace it well and add No Rez.


It is still necessary to properly brace the box. And as to NoRez i have not had positive results using it.

Quote
That's well up into the heart of the mid-range. that's not a good place for a phase shift of chopping up the vocal region into dis-similar drivers.

There are compromises involved no matter where you XO. A lower XO is advantageous in that it keeps more of the music in the FR/midTweeter. Cross a bit higher and you can play louder. That is of particular importance with a small 3” FR used as midTweeter.

It is important that the helper driver has very good HF extension, the SF W14 reach up to something like 5k on-axis, the Alpair 12pw and Peerless 830870 we use in some other applications reach 10k, the other midbass we have used only goes to 1.5k but that speaker is also on a wide baffle and we have pushed the XO down to 160-180 Hz.

All of these designs were executed with an eye towards 1st order PLLXO, and that coupled with less than 1/4 wl XO point makes a phase coherent result easy.

Quote
The plate amps are not a problem if crossing below 200Hz. The size of passive crossover parts for the low crossover point are more of an issue if the order of the crossover gets high. Really large cap values or iron core inductors are much more harmful.

I have not yet tried every plate amp on the planet but of the many i have, the amps are of 2nd rate execution and the XOs not of the type we’d be inclined to use and usually only reach up to about 150 Hz.

I am a firm believer that an XO at about 200 Hz or lower is much better done actively, and all of our WAW have been designed with an eye to that. The passive XOs were designed to give a complete set of options for anyone purchasing a paid planset subscription.

dave
Title: Re: Forum developed GR-R FAST speaker
Post by: Danny Richie on 5 Jul 2018, 11:58 pm
Quote
You make the assumption here that most of the energy that can excite box resonances comes from the changes in air pressure inside the box.

No, I am saying that the alignment you are using doesn't change that aspect.

Quote
It is very easy to show that this is not the case and much more energy comes from the direct connection of the driver basket to the box.

Oh, I am well aware of what can be transmitted from the driver frame to the baffle. This is why we use polymer frames on all of our drivers. And the difference between our polymer frames and some metal frames can be fairly significant. 

Quote
Push-push done properly kills some (guess) 90% of the driver vibration thru active cancelation, and even more so at the lower frequencies that are more likely to excite the box in a negative manner..

It can kill some of the overall force placed on the box, but it will have little to no effect on the baffle that each driver is mounted to. The vibration that the driver causes on the baffle is still present. You are just causing an out of phase vibration on the opposite panel to cancel it. That can have a positive effect. But there are easier and more effective ways to handle cabinet wall vibration.

Quote
And as to NoRez i have not had positive results using it.

I have never heard of anyone not having positive results with it. It is known to be quite effective. When did you order some?

Quote
I am a firm believer that an XO at about 200 Hz or lower is much better done actively, and all of our WAW have been designed with an eye to that.

I agree.
Title: Re: Forum developed GR-R FAST speaker
Post by: planet10 on 6 Jul 2018, 01:03 am
It can kill some of the overall force placed on the box, but it will have little to no effect on the baffle that each driver is mounted to. The vibration that the driver causes on the baffle is still present. You are just causing an out of phase vibration on the opposite panel to cancel it. That can have a positive effect. But there are easier and more effective ways to handle cabinet wall vibration.

If you execute this properly the drivers are physically connected other than just thru the baffle. Vibrations are instantly cancelled by one driver working actively against the other. Differences can be quite dramatic. Enuff that i would not bother with a single driver subwoofer any more… and subsequently being able to get away with 15mm plywood for sub boxes.

this proves to be pretty good:

(http://www.planet10-hifi.com/images/push-push-SDX10-inside-view.jpg)

But one could take it to the next level by replacing the mounting screws with threaded rod that couples the basket bezels together.

(http://p10hifi.net/planet10/TLS/drivers/images/pushpushPR2.gif)

http://www.t-linespeakers.org/FALL/push-push.html

The particular subject of this drawing having motor and basket allowing a bolt to be run theu the pole-piece.

dave
Title: Re: Forum developed GR-R FAST speaker
Post by: Danny Richie on 6 Jul 2018, 02:21 am
Yes, there is a lot of merit to that aspect, but an un-braced plywood box? 

This is a little more like what I am used to: https://www.audiocircle.com/index.php?topic=53675

That is a non-resonant box.
Title: Re: Forum developed GR-R FAST speaker
Post by: Folsom on 6 Jul 2018, 02:32 am
With the plan as is now the two M165's should be able to be braced against each other with nothing maybe as little as a washer or such.
Title: Re: Forum developed GR-R FAST speaker
Post by: planet10 on 6 Jul 2018, 04:09 am
Yes, there is a lot of merit to that aspect, but an un-braced plywood box? 

None of my boxes are unbraced. The braces are often more work to do than the rest of the box.

Quote
This is a little more like what I am used to:
That is a non-resonant box.

(http://www.gr-research.com/images/solid1.jpg)

Well braced side, top/bottom, but the back, and the baffle — the weakest panel of the box — have no bracing.

And hard to tell: Is that MDF?

dave
Title: Re: Forum developed GR-R FAST speaker
Post by: Danny Richie on 6 Jul 2018, 04:28 am
The front and back panel get another 3/4" layer making them 1.5" thick solid MDF.

(http://www.gr-research.com/images/solid6.jpg)

Then the back panel gets a through hole cut out for the servo amp. So the front and back panel aren't much of a panel at all. They are just a frame that supports the woofer or the amplifier. So there is nothing there to resonant.

And of coarse the whole thing is filled with sand.

(http://www.gr-research.com/images/solid8.jpg)
Title: Re: Forum developed GR-R FAST speaker
Post by: planet10 on 6 Jul 2018, 05:09 am
solid MDF

Not the best material for speakers, and subwoofers in particular.
 
Quote
And of coarse the whole thing is filled with sand.

I usually avoid brute forcing the enclosure. In particular this is hard on the back.

dave
Title: Re: Forum developed GR-R FAST speaker
Post by: Danny Richie on 6 Jul 2018, 02:52 pm
Actually I prefer MDF over plywoods. Plywood is better for bracing only.
Title: Re: Forum developed GR-R FAST speaker
Post by: gregfisk on 6 Jul 2018, 05:15 pm
It seems to me that the more dense the material the better. mdf certainly fits that bill. I suppose the reason for plywood bracing is it's more stiff which I think would make a better brace.

I do like the idea of connecting the two drivers together, if they are locked together they make a very strong brace across the box. It seems to me they would need to be physically connected though so they can't move at all. Otherwise they could still vibrate some.

Of course if you are building a double box full of sand it's probably a mute point. 
Title: Re: Forum developed GR-R FAST speaker
Post by: planet10 on 6 Jul 2018, 06:19 pm
It seems to me that the more dense the material the better.

That is what lots of people think, but it is untrue. Unless the extra weight brings more stiffness it is not worthwhile.

And in a subwoofer in particular stiffness is king. And a piece of quality 15mm plywood has greater stiffness than 18mm MDF.

MDF’s only reall asset is that it is cheap. Cheap to buy, cheap to finish, cheap to work with.

dave
Title: Re: Forum developed GR-R FAST speaker
Post by: Tyson on 6 Jul 2018, 06:22 pm
I like using material of one density for the main box build (say, MDF), and of another density for bracing (say, oak), and then using something like Norez which has a different density entirely.  These 3 things together get you close to a "constrained layer damping" effect.  The idea is that if you use the same material throughout, then the material itself will have a common resonating point.  By using materials of different densities, the resonances are spaced out and are less likely to overlap.  Another way to do something similar is to use MDF bonded to a higher density wood like Baltic Birch.  Brace it with oak, line it with Norez, I don't think there'd be a less resonant enclosure possible (outside of exotic materials).
Title: Re: Forum developed GR-R FAST speaker
Post by: planet10 on 6 Jul 2018, 06:36 pm
The idea is that if you use the same material throughout, then the material itself will have a common resonating point.

That is easily dealt with in the design process. Potential resonance modes can be shifted around by just changing a dimension (ie placement of a brace). And in a subwoofer, the goal is simple, to raise any potential resonances above the bandwidth of the device so that they can never be excited. That requires stiffness as a foremost property.

If 2 materials have the same stiffness the lighter one will have a higher frequency resonance potential. I can/have built subwoofers with 15mm ply and push-push drivers that do not have any resonance until the amplifier is well into clipping and it is WAY louder than it would ever be used… why go to more effort & expense to build something that may (or may not) work as well?

dave
Title: Re: Forum developed GR-R FAST speaker
Post by: Tyson on 6 Jul 2018, 06:55 pm
That is easily dealt with in the design process. Potential resonance modes can be shifted around by just changing a dimension (ie placement of a brace). And in a subwoofer, the goal is simple, to raise any potential resonances above the bandwidth of the device so that they can never be excited. That requires stiffness as a foremost property.

If 2 materials have the same stiffness the lighter one will have a higher frequency resonance potential. I can/have built subwoofers with 15mm ply and push-push drivers that do not have any resonance until the amplifier is well into clipping and it is WAY louder than it would ever be used… why go to more effort & expense to build something that may (or may not) work as well?

dave

Because this is a DIY area and we're not really concerned about the feasibility of a design from the standpoint of producing multiples - we're free to try out different things that might (will) work better, even if it takes more time or a little more effort. 

I'd also point out that most here know that bracing location will absolutely affect panel resonance.  We're ALREADY going to do things like that.  The discussion (IMO) is really around "what other things can we do that will give us even better results".  IME, using a multi-material approach to cabinet construction can do that.
Title: Re: Forum developed GR-R FAST speaker
Post by: planet10 on 6 Jul 2018, 07:02 pm
I much prefer elegance of design over brute force. And if you can achieve the same result with less effort, why not.

dave
Title: Re: Forum developed GR-R FAST speaker
Post by: Danny Richie on 6 Jul 2018, 07:05 pm
Quote
MDF’s only reall asset is that it is cheap. Cheap to buy, cheap to finish, cheap to work with.

And it's denser, less resonant, and more readily available.

Other than plywood being stiffer, making it better for braces, I so no other advantage. And stiff is easy to achieve with bracing.

Quote
I can/have built subwoofers with 15mm ply and push-push drivers that do not have any resonance until the amplifier is well into clipping and it is WAY louder than it would ever be used… why go to more effort & expense to build something that may (or may not) work as well?

Some of us like to take things to another level. I for one can't go back to using un-controlled drivers for low bass response. It's hard to go back from servo control. And then using servo control in an open baffle is in a different league.
Title: Re: Forum developed GR-R FAST speaker
Post by: planet10 on 6 Jul 2018, 07:09 pm
And it's denser, less resonant, and more readily available.

Dense is not an asset, less resonant is questionable, and the last is unfortunatly true for many (not here thou).

Quote
Some of us like to take things to another level.

But if you get the same result with less effort, over the top is just stoking the ego.

Servo is certainly nice, but all the same issues are involved in the box design.

dave
Title: Re: Forum developed GR-R FAST speaker
Post by: Tyson on 6 Jul 2018, 07:12 pm
I much prefer elegance of design over brute force. And if you can achieve the same result with less effort, why not.

dave

I agree with your first statement.  It's the second sentence I'd dispute. 

You act like you're the only one that has ever thought of a well braced cabinet.  You're not.  IMO, a well braced cabinet is the starting point for a design, not the ending point. 

I also tend to dislike push-pull subs from a sound quality standpoint.  Which is hard to measure but is easy to hear.  Just like going to the time/effort of building a box with differing mechanical resonance points won't measure much better but it will sound better. 

That's the whole point of DIY - to experiment.  Try something out.  Does it sound better?  Great!!  Does it sound the same (or worse)?  Well only a bit of time and some materials costs are lost. 
Title: Re: Forum developed GR-R FAST speaker
Post by: Folsom on 6 Jul 2018, 07:55 pm
I suspect this project will be successful with any cabinet material. It's small, easy to brace, easy to line and stuff, and has cancellation properties.

But I will say I prefer nice plywood. MDF is easy but gross.
Title: Re: Forum developed GR-R FAST speaker
Post by: Folsom on 6 Jul 2018, 07:56 pm
I also tend to dislike push-pull subs from a sound quality standpoint. 

Ah, Good, because they aren't part of this project.
Title: Re: Forum developed GR-R FAST speaker
Post by: planet10 on 6 Jul 2018, 08:12 pm
I also tend to dislike push-pull subs from a sound quality standpoint.

Push-Push is not the same as push-pull. Push-pull cancels 2nd harmonics, probably enhances 3rd, and as has been pretty well established, the ear/brain prefers 2nd harmonics to be higher than 3rd, and everything that follows (hard to generate too many higher orders).

I am a big fan of push-push, it solves a lot of problems, push-pull not something i would build.

dave
Title: Re: Forum developed GR-R FAST speaker
Post by: Danny Richie on 6 Jul 2018, 08:14 pm
Quote
Dense is not an asset, less resonant is questionable, and the last is unfortunatly true for many (not here thou).

I lively more resonant box is typically not desired.

Quote
But if you get the same result with less effort, over the top is just stoking the ego.

Getting the same result is questionable. Sound quality in subs really revolves around a solid and non-resonant box regardless of configuration. And opposing drivers can have some advantages, but it doesn't solve all box design issues or exclude any considerations. A solid non-resonant box is still necessary.

Quote
Servo is certainly nice, but all the same issues are involved in the box design.

Yes, just as I said above. The same issues exist with all configurations. But servo control is beyond nice. It is the pinnacle of low bass reproduction.   
Title: Re: Forum developed GR-R FAST speaker
Post by: planet10 on 6 Jul 2018, 08:31 pm
I lively more resonant box is typically not desired.

True. But i am not building resonant boxes. Any potential resonances are pushed up high enuff that they never get excited, so the box does not resonate. The other option is to try to push potential resonances below the DUT’s bandwidth, but that isn’t going to happen in a subwoofer.

dave
Title: Re: Forum developed GR-R FAST speaker
Post by: Folsom on 6 Jul 2018, 08:49 pm
ANYWAYS how about back to the speaker.

Dave, thoughts on crossover, with these drivers?

The LGK has a good Xmax, so it isn't really necessary to go over 200hz, but I suspect it'll be easier at 300-400hz. Is a low order on the LGK not advisable, with a steeper one on the woofers? Again I want to have a 3 way switch for the woofers for 0 -3 -6db, to adjust for room size.

The LGK will be bypassed by a 5uf 9ohm RC regardless of everything but my sims show that does nothing more than lower the highest registered slightly, which may make it a little more on-axis friendly anyways.
Title: Re: Forum developed GR-R FAST speaker
Post by: Tyson on 6 Jul 2018, 08:54 pm
Push-Push is not the same as push-pull. Push-pull cancels 2nd harmonics, probably enhances 3rd, and as has been pretty well established, the ear/brain prefers 2nd harmonics to be higher than 3rd, and everything that follows (hard to generate too many higher orders).

I am a big fan of push-push, it solves a lot of problems, push-pull not something i would build.

dave

You are right, and I mis-typed.  I did mean push-push, not push-pull.  Push-push configurations just sound odd to me (although not as bad as push-pull configs).

Anyway, Folsom is right, we've gone way off topic from his thread.  Sorry man!
Title: Re: Forum developed GR-R FAST speaker
Post by: Folsom on 6 Jul 2018, 09:11 pm
I think sometimes with push-push speakers people forget to do some room correction. That will definitely give it a lot of bass. When it comes to fullrange it's a different kind of sound. I don't think it's bad at all. In some ways it is nice because there isn't much of a sweet spot and toe-in and speaker placement doesn't change much of the sound so long as it isn't against the wall. But it is different. I don't think it's quiet as intimate with fullrange push-push.
Title: Re: Forum developed GR-R WAW speaker
Post by: Danny Richie on 6 Jul 2018, 10:28 pm
Folsom, you can use a first order high pass filter on the LGK with no issues.

And if you are wanting to make the low end variable keep in mind to never use a resistor inline with a low frequency driver.
Title: Re: Forum developed GR-R WAW speaker
Post by: Folsom on 6 Jul 2018, 11:12 pm
I certainly wouldn't use a single series resistor that couldn't handle the load.

It does look like a high pass filter along with your baffle step filter (guessed values from pics) does create a little bit of a 4khz dip. It's not enough to be bad, and not nearly as bad as the ragged response of other fullrangers either.
Title: Re: Forum developed GR-R WAW speaker
Post by: Danny Richie on 6 Jul 2018, 11:29 pm
I certainly wouldn't use a single series resistor that couldn't handle the load.

Even if it is a bundle that can handle the load it's still never a good idea to do that. You'll be changing the damping and control of the driver. it will also be converting output to heat instead. There are better ways to vary the output. 

Quote
It does look like a high pass filter along with your baffle step filter (guessed values from pics) does create a little bit of a 4khz dip. It's not enough to be bad, and not nearly as bad as the ragged response of other fullrangers either.

Where are you seeing that?
Title: Re: Forum developed GR-R FAST speaker
Post by: planet10 on 6 Jul 2018, 11:36 pm
thoughts on crossover, with these drivers?

I have seen the midBass, but am not at all familiar with the FR.

At the frequencies involved going active makes things way easier… and depending on how you do it, could even be cheaper. The size of the passive parts is significant, and good ones are not cheap — ie in our A12pw/A7.3 MTM XO @ about 250 Hz, there are a couple 240 uF caps and some equally big chokes. Even the 450 Hz passive on Tysen V2 has a 240 uF cap.

We design to be able to use 1st order PLLXOs which have minimal parts cost and no electronic haze. Some HT receivers have built in XOs that use channels 6&7 to be able to bi-amp. Something like miniDSP has huge flexibility, but the analog sections in the cheapest ones may not be up to your standards and the best ones avoid ADC at the front end (but sort of restrict you to digital sources). Often people will use something like miniDSP to work out whta filter is best and then try to implement that pssively or in an anaog line-level XO (ie the modified FirstWatt B5 Nelson is working up for the Linkwitz LX-mini).

A PLLXO can be 2 Cs and an R, and if you want to minimize one of those Cs can be the input cap on the HF amp. A series R & shunt cap are the LP filter (one assummes when Danny says no series R in the LP he is talking about a passive XO). At least 1 amp (usually the LF amp) needs a gain control to set levels.

A passive XO is much harder, as one has to deal with the variable impedances of the drivers particularility down around their resonance frequency where the XO is being placed.

When choosing the XO frequency one has to take into account driver centre-to-centre and baffle step F3. Ideally one targets the XO at less than a quarter wavelength of the centre to centre, and within the range BS (F3) times (the range 0.707 to 1). No BS is then needed on the midTweeter.

As the XO approaches 200 Hz the greater is the need to go active.

Do you have a 2nd amp already?

dave
Title: Re: Forum developed GR-R WAW speaker
Post by: Folsom on 6 Jul 2018, 11:39 pm
Even if it is a bundle that can handle the load it's still never a good idea to do that. You'll be changing the damping and control of the driver. it will also be converting output to heat instead. There are better ways to vary the output. 

Where are you seeing that?

Yes, I wouldn't do it to change the response in that way. HOWEVER if the speaker was made into a TL I'd use in series resistor to move the QES up a bit because TL lines don't like well damped woofers. It would be good it move it to .45-.55 range.

I use BoxSIM. I put the LGK in a baffle & box size approximate to what this would be with a smaller front on one side. Then I put in the BSC filter you have (approximate) and a 66uf cap to see how it would do. Phase is pretty good, impedance fine, but response does have bit of a dip at 4khz. I suspect that a 66uf cap inline with a 1mH~ inductor could be at 4khz, but I have not checked. The overall response is pretty good, if not maybe a touch lower than expected in the midrange region. How much attenuation did you shoot for to balance it?
Title: Re: Forum developed GR-R WAW speaker
Post by: Jonathon Janusz on 6 Jul 2018, 11:40 pm
Well, if the initial sketch is the basic design which we are allowed to work with...

I know the aperiodic venting will mess around with the "rules" regarding how big a box should be used and how it will be tuned, but if I'm remembering correctly from the last designs I heard from Danny that use a pair of M165 (or comparable) boxed (the x-static and the x-cs sealed, the x-mtm ported):

Isn't the box a little big to be ideal for a sealed box (where the aperiodic venting would allow one to "cheat" and make the box even smaller, which is more or less backwards from what we see here)?

Isn't the box significantly smaller than ideal if going ported (the x-mtm is a lot bigger than most realize until it is sitting in front of you, and even then I know from way back in the day that folks often struggled to get that speaker to play nice in the low end because of the amount of bass a pair of M165 ported could put out)?

All this is maybe to ask, would a pair of M130 be a better plan for the intended application?  They could be moved around the baffles more because they are smaller around, use a lot smaller optimum box no matter how you cut it (sealed or vented), and considering the LGK isn't super sensitive nor moves a lot of air, what one loses taking away the M165 seems to be suggested will be made up for with the strengths of the intended design (frequency extension through porting and output through using a pair of drivers) with the benefit of getting both speed and the ability to play better (relative quality versus quantity) higher up the frequency range if the goal is to basically use the LGK as a (sort of?) really low crossing tweeter?

To close, might I suggest checking out Santiago's prior threads showing his skinny-six builds for inspiration on how to build out the boxes for this project; his seemed an elegant and clever solution to beefing up the box (in this case for the woofers) without compromising the narrow baffle width for the LGK.
Title: Re: Forum developed GR-R WAW speaker
Post by: planet10 on 6 Jul 2018, 11:47 pm
...and considering the LGK isn't super sensitive

reminds of a point i forgot to make. If one is going for a passive XO, the woofers need to be some 3-5dB greater sensitivity than the midTweeter. This allows for choice of XO such that the extra sensitivity gives the BSC needed. If you go active sensitivities are way less important.

dave
Title: Re: Forum developed GR-R WAW speaker
Post by: Folsom on 6 Jul 2018, 11:50 pm
Dave, active will come second to passive is the plan. Any able to fiddle with a MiniDSP can cross them over themselves with little to no trouble.

240uf would be huge. I'm getting 66uf at 300hz as a workable figure. At 44" we're looking at 11" distance for a quarter. It looks like that might be a bit to close while using separate internal enclosures. Where as at 200hz and 16.5" should be pretty easy. The M165's would just be close to the top. The price of a Solen cap does jump a good amount to get there from 60uf to 100uf. (almost doubles the cap price,  but still not that bad)
Title: Re: Forum developed GR-R WAW speaker
Post by: Folsom on 6 Jul 2018, 11:58 pm
Isn't the box a little big to be ideal for a sealed box (where the aperiodic venting would allow one to "cheat" and make the box even smaller, which is more or less backwards from what we see here)?

All this is maybe to ask, would a pair of M130 be a better plan for the intended application?  They could be moved around the baffles more because they are smaller around, use a lot smaller optimum box no matter how you cut it (sealed or vented), and considering the LGK isn't super sensitive nor moves a lot of air, what one loses taking away the M165 seems to be suggested will be made up for with the strengths of the intended design (frequency extension through porting and output through using a pair of drivers) with the benefit of getting both speed and the ability to play better (relative quality versus quantity) higher up the frequency range if the goal is to basically use the LGK as a (sort of?) really low crossing tweeter?

The vents really don't help you change the speaker box size. What it really does is changes the Q, in terms of measurements. You still need volume for output. While the vent maybe could allow a driver to play in a smaller box than usual, you still throw away bottom end as you shrink. The intention of them in this case is purely for the sound. Ported would work away from the sound intention. A TL technically could be used with an Ape vent and may be very pleasurable with one. I personally prefer a TL over a port, but the obvious work load is extensive to pull it off. In this case I think it'll just be sexy enough to have the Ape vent. Consider that the M165's won't have any baffle step loss, and they'll need to be brought down to the sensitivity of the LGK at 85db, and then will have a switch to lower them even farther because the straight forward no BSL may leave the bass too strong. The way crossovers work this should give very nice bass extension, lower than the regular bookshelves that are ported because you're huffing off sensitivity that will also flatten the response below the crossover.

The M130 cannot play nearly as low without a port. It also has a VAS difference that makes it more demanding on box size for low output.

reminds of a point i forgot to make. If one is going for a passive XO, the woofers need to be some 3-5dB greater sensitivity than the midTweeter. This allows for choice of XO such that the extra sensitivity gives the BSC needed. If you go active sensitivities are way less important.

dave

As Luck would have it the M165 is 89.9, and the LGK is really more like 85db.

You don't thhink the LGK needs any BSC at 200hz+ ?
Title: Re: Forum developed GR-R WAW speaker
Post by: planet10 on 7 Jul 2018, 12:06 am
You don't thhink the LGK needs any BSC at 200hz+ ?

The width of the box is the primary input into that factor. 200 Hz is going to need ~ 20”+ wide cabinet (on 1st pass), but there are other factors. Our XOs are usually lower than theory would suggest.

Also a factor is how low the midTweeter goes. We tend to put them in an aperiodic midTL that aims to eliminate any effect of the output of the rear side of the cone, but it also means the midTweeter does not reacg as low.

As an aside, aa aperiodic  midTweeter TL enclosure does a good job of lowering the magnitude/fattening the resonant peak making a passive HP filter less problematic.

dave
Title: Re: Forum developed GR-R WAW speaker
Post by: Folsom on 7 Jul 2018, 12:08 am
Oh, and Danny, can I confirm if the M165 is 2.93" insertion depth? It isn't listed, but that's what the other M165 variants are I believe.
Title: Re: Forum developed GR-R WAW speaker
Post by: Folsom on 7 Jul 2018, 12:12 am
Dave can you give me an example of such a midtweet TL setup? A pic? I was hoping to keep it simple but it could be advisable to have an ape vent chamber feeding a TL that's at resonance to help it drop off faster... That is the point, right?

The front baffle is not going to be 20" wide, so I don't see how a BSC is going to be unneeded.

The example WAW speaker you showed earlier at 450hz passive xover obviously wasn't within 1/4 wavelength distance for 350hz.
Title: Re: Forum developed GR-R WAW speaker
Post by: Danny Richie on 7 Jul 2018, 12:31 am
Quote
At the frequencies involved going active makes things way easier… and depending on how you do it, could even be cheaper.

Dave is correct about that one. And it gives you all the flexibility you need.

Big wide baffles should be avoided. That really hurts imaging and sound stage depth. The LGK doesn't require a baffle any wider than 4.5".

Quote
Oh, and Danny, can I confirm if the M165 is 2.93" insertion depth?

Yes.

And you will need a filter for baffle step loss.
Title: Re: Forum developed GR-R WAW speaker
Post by: Danny Richie on 7 Jul 2018, 12:37 am
Quote
I use BoxSIM. I put the LGK in a baffle & box size approximate to what this would be with a smaller front on one side. Then I put in the BSC filter you have (approximate) and a 66uf cap to see how it would do. Phase is pretty good, impedance fine, but response does have bit of a dip at 4khz. I suspect that a 66uf cap inline with a 1mH~ inductor could be at 4khz, but I have not checked. The overall response is pretty good, if not maybe a touch lower than expected in the midrange region. How much attenuation did you shoot for to balance it?

I wouldn't put much faith in modeling software. Taking real measurements tells you exactly what's going on.

And the filters needed are easy to design based on the real output. It is much easier and faster to actually put the filters on the speaker and shoot a response than to theorize what is going to happen.
Title: Re: Forum developed GR-R WAW speaker
Post by: planet10 on 7 Jul 2018, 12:38 am
can you give me an example of such a midtweet TL setup?

http://www.planet10-hifi.com/planset/Tysen-V2-extents.gif

It is a simple 10:1 line the depth of the box. As long as it has sufficient volume the other detail are not overly critical.

Quote
That is the point, right?

The point is to completely absorb the rediation off the back of the driver.

Quote
The front baffle is not going to be 20" wide, so I don't see how a BSC is going to be unneeded.

That is a 1st order quess for a 200 Hz XO.

Quote
The example WAW speaker you showed earlier at 450hz passive xover obviously wasn't within 1/4 wavelength distance for 350hz.

That is true, and why the active XO is at 250 Hz. The fellow that did the passive XO took a lot of measures of the drivers in the box and that is where he ended up.

(http://www.planet10-hifi.com/planset/TysenV2-modeledXO-FR.png)

This is the simmed response but after the fact measures showed it to be spot on.

dave
Title: Re: Forum developed GR-R WAW speaker
Post by: Danny Richie on 7 Jul 2018, 12:55 am
Man, those lower woofers are playing up really high. That could cause a lot of out of phase cancellation in the vertical off axis. There might even be some in the on axis.

That looks like about a 5db peak at 9kHz. I can design a filter to knock that out of it if you like. But then again, it is just a simulation so who knows.
Title: Re: Forum developed GR-R WAW speaker
Post by: planet10 on 7 Jul 2018, 01:10 am
Quote
That looks like about a 5db peak at 9kHz.

A bit bigger on the stock drivers. It does not intrude much being that high up. The Alpair 5.2 is probably smooth up there but wasn’t on the radar when we started these. A wide variaety of 3” drivers could be loaded into that TL including the LGK, and not much change needed to accomodate larger drivers in the Alpair6.2/7.3 class.

The woofers go up to 5k or so on-axis. I had misgivings about the high XO when i 1st got the design, but in practise it works well. Edged out by FHXL, but a much smaller footprint. Bipolar “viewed” on edge has perfect bafflestep fill-in, issues near the XO will be dependent on the off-axis.

The fellow who did the XOs is a measurement based fanatic, these were measured in a big space, with them 20+ ft in the air. That the measured responses after XO were very close to the sim is an indication of how well he did.

dave
Title: Re: Forum developed GR-R WAW speaker
Post by: Folsom on 7 Jul 2018, 03:47 am
I guess the next step is waiting for the measurements from raw drivers in box.

The TL isn't what I thought. I would not use it with this driver, I believe the ape vent will be much better.
Title: Re: Forum developed GR-R WAW speaker
Post by: planet10 on 7 Jul 2018, 04:35 am
I would not use it with this driver, I believe the ape vent will be much better.

2 ways of doing much the same thing.

dave
Title: Re: Forum developed GR-R WAW speaker
Post by: Folsom on 7 Jul 2018, 05:19 am
The LGK has almost no dampening, some back pressure that isn't highly resonate will make the amplifier put on the e-brake better.
Title: Re: Forum developed GR-R WAW speaker
Post by: Danny Richie on 7 Jul 2018, 04:12 pm
Quote
The fellow who did the XOs is a measurement based fanatic, these were measured in a big space, with them 20+ ft in the air. That the measured responses after XO were very close to the sim is an indication of how well he did.

I am a bit fanatical about the measured responses myself. But while most just look at the on axis response, I look heavily at the horizontal and vertical off axis response to see how well drivers remain in phase over a wide range or if there are out of phase cancellations taking place that might cause an uneven room response.

That design is one that I would be curious to measure myself as I kind of question what will happen in the off axis in both planes.