Jump to content


Photo

(4) Sealed 21": Funk Audio UH-21v1


  • Please log in to reply
122 replies to this topic

#1 SME

SME

    Super Bass Overlord

  • Members
  • PipPipPipPipPipPipPipPip
  • 965 posts

Posted 24 August 2016 - 09:32 PM

I got a chance to place an early order for four of the upcoming Funk Audio UH-21v1 drivers.  This driver is an underhung design, similar to the Focusworks GUJ18V1 and GUJ21V1 drivers, but it uses the magnet structure from the TSAD21v2 driver featured in the Funk Audio FW21.0 powered sub.  I requested coils with lower resistance for a better match to my choice of amp and for getting the most motor strength possible.

 

Here are some pictures of the finished product:

 

hC1UUkvCpK8Ha4MKza72HNg0DcZJTA0Xu94UczOp

 

hRHIvS-rVoZvToSi81M-7FN3luMRJRIqHo7CN6NM

UH-21v1 driver

 

Forgive me for my pictures are not the best.  Here is a driver undergoing inspection:

 

UH-21v1-with-cat.jpg

 

My cat likes bass and sine sweeps but doesn't like dynamic movie sounds or music with horns.  Here are a couple more:

 

UH-21v1-top.jpg

 

The carbon fibers have a mirror like reflection when viewed from the right angle.  None of the pictures I've seen do the motor/magnet assembly justice.  It's about 10" diameter like the spider:

 

UH-21v1-underside.jpg

 

And of course, they deserve nothing less than an SP2-12k drive them all:

 

sp2-12k-front.jpg

 

These are T/S specs taken by Funk Audio, using the lower DCR coil:

  • Fs: 22.32 Hz
  • Re: 4.65 ohm
  • Qms: 5.52
  • Qes: 0.235
  • Vas: 296 L
  • Mms: 590g
  • Bl: 40.4 T*m
  • Le (1 kHz): 2.93 mH

The (BL)^2/Re comes in at 350 putting it among the top brass for motor strength, approaching the Rockford Fosgate TS3-19 and the B&C IPAL-21.  It appears to have a more compliant suspension than either of these, which aids low frequency efficiency in medium to large size boxes.  Compared to the TS3-19, the mass of the UH-21v1 moving assembly is lower, and the driver has greater upper frequency efficiency.  It also requires less mounting depth.  Compared to the IPAL-21, the UH-21v1 has more cone travel and a much more usable DCR.

 

I plan to install the drivers dual-opposed in two boxes with about 4-5 cuft per driver.  They will be designed to fill the spaces I have on each side of my equipment racks up front, about 40" wide, 28" deep, and 24" high.  At 4.5 cuft per driver, the Fb models at around 35 Hz and a low Qtc of around 0.4.  Each box will be run with a channel from a SpeakerPower SP2-12000 with drivers wired in parallel.   After a lot of research, these drivers appeared likely to offer the most ULF bass of any driver for the space I had to work with, and they should still be usable to frequencies well beyond 100 Hz.  They will also be insanely efficient above 20 Hz and a good fit to my room gain profile.

 

I also have an upgrade path if for some reason I find the headroom in the deep bass to be insufficient.  I can add a second SP2-12k amp and run each channel on a pair of coils in parallel for DCR of 1.25 ohm and then use frequency selective limiting to avoid over-driving them at the lowest frequencies.  I doubt I'll opt to go this direction, especially since I'm not even one to run my bass hot.  I just want as much ULF as I can get in this fairly large space, and lots of headroom for realistic high SPL peaks.  As it is, the numbers I'm looking at are quite absurd.  I can also envision moving these to horn-loaded cabinets at some point in the distant future but probably not for my living room.  :)

 

Dual-opposed Sealed Cabinets

 

Here is one of the cabinet sides attached to the bottom, with dados for the opposing front baffles, and for internal braces:

 

sub-cabnet-bottom-and-side.jpg

 

Here is one of the internal braces along with one of the front baffles:

 

sub-cabnet-baffle-and-brace.jpg

 

Here is the inside of the cabinet, right after gluing the major panels together.  The cabinet is actually divided into separate chambers for each driver.  I'm not sure what's up with the weird camera lens flare on the left.

 

sub-cabnet-glued.jpg

 

This is with the braces installed parallel to the axis of the drivers, connecting the baffles:

 

sub-cabnet-baffle-braces.jpg

 

Here is a cabinet after patching and sanding, ready for Duratex application:

 

sub-cabinet-ready-for-duratex.jpg

 

Here is look at the inside of the fully finished cabinet, without the driver:

 

wv-jGsQaEsdw_szT3xtNi2vUdzw-oXHBS8wG2VGB

 

I also ordered hand-machined anondized aluminum trim rings from Funk.  Each piece fits the frame profile perfectly and interlocks with the adjacent pieces:

 

0SQc9mNJGoMaEPOpTLLvMQi__h9DC6_EUM0BqnYa

 

 

FacQjgu1degTVkJ9HF6e2-Iw-OdECDqM5PQlIjl0

 

Impedance

 

UH-21v1-sealed-impedance.png

 

 

The resonance comes in at around 39 Hz and with a Qtc of 0.33.  Wow!  These things appear to be absurdly efficient.  The DCR is 4.75ish ohm, and I measured impedances of 8.5 ohm at 10 Hz and 14 ohm at 20 Hz :blink:.

 

 

Close-mic Response

 

This was measured approximately 6" from the cone, and a signal level normalized to 2V.  It should not be taken as quantitatively accurate but reveals the approximate native response shape:

 

UH-21v1-close-mic-response.png

 

 

In-room Sensitivity

 

Here are in-room sensitivity measurements, normalized for 2V (more precisely than above) and taken at the main listening position:

 

UH-21v1-sealed-in-room-sensitivity.png

 

The sum just assumes I send each sub an identical signal.  I can likely get a bit more output in some spots with help from some phase corrections.

 

My amp is rated for 126V maximum with 240V supply, so if the subs are capable, the amp can supply the current, and there's no compression, I'll see +36 dB with full power.  (NoteAt this time, I'm running on 120V power and can probably do closer to +33 dB.)

 

The impedance is 7 ohm (for two drivers in parallel) at 20 Hz, so the amp is unlikely to have any problem at or above there.  It appears that I now possess ridiculous output capability.

 

 

Edit: Added lots more pictures.


Edited by SME, 25 April 2017 - 07:01 AM.

  • Infrasonic, lukeamdman and chrapladm like this

#2 SME

SME

    Super Bass Overlord

  • Members
  • PipPipPipPipPipPipPipPip
  • 965 posts

Posted 24 August 2016 - 09:32 PM

Reserved.



#3 SME

SME

    Super Bass Overlord

  • Members
  • PipPipPipPipPipPipPipPip
  • 965 posts

Posted 24 August 2016 - 09:32 PM

Reserved.



#4 SME

SME

    Super Bass Overlord

  • Members
  • PipPipPipPipPipPipPipPip
  • 965 posts

Posted 24 August 2016 - 09:33 PM

Reserved.



#5 lukeamdman

lukeamdman

    Bass Overlord

  • Members
  • PipPipPipPipPipPipPip
  • 829 posts
  • LocationBurnsville, MN

Posted 24 August 2016 - 10:04 PM

Say what now?

 

EDIT:  Really looking for to this :D



#6 chrapladm

chrapladm

    Ultra Member

  • Members
  • PipPipPipPipPipPip
  • 271 posts
  • LocationAdelaide, South Australia

Posted 25 August 2016 - 06:58 AM

Glad to see you bought these drivers. I chatted in an email about getting an 18" version and was curious about T/S. So, looking forward to your updates.


"Music is not in the notes, but in the silence between."


#7 Ricci

Ricci

    Super Bass Overlord

  • Moderators
  • 1,067 posts
  • LocationLouisville, KY

Posted 25 August 2016 - 02:35 PM

I got a chance to place an early order for four of the upcoming Funk Audio GUJ21V2 drivers.  This driver is an underhung design, similar to the Focusworks GUJ18V1 and GUJ21V1 drivers, but it uses the magnet structure from the TSAD21v2 driver featured in the Funk Audio FW21.0 powered sub.  I custom ordered the drivers to use the coil from the GUJV1 in order to eek out just a bit more motor performance.  I'll post final T/S specs when I get them, but I'm anticipating a (BL)^2/Re of around 400 (SI units), approaching that of the IPAL-21.  Unlike the IPAL-21, however, I get a more reasonable DCR of ~5 ohm per driver, a more compliant suspension, and more linear cone travel.

 

Edit: Nathan tells me these are to be branded as "Funk Audio" rather than "Focusworks".

 

 

A 400 motor rating with an underhung coil? Are you sure about that? Underhung drivers sacrifice some motor strength typically. That would really be something out of the ordinary.

 

Nathan you got any details?



#8 Funk Audio

Funk Audio

    Advanced Member

  • Members
  • PipPipPip
  • 45 posts

Posted 25 August 2016 - 02:51 PM

s884891237978113168_p48_i27_w2560.jpeg


  • lukeamdman likes this

#9 Ricci

Ricci

    Super Bass Overlord

  • Moderators
  • 1,067 posts
  • LocationLouisville, KY

Posted 25 August 2016 - 03:58 PM

Cool...What's the gap overhang?



#10 Infrasonic

Infrasonic

    Bass Overlord

  • Members
  • PipPipPipPipPipPipPip
  • 779 posts

Posted 25 August 2016 - 04:52 PM

Sweeeettt!



#11 Funk Audio

Funk Audio

    Advanced Member

  • Members
  • PipPipPip
  • 45 posts

Posted 25 August 2016 - 04:57 PM

Cool...What's the gap overhang?

Geometric "underhang?" is 18mm, same as the GUJV1 motor. So has the same overall stroke capability but can maintain it down to lower frequencies in smaller box with less distortion.

 

Testing the first one of these, with ~5kw in ~200L box, reminded me of this scene, Loki as cone, looks like its just being thrashed, but takes it;

Hulk-vs-Loki-04.gif



#12 Infrasonic

Infrasonic

    Bass Overlord

  • Members
  • PipPipPipPipPipPipPip
  • 779 posts

Posted 25 August 2016 - 05:12 PM

5kw in ~200L?

 

Wow.



#13 Funk Audio

Funk Audio

    Advanced Member

  • Members
  • PipPipPip
  • 45 posts

Posted 25 August 2016 - 05:25 PM

5kw in ~200L?

 

Wow.

Didn't quite use it all under 20hz, that test box was a little larger than needed but yes it took it 20hz up. Around 120-150 L is about perfect, for ~3-5kw.



#14 Ricci

Ricci

    Super Bass Overlord

  • Moderators
  • 1,067 posts
  • LocationLouisville, KY

Posted 25 August 2016 - 06:11 PM

Geometric "underhang?" is 18mm, same as the GUJV1 motor. So has the same overall stroke capability but can maintain it down to lower frequencies in smaller box with less distortion.

 

Yes overhang of the gap beyond the coil. Or underhang of coil below gap depending on how it's looked at. Same as the v1 then...Cool. 



#15 SME

SME

    Super Bass Overlord

  • Members
  • PipPipPipPipPipPipPipPip
  • 965 posts

Posted 25 August 2016 - 07:38 PM

Didn't quite use it all under 20hz, that test box was a little larger than needed but yes it took it 20hz up. Around 120-150 L is about perfect, for ~3-5kw.

 

So I guess I need to plan on that second SP2-12k?  :D  Seriously though, I expect I'll be fine without it.  I'm worried that a single SP2-12k on four of these will blow my house apart, like this guy.

 

Maintaining silence about these drivers for the last couple months has made me antsy, to say the least!  The timing for me couldn't have been better.  After a brief love affair with the IPAL-21, I had a wish list of things I wished the IPAL did differently.  This driver looks like it fits that mold very well, at least on paper.  And considering what we've seen with Ricci's measurements of a GUJV1 driver and the TSADv2 driver in the FW21.0 system, I am confident that these drivers will deliver.

 

So one design decision I must make is how far away from my equipment racks to place the side-firing cabinets.  I definitely want to avoid slot loading one of the drivers, not just because it would break the D.O. symmetry but also to ensure excursion is shared equally.  I'm thinking of allowing at least 6, maybe 7 inches of clearance, but every extra inch of clearance sacrifices internal volume, given the space I have to work with.



#16 3ll3d00d

3ll3d00d

    Ultra Member

  • Members
  • PipPipPipPipPipPip
  • 413 posts

Posted 25 August 2016 - 09:15 PM

 

So one design decision I must make is how far away from my equipment racks to place the side-firing cabinets.  I definitely want to avoid slot loading one of the drivers, not just because it would break the D.O. symmetry but also to ensure excursion is shared equally.  I'm thinking of allowing at least 6, maybe 7 inches of clearance, but every extra inch of clearance sacrifices internal volume, given the space I have to work with.

 

I modelled it beforehand but the best fit was only found after I had actual data, see http://www.avsforum....omaly-what.htmlfor details

 

In retrospect, and this is perhaps not surprising, the loading is extremely sensitive to the nature (size, lossiness) of the slot. If you really want to eke out every last cc of your enclosure then build a test box of the appropriate size & measure it. 



#17 Funk Audio

Funk Audio

    Advanced Member

  • Members
  • PipPipPip
  • 45 posts

Posted 26 August 2016 - 03:47 PM

If you built the enclosure with panels attached on both sides in front of the drivers spaced away from the cabinet with standoffs, you would have perfectly symmetrical driver loading and you could place it tight to the wall, and also have something close to the other side as well without worry, the "slot loading" affect could be reduced by having both the front and top edges open, vs just the front



#18 SME

SME

    Super Bass Overlord

  • Members
  • PipPipPipPipPipPipPipPip
  • 965 posts

Posted 28 August 2016 - 08:31 AM

I modelled it beforehand but the best fit was only found after I had actual data, see http://www.avsforum....omaly-what.htmlfor details

 

In retrospect, and this is perhaps not surprising, the loading is extremely sensitive to the nature (size, lossiness) of the slot. If you really want to eke out every last cc of your enclosure then build a test box of the appropriate size & measure it. 

 

Interesting data.  I'm not sure the slot in your setup is necessarily lossy, just because a low Q value was needed in the bandpass model to make the frequency response appear to match your measurements.  I suppose it could be if a lot of the sound is passing through the walls and/or floor.  That's not too unusual in the mid bass.  But in any case, the slot definitely had a big impact on your response.

 

If I were to build a test box, I think I would have to build it pretty much like the final product to get a good idea of the loading effect.  So I will probably have to choose something and then see what I get.  At least I know my drivers will get more breathing room than yours do.  There's a 9" gap between the wall and where my rack cabinet side starts, so I think a lot of air will be able to escape behind the racks.

 

If you built the enclosure with panels attached on both sides in front of the drivers spaced away from the cabinet with standoffs, you would have perfectly symmetrical driver loading and you could place it tight to the wall, and also have something close to the other side as well without worry, the "slot loading" affect could be reduced by having both the front and top edges open, vs just the front

 

This is a very good idea, and I have to admit taking some inspiration from Josh Ricci's M.A.U.L. and contemplating something similar.  I gave up on the mini-M.A.U.L. idea when I realized I'd want a bigger box than I have room for.  But maybe a shaped slot on the sides is worth considering?  I wouldn't turn down a bit of boost in the 20-40 Hz range, but I'm guessing I'd have to give up too much box volume for that or I'd lose more upper bandwidth than I'd like.  I have no idea.  Maybe it's worth looking into.  I guess the closest analog is 4th order bandpass?



#19 SME

SME

    Super Bass Overlord

  • Members
  • PipPipPipPipPipPipPipPip
  • 965 posts

Posted 31 August 2016 - 08:43 AM

So I decided to learn Hornresp and used it to try to gain some insight into slot loading effects.

 

First off, the rack cabinet is only in the way of one of the two drivers.  The other driver is open to the room, so I kind of get clearance on that side "for free", but I'd lose that free space if I slot loaded that side too.  I'm still space constrained there because the front right part of the room is actually open to the front door and foyer area.  Instead, I'm thinking of facing the drivers into a single slot in the middle.  This has the added benefit of making off-axis dispersion of bass above 100 Hz favorable and a bit easier to predict.  With the top open to the room, I may see even more upper bass directed toward the sides than the front, which is beneficial for wide coverage in a frequency area where localization begins to matter.

 

So I tried modelling some cabinets with middle slots.  I took two different approaches.  In one approach, I simply defined a single conical segment of horn with constant cross-sectional area and length equal to the slot length.  This involves ignoring a bit of whining from Hornresp because the exit area of the slot is less than the area of the drivers until it gets pretty wide.  The other approach was to treat the region of the slot between the drivers as a throat chamber whose exit just so happens to be the same area as the single conical segment of horn that it's connected to.  The length of this imaginary horn segment is much less than in the first approach because most of the slot is allocated to the throat chamber.

 

Using these two methods yielded remarkably different results.  I had hoped these two models would perform similarly.  They do not.  The features they share in common are: a decrease in box resonance frequency, a boost in sensitivity and efficiency that starts in the mid band and grows with increasing frequency up to a strong peak somewhere in the upper end of the response.  There is a corresponding impedance peak here as well.  I believe excursion is also reduced somewhat over most of the range.  Unfortunately, the location of the first peak from the slot load varied substantially depending on what modelling approach was used.  The model with the throat chamber had a much smoother top-end as well.  However, regardless of the modelling approach, I did not find a slot-load cabinet design with suitable dimensions whose first peak fell below 100 Hz.  Enhancement of output in the 20-40 Hz range was pretty poor, IIRC hitting +3 dB at 40 Hz in the best case.  I suspect my problem here is simply one of geometry.  I don't know horn theory too well yet, but I imagine I need either a longer slot length or greater exit area than I can accommodate without giving up valuable internal volume.

 

Another concern I have with slot loading, which IMO may be relevant to the M.A.U.L. is the effect of varying driver stroke on slot/horn geometry.  For example, with the driver pair facing one another and both moving with 2 inch peak-to-peak stroke, that effective width of that part of the slot (that part being almost all of it in my case) is modulated across a range of 4".  With drivers at 6-8" apart (at rest), the consequences are pretty severe.  I modelled and compared the two extremes, and the responses are totally different above 100 Hz.  This leads me to worry about inter-modulation distortion.  Even below 100 Hz, the response differs by more than a dB at 2" PTP stroke.

 

At this point, I'm leaning toward going with the original design with the drivers facing outward and hoping 6-7" is enough clearance from my racks to prevent "serious problems".  I gather I could take the time to learn Akabak and improve on the simulations for the center slot, but I don't think I'm going to gain any new useful information.  Just to make sure, I did some additional HR runs with a larger 12-14" slot to ascertain the effects on a single driver loaded into half that space, i,e, the 6-7" spacing I intend to use.  I increased the exit area somewhat to account for the top of the slot being open.  I assume the bottom and rear are closed because there is less room for air flow in those directions.

 

My simulations suggest there is still quite a bit of loading effect in the upper frequencies, but the wider slot is much less sensitive to relative cone displacement.  IIRC, mid bass varies < 1 dB with 2" PTP stroke..  This is a big plus for controlling distortion.  On top of that, only half my drivers are loaded this way, so this type of distortion should drop by half again.  The impedance curve is also much more stable through the stroke.

 

The box resonance frequency does shift slightly for a loaded driver vs. an unloaded driver, when simulated in isolation.  One consequence of this is that the motion of the drivers will not be perfectly opposed, particularly around the resonance frequencies.  What if they share the same air volume?  I don't really know.  I could try to model it with Akabak, or just try to reason through it with theory.  Or I could just seal the chambers behind them separately and not worry about it.

 

I'm not as worried about off-axis dispersion with this design either.  The driver that's loaded should have quite a bit more acoustic output than the other driver and so the dispersion won't be as narrow as I originally thought.  In fact, the slot formed with the single driver and cabinet will disperse upper bass very similarly to the middle slot.  Additionally, the direct upper bass sound from the driver on the outside face will likely be weaker on the opposite side of the room due to baffle diffraction effects.  I may yet get a favorable dispersion pattern.  Or I'm just waving my arms around and hoping something good will happen.

 

Even if it doesn't, the 100-200 Hz capability is more of a "nice to have feature".  The real purpose of these subs is the deep bass and ULF.  I am on a suspended floor, so why not play as low as possible?  Basically, I'll get as much if not more output at 5 Hz than at 10 Hz, based on my signal chain and room gain measurements.  Unfortunately, I'm not especially confident in my room gain measurements in an absolute sense.  After thinking about it, I realized room gain is very tricky to measure correctly in an absolute sense, contrary to what Bossobass Dave and others may suggest.  Actually, it's rather tricky to even define as seating distance comes into play.  But anyway, from the data I have, I think 110 dB @ 5-10 Hz at the seats is a reasonable expectation, and I could end up with more like 115 dB.  We'll see.



#20 3ll3d00d

3ll3d00d

    Ultra Member

  • Members
  • PipPipPipPipPipPip
  • 413 posts

Posted 31 August 2016 - 11:31 AM

The drivers facing each other version sounds like the way a PPSL is modelled. AIUI this is generally described as an OD so S1-S2 forming the rear of the slot and S2-S3 the front of it. This means you'd set S1-3 to some constant value according to the area of the slot and set the lengths accordingly, set to a sealed back chamber and use Vrc/Lrc as usual to define the size of it.






0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users