D75 chamber sizing

4thtry

I made a small 3" diameter x 1.75" long PVC chamber for my new (but slightly used) D75MX-21 midranges (thanks again, Dan!).   Below are a few construction pics with comments.  Qtc with this test chamber came out to 1.2 (see DATS V2 screenshot).  Is this an acceptable Qtc for a midrange of this size?  Should I make the chamber bigger or smaller?

When I removed the green felt along the perimeter to glue the PVC in place, I discovered a large number of vent holes. 

  The PVC tube blocked off these vent holes, so I ground clearance chamfers using my dremel tool.

I made a cap out of 1/4" masonite and then glued a white rubber bumper to the center of it to spread out internal resonances.

Finished chamber:

DATS V2 measured Qtc with test chamber sealed in place with caulking cord:

FR measurement at 31", midrange placed on floor flush with thick blanket:

Wolf

For what it's worth, the Dyn D76AF chambers are about another half of that volume, and they are 3" domes, but I've had 2" domes also at about the volume you have there. Honestly, it depends on your wants/needs. 1.2 will provide an earlier rolloff, but bump the response a bit (not unlike a waveuide). This can be beneficial and allow smaller parts sizes for the same F3 of larger chambers.

PWRRYD

That frequency response looks good.  Interested in what the HD measurements look like.

4thtry

Thanks.  The DynAudio D76AF spec sheet shows a Qts of 0.9, so I think I need to go a bit larger.  I'll slice off a 3.5" length of PVC, re-measure, and report back.  I'll also run a few HD measurements.

Kornbread

What size pvc is that?  Instead of going to the trouble of cutting reliefs into the pvc, what about taking the mid to the local hardware store and going through their bin of pvc pipe and fittings.  Bet you'll find something that will work w/out all the fuss.  If you can't find it in the plumbing section, try the grey pvc in the electrical section.  

Wolf

Fill it with Ultratouch,

4thtry

Kornbread said:

What size pvc is that?  Instead of going to the trouble of cutting reliefs into the pvc, what about taking the mid to the local hardware store and going through their bin of pvc pipe and fittings.  Bet you'll find something that will work w/out all the fuss.  If you can't find it in the plumbing section, try the grey pvc in the electrical section.  

You are correct.  This is  3" ID PVC (sch 40) and the 3" coupler used to join two 3" pipes together is almost the perfect fit to the magnet.  I bought one coupler, which, when cut in half, will work for 2 of the mids.  Thanks.

4thtry

As another test, I doubled the enclosure size from 3" ID x 1.75"L to 3" ID x 3.5"L by stacking two enclosures end to end (have not used the coupler yet).  Qtc dropped from 1.204 to 1.029.  Fc dropped from 386.9 to 319Hz.  I ran an HD test with the 1kHz SPL level set at 95dB.  Microphone tip was 32 inches from the dome.  The attached HD graph is hard to interpret, so I pulled the following HD data using the mouse.  Graphs are also attached:

Freq   Percent
100Hz  10.75%
200    5.52%
300    2.19%
400    2.43%
500    2.29%
600    0.79%
700    0.52%
800    0.64%
900    0.52%
1000   0.64%
1500   0.62%
2000   0.44%
2500   0.35%
3000   0.57%
3500   0.43%
4000   0.47%
5000   0.28%
6000   0.18%
7000   0.21%
8000   0.15%
9000   0.13%
10000  0.03%

4thtry

I noticed that the response below 1kHz dropped off a little bit when I increased the enclosure size.  For my next test, I will cut the enclosure in half, raising to Qtc to 1.5 to see what happens.  Maybe this will give me the flattest measurement.   

4thtry

Wolf said:

Fill it with Ultratouch,

I have been using some light polyfill material so far.  Will test Ultratouch as well.

PWRRYD

I would definitely recommend using some heavier damping material like the Ultratouch.  That or some fiberglass insulation.

Tom_S

The HD is sure telling you when these are not happy.

Wolf

Yeah, 600-2.5k is the sweet spot for sure.

DanP

Lol I had these things for 5 years and never even took a frequency measurement.  Bill's had them for 2 weeks and already has them figured out!

4thtry

Wolf said:

Yeah, 600-2.5k is the sweet spot for sure.

I agree.  I'm thinking a TMW or TMWW with a SB19 on top crossed with an initial target of 3k or so.  Then I will use one or two 5.25" woofers crossed with an initial target of 800Hz or so.   Maybe a single ND140 in a 0.75cu.ft cabinet, tuned to 40Hz or so.  Baffle would be about 17H x 6.5W.  Depth about 14".   Power handling would be limited, but I want to keep the baffle width to 6.5" and I have been unable to find two reasonably priced 5.25" 8 ohm woofers that would work well in a 0.75 cu ft box.   

4thtry

PWRRYD said:

I would definitely recommend using some heavier damping material like the Ultratouch.  That or some fiberglass insulation.

I tested this using my 3x3.5" enclosure.   Qtc dropped from 1.2 to 1.1 when I inserted the Ultratouch.    So, Ultratouch must be a very reactive type material, similar to fiberglass.  I did not know this.  Thanks for the tip.  

4thtry

I made a flange out of two layers of 1/8" thick tempered hardboard (HDF).  I used the type that is smooth on both sides.  Both layers are 5.25 x 5.25".  The top layer has a 3.625" hole cut to match the dome.  I partially rounded over the internal edge with a 1/4" round over bit.  The bottom layer has a 4.5" hole to clear the magnet and terminals.

When glued together, this flange will be very strong.  I think I will drill 8 mounting holes instead of 4 to prevent flange distortion as I tighten the screws.  When finished in flat black with a satin top coat, it should look like a brand new driver!

 

4thtry

I glued the 1/8 + 1/8" flange sandwich together with Titebond II.  Then, to prevent warping, I inserted all the finished flanges into a makeshift "press" for drying.  Also, after several stages of glue "squeeze out" and clean-up, I placed a single layer of paper towel between each flange in the press to act as a moisture wick.  All flange surfaces were completely dry when I inserted these towels, so they should not stick and create a mess later on.   I'll probably leave the flanges in this "press" for a least a week before taking them out for painting.
 
I made a total of 4 flanges.  This gives me one extra in case one gets messed up.

4thtry

I finally got around to selecting a chamber size and ran up a new set of measurements (which look very good).  I used a 3" PVC end cap, silicone glued onto the back.  My local Menards had two types, flat bottom and rounded.  I selected the rounded type, which should help to spread out internal reflections a little.  It was a very snug fit at first, so I turned or shaved out about 1/32" x 1/2" on the internal opening to get it to slide onto the magnet well.  Thanks, Kornbread, for the suggestion.  No need to radically dremel out clearance holes this time.

  I decided on a target QTc of .78 or so, near maximally flat.  Fs came out in the 350 to 390Hz range, which should be perfect for an 800Hz crossover.  I stuffed each end cap with 8.5 grams (not ounces) of Acousta-stuf, which hit the target perfectly.  I did not use ultratouch, as my tests with this material dropped the Qtc lower than I wanted (in the .5 to .6 range) which rolled off the bass too much.

I finished the flanges with 2 coats of Sealcoat and then 3 coats of semi-gloss enamel.  The Sealcoat dried fast, but I had to wait at least a week between each coat of enamel, lightly sanding with 220 between each coat.  I glued the flanges on with 100% silicone black gasket maker, careful not to get any squeeze out on the dome suspensions.  I pealed and transferred the driver label to the outside before sealing the chamber in place.  They now look like 3 "brand new" midrange drivers, ready for a new project!!!  I'm thinking high end:  Vifa D75's in the middle with JR's Esoteric tweeters on top and the Rival 176 door prize woofers on the bottom.  

Kornbread

4thtry said:

I finally got around to selecting a chamber size and ran up a new set of measurements (which look very good).  I used a 3" PVC end cap, silicone glued onto the back.  My local Menards had two types, flat bottom and rounded.  I selected the rounded type, which should help to spread out internal reflections a little.  It was a very snug fit at first, so I turned or shaved out about 1/32" x 1/2" on the internal opening to get it to slide onto the magnet well.  Thanks, Kornbread, for the suggestion.  No need to radically dremel out clearance holes this time.

    Hey, you're welcome, glad I could help.  Spend a lot of time at the local hardware store.  Most everyone there knows me by name. 

ani_101

That is neat - the Label on the end cap!!!

4thtry

ani_101 said:

That is neat - the Label on the end cap!!!

I was extremely careful when removing the labels.  Took me several hours to do this.  I wanted to preserve them for future generations of speaker building aficionados.  After all, we would not want anyone to confuse these drivers with ScanSpeak or DynAudio D76's.  Sure wish I could find a spec sheet for this driver.  All of my internet searches have come up empty.

Ed_Perkins

Bill -  Is this the Vifa D75 MX-41 mid?  Lot's of hits (e.g. https://www.hificollective.co.uk/components/vifa-d75mx-41-8ohm-midrange-driver.html ) on that one, so I'm guessing it's something different.

Wolf

Nope, these are much older than that.

4thtry

They are the D75MX-21 8 ohm units.   The label is a little damaged toward the end.  Dan said that he had them for 5 years, so he must not have been the original owner.  I looked at the pdf spec for the -41 version and find that the FR response graph is substantially different from my -21 units.   I'll post my latest FR measurements with the drivers flush mounted on my large test baffle for comparison.   

4thtry

A, B, and C marked units are very similar'

The dip at 650Hz is due to baffle size:

I added three 20x30 foam core sheets to the baffle temporarily to smooth out the 650Hz dip:

4thtry

 I'm going to make a foam core expansion interface for my 24x30 test baffle which will convert it into an IEC 268 test baffle.   Then I can start measuring like the "pros"