The #1 Internet Source Of Audio, Video and Speaker Building     Components, For Home, Office, Mobile And Prosound.       Call: 800-338-0531 Site Last Updated   Hello music lovers and welcome.  This is the the long anticipated and somewhat overdue account of the odyssey through Tweetertown and beyond of a humble tweeter of noble genetic lineage but far eastern birth.  The purpose of this quest was to take what is quite possibly the best tweeter bargain in DIY audio today and make it even better.  The journey, as you will soon see, has been a  rewarding one, and I hope you will read on and learn how this "boy" became a "MAN".

Objectives of the 275-070 Modification

 

The ultimate goal of this adventure in "tweaking" is to coax a higher level of performance from this tweeter than the stock version is capable of delivering.  So this said, the task immediately before me was to decide what mods could be done to achieve the desired heightened fidelity.  To determine what course of action to take I studied the steps taken by the makers of the '070's "blueblood" kin folk to see what they had done.  I also wanted to go about it in a series of "Phases" moving from the simple to the complex.  This would serve two main purposes 1.  It would allow the potential tweaker to pick a level that matched their desired level of expenditure of skill/time/effort/money. 2.  It would allow a stratified evaluation of individual mods

 

The "Phases"

 

Phase I

 

The first, very obvious, mod that would be made was the addition of a sealed rear chamber to lower system Fs and Q.  Virtually all highly regarded dome tweeters available to the DIY community sport this feature.  You may be wondering what it is about this rear chambering that seems to make it a necessary component for achieving high quality tweeter operation.  The answer lies in the fact that a tweeter is governed by the same laws that apply to a low frequency system.  A tweeter consist of a driver and an enclosure just as bass systems do.  Have you ever noticed that tweeters often exhibit a second order(12db/octave) low end roll off, just as sealed box woofers do?  You are probably familiar with the profound effect that box volume, shape, etc. have on a woofer's Fs and Qts and therefore Fcb and Qtc.  The same applies to the operation of a tweeter only on a miniature scale.  In its Phase I incarnation the rear chamber is constructed of materials that can be obtained inexpensively at your local Walmart and Home Depot.

 

Phase II

 

The defining feature of the Phase II level is the inclusion of a modification to the voice coil leads.  This mod, as is the rear chamber, is a very common feature on high end tweeters most notably the the Morel MDT 30 which the '070 bears strong resemblance to.  In this mod the stock leads are removed from their stock location and reconfigured in such a way as to increase their length.  This increase in length makes them vastly more compliant thus facilitating free movement of the dome/coil assembly.  I feel this mod is of great value in giving the diaphragm the ability to react to small as well as large signal input.  This freedom is evidenced by a reduction in Fs of around 100 Hz beyond that of the Phase I.   This increase in free movement will also facilitate the use of lower crossover frequencies made possible by the reduction in Fs produced by the rear chamber.

 

Along with the leads mod, the Phase II's receive more esoteric rear chamber stuffing material.

 

Additional Phases?

 

In the planning stages are the following stages.

 

Phase III "Aegis":  This Phase is the Phase II with the addition of a bucking magnet for reduction of stray magnetic field.

 

Phase IV "33":   This Phase is really exciting to me.  The "33" will be a triple magnet  maximum effort.  I will be attempting to push the '070 to the very edge of the envelope.  A veritable "UberDayton".  This version will sport a custom aluminum faceplate featuring totally flush mounted fasteners among other things I have brewing.

 

Phase V "Gemini":  This "beast" is simply two Phase IV's mounted on a custom faceplate as closely together as possible.  Driver distortion generally increases with increases in input power.  Two drivers sharing the load equally should cut distortion levels by a very substantial amount.  How will this effect interdriver radiation patterns?  Stay tuned tha' Mayor intends to find out.

 

 

Performing the Modifications

The Rear Chamber

 

Raw materials you'll need:

 

For Phase I

 

1- length of 1" schedule 40 PVC pipe (a scrap piece a couple feet long will be plenty)

1- 1" schedule 40 PVC pipe cap

1- 1' square piece of kraft felt (available at Wal-Mart in many colors, 20 cents/square)

1- 1" schedule 40 PVC pipe sleeve connector

1- Bag of polyester fiber fill

1- Package J-B Weld epoxy cement

 

For Phase II

 

All of the above plus:

 

1- Piece of SAE grade F-11 felt, 1/8" thickness, 87% wool content (available from McMaster-Carr in 6' widths, 1 lineal ft. minimum order, cost: about $12)

Okay,  after getting all the raw materials together it's time to convert them into finished goods.  Begin by cutting a 1 1/2" length of the PVC pipe.  This will form the body of the rear chamber.  For this I use my 12" Dewalt compound miter box saw (yeah, I'm braggin').  Another good method would be to use a miter gauge on a table saw.  If you don't have power tools I recommend buying a hand mitre box saw and a hand mitre box (available at most hardware and home centers).

 

Now its time to cut the mounting sleeve which will be cemented to the back of the tweeter and will hold the chamber body in place.  Cut a 5/8" piece off the end of the sleeve connector.  It is important to make a good square cut here because any error will cause the rear chamber  to tilt and the chamber body  will not bottom uniformly against the back of the tweeter.  This bottoming holds the kraft felt in place (more about that later).  The way I do this and keep all my fingers still attached to my hand is thusly.  Cut a piece of 1" sch. 40 pipe about 1' long.  Clean any flashing off that may have been produced by the cutting process.  Now, slide the sleeve connector onto the pipe.  This should ensure that the two pieces are held together as a single unit so that true cuts can now be safely made. Since the outside diameters of the pipe and sleeve connector are different it will be necessary use a spacer with the sleeve overhanging it to make the cuts.  When using my power miter box saw I use a couple of pieces or 1/2" plywood for this, one on the bed of the saw and one on the rear fence.  If you use a table saw, a single piece placed where the miter gauge and the saw table meet will provide the necessary spacing.

   

Next cut a disc of felt 1 1/4" in diameter.  You can use a piece of the pipe as a pattern for marking and scissors to do the cutting.  All that remains to be done now is to weigh out a 2 gram piece of polyester fiber fill which will be the stuffing for the rear chamber.  Your collection of parts should now look like figure 1(Finparts.jpg).   For the Phase II rear chamber follow the above steps with the following changes.  Reduce the amount of fiber fill to 1 gram and cut three 1 1/4" discs of the F-11 wool felt.

 

---

Take the following steps to install your rear chamber

 

1.  The first order of business is to remove the foam from the rear surface of the tweeter uncovering and exposing the vent.  I found the best way to do this was to first scrape the foam off using a razor scraper, see Fig. 2.  Prince (I'm sorry, I meant to say the artist formerly known as Prince) dropped by to give me a hand with this one wearing his trademark purple nitrile gloves.  Great guy that Prince, and an avid speaker builder.  Ya know, they don't call him "His Royal Badness" for nothin'!  Yeah Baybee!  (Just kiddin' your Royal Badness....If you read this please don't sue me!) 2.  After you have scraped as much of the foam and it's adhesive off the backplate as possible,  the rest can be removed with lacquer thinner.  Wet a small piece of cotton cloth with thinner and squeeze out the excess.  Now rub the remaining adhesive until it dissolves and is removed.  Be careful not to let any of the thinner run down into the vent.   After you think it is all removed, wet a fresh piece of cloth and clean the surface a second time.

 

3.  After the solvent has thoroughly dried from the rear surface, plug the vent (be sure to do an especially good job doing this).  Now using fairly coarse sandpaper (I used 60 grit) rough up the rear surface especially the area where the mounting sleeve will be adhering to the backplate.  As tempting as it is to skip this step, don't do it.  I did, and let me tell you the epoxy just wouldn't hold.  After the roughing up is completed, remove ALL of the filings before carefully removing the plug.  A damp (water) wash cloth does a good job of cleaning off the filings.  Now remove the vent plug.  Your tweeter should look like Fig. 3.

Adjusting the rear chamber volume

During the development of the rear chamber just described I experimented with chambers of various sizes.  I found that chambers of too great a volume lead to the occurrence of spikes in the impedance curve at frequencies between 5- 10K.  On the other hand a chamber of too small a volume produced curves very similar to stock.  Through trial and error, I eventually arrived at what seems to be the correct volume.  I did,  however, design the chamber to have the ability to vary its volume to some extent to suit the individual requirements of  a given tweeter.

A tweeter will react to changes in rear chamber volume much like a sealed enclosure low frequency driver.  Lessening the volume will cause a rise in Fs and Qtc while enlarging it will have just the opposite effect.  The goal is to find the point in the volume range where the Fs ceases to drop.  At this point the tweeter will have reached a condition were Qts equals Qtc.  Essentially there will exist a "free air" condition.

If you don't have the ability to generate impedance curves or just don't want to, I've found that with a spacing of 1/2"-5/8" between the mounting ring (which has the chamber body inserted and bottomed out in it) and the end cap you can expect very good results.

Modifying the Voice Coil Leads

Things you'll need:

1- #1 darner (sewing needle)

1- Pencil type soldering iron (around 30 watts) with small tip

1- Small package of solder (silver bearing is nice but plain old 60-40 is easier to use)

1- Scrap piece of 2" x 4" lumber about 6" long and a drywall screw

1- Pair of tweezers or better yet forceps

1- Magnifier on a stand (not absolutely necessary but it would help)

1- 2.5 mm allen wrench

1- Smal standard screwdriver

1- Rainy day and lots of patience :-)

The Procedure

Here's how to perform this mod step by step:

IMPish Behavior - Let your impedance curve be your guide

I was guided in this moding expedition by basically two sources of  quantification and evaluation.  The first being impedance curves that I generated using my "trusty" Woofer Tester.  I have said it many times before and I'll say it again now.  The Woofer Tester is the best two hundred dollars I've ever spent on speakerbuilding.  I just keep finding new uses for it!  The speed, ease, and accuracy with which it generates impedance curves is remarkable.    It was an invaluable testing tool during the duration of the development  of the rear chamber.

The second being my ears......which is really the true master we're all trying to please anyway, right?  I mean, not counting our wives....

The first order of business was to generate a stock impedance curve. It would serve as a baseline with which to compare subsequent curves produced after modifications were made.  We will trace the progression of a typical '070 as it moves from stock to the Phase II level of modification.

Altered States

The following shows the effects on system impedance displayed at various stages of tweeter modification.

1.  The first action taken was to generate an impedance curve of the tweeter in stock form see fig. 20.  As can be seen in the graph, the Fs is around 1100 Hz and the impedance maximum is about 9 ohms.  The rough nature of the curve above 7.000 hz. is due to the Woofer Tester and not representative of the tweeter's actual impedance curve and should be disregarded.    

Figure 20. Impedance Results

2.  When the polepiece vent is uncovered several things occur and are indicated by the resulting impedance curve fig.21.  The single peak at Fs is replaced by what at first appear to be two peaks,  but after closer inspection it becomes evident that there are actually three present.  If you carefully inspect the peak centered around 700 hz you can see that it is not a single resonance but instead is a composite peak comprised  of two resonances.  There is a slight "swag" between the two maximas occurring between 550 and 600 hz..  I feel the lower peak is the actual fundamental resonance of the tweeters moving system while the upper is a pipe resonance produced by the uncapped vent.  Then at around 1800 hz another peak is observed. I believe this is a secondary pipe resonance also produced by the polepiece vent.

Figure 21. Impedance, Polepeice vent uncovered

3.  The addition of a rear chamber brings about several changes in the impedance curve fig. 22.  The driver fs is elevated by around 100 hz.  The primary pipe resonance is raised in frequency by  something in the neighborhood of 300 hz. And the secondary impedance resonance is damped to near elimination.  The impedance curve featured in the illustration is of a Phase I unit ("Uneasy E" to be exact) and although generally effective the superiority of the Phase II chamber is clearly evident (example later on).  Also the sharp peak of the primary pipe resonance shown is not really characteristic of the average unit.  It is usually not that high Q (sharply peaked). Its zenith is more commonly of a smoother nature.  I use this example simply because it is the only one I could locate at the time of this writing.  It does allow me to point out, however, that each of these tweeters seems to be as individual as people are, and although each will exhibit similar behavior the frequencies of occurrence may be shifted somewhat from the ones in the examples.  This "individuality" stems from differences in manufacturing tolerances.  What I'm really saying is don't expect identical impedance curves unit to unit, but instead look for trends and strive for as much consistency as possible.  

Figure 22. Impedance with rear chamber present

4.  Adjusting the ferrofluid in the magnetic gap is the last step in the tuning process.  Most of the '070's I've seen come from the factory with too much ferrofluid in the magnetic gap.  I believe the manufacturer does this intentionally  as an "insurance policy" against failures in the field.  They probably feel that the loss of a some resolution is better than risking having a distributor plagued with burnt voice coils.  We, however, are not constrained by the same concerns. Optimum performance is our goal so let's adjust that ferrofluid shall we?

The trick to adjusting the fluid level is removing enough.......but not too much.  First, I believe I must state the relationship between impedance magnitude at fs and the amount of magnetic fluid in the gap.  They have an inverse relationship.  When the fluid quantity goes up impedance magnitude goes down, and vice versa.  So we can use this impedance magnitude as a direct indicator of fluid level.  Now all that remains is to establish what is "ideal" and therefore a goal.  I chose to use the '070's illegitimate cousin as the source of my goal...the Morel MDT 30.  I reasoned that Morel had spent countless man hours determining what is ideal and that they're published impedance curves would surely represent their findings.

According to Morel's curves a 10 0hm peak at resonance represents the best balance between damping, resolution and, cooling.  If after moding your tweeter it has a Fs of less than 10 ohms (most will) it may be adjusted up by following the procedure below.

To adjust the magnetic fluid level follow these steps:

A.  Make an impedance sweep to serve as a baseline.

B.  Remove the faceplate and dome assembly.

C.  Using a lint free tissue (such as kimwipes or paper toweling) remove a small amount of magnetic fluid from the voice coil.

D.  Re-assemble the tweeter, but before installing the faceplate and screws rotate the dome/coil assembly three turns in both directions this will help to redistribute the magnetic fluid. (Note: this procedure can  be used on a stock unit which has an unusually high or strangely shaped impedance peak which can be caused by an over accumulation of ferrofluid in the magnetic gap.  However, the effects of this exercise may be temporary because the fluid will return to the gap after a period of use.)

Run a music signal through it for at least 30 minutes to allow all remaining ferrofluid to re-gather in the magnetic gap.  Be sure to protect the tweeter with a high pass filter.  I used a first order filter comprised of Dayton poly caps yielding  around 7.5 microfarads.  The gentle roll off also allows enough passage of frequencies from the lower range of the tweeter's operation to produce sufficient excursion to aid the ferrofluid in its migration to the gap.

E.  Run another impedance sweep and note magnitude at Fs.  Repeat as necessary.  When nearing the goal remove a very small quantity of ferrofluid at a time to prevent overshooting the goal.  Remember, you can take it out but you can't put it back in (unless you have a supply of tweeter grade fluid)

Note:  Should you run an impedance sweep before passing a music signal through the tweeter after removing some fluid, you will observe that the Fs impedance will rise what seems like a drastic amount.  Don't be alarmed.  After the music signal is ran through the unit,  the ferrofluid will pool back into the magnetic gap and impedance will fall back down quite a bit.

 Setting the Curve

How close can a '070's impedance curve come to the Morel ideal you ask?  Well, in my opinion astonishingly close.  Let's compare.

I own a pair of MDT 30's which I've had for quite sometime,  and I love the way they sound.  Anyone who has been around the P.E. Tech Board for awhile knows I love the Morels.  Along with they're excellent sound quality, they have always been known for their unusual "camel's back" impedance curve.  There's a graph of one of my 30's pictured in fig.23. 

Figure 23. The Morel MDT 30

Very typically Morel.  Now take a look at the impedance graph of a Phase II '070 fig. 24.  The similarity is amazing.  My woofer tester automatically switches over to a 20 ohm graph for anything over 10 ohms, but the closeness between these two graphs is still readily apparent.

Figure 24. Phase II Impedence

The Tweetertown 4 World Tour

About now you're probably saying to yourself, " Impedance curves are important but you can't hear them.  How do the moded tweeters sound?!".  Well, that's exactly what I wanted to know.  Or more to the point,  how do they measure?  I was able to listen to them (Alot!) and certainly thought I heard a distinct improvement but the "placebo effect" is strong and the need for objective testing was undeniable.  It was time for the "Tweetertown Four" to hit the road in search of truth.

Enter P.E. Paul (AKA Pepe, Pauly, Pavilion, Tha' Man, Mr. Smartypants, etc. and the list is still growing).  I must take the opportunity here to say that Paul is truly a "prince among men".  I've been posting at the P.E. tech board for a long time now and watched it grow and flourish under Paul's direction.  We ALL own him a debt of gratitude.  Paul, you truly are "Tha' Man"!  From the very beginning of this little adventure Paul had been involved with its development.  It had been agreed early on that I would perform the "black art" and when the spell was done the spellbound subjects would be whisked away to "Ye Olde Pauly's Speaker and Scnooker Shoppe" for LMS evaluation. 

 The time arrived for the "men", as they had come to be known,  to make the journey to Paul.  I built a suitable "strongbox" knowing that those fine young men in brown would probably use the package to practice their favorite WWF moves,  and man they didn't let me down. The trip was not a smooth one.   They reached Paul 3-4 days late and suffered some damage in transit.  One of the Phase II's had a cracked faceplate and was thrown badly out of adjustment.    I will forever have the mental image indelibly etched in my mind of a uniform clad "ups" man "piledriving" my package,  and once the dust settles whispering gently in a soft voice to the carton, "Don't take this A** beating personally son!"

When Paul finally received the tweeters, and after assessing the shipping damage, he set about the task of measuring their freq. response and impedance using LMS.  Here's a couple of photos taken by Pepe of "tha' Men" shortly after their arrival at his home figs. 25 & 26.

Figure 25. The Tweeter Box Arrives

Figure 26. The Phase IIs Revealed

All testing was performed with the tweeters mounted on a Dayton BR-1(Budget Reference-One) baffle that measures 8-5/8" wide x 14-1/4" high.  The tweeter is oriented 1" off the centerline in this design with a center to center driver spacing of 5-3/4".  The SPL tests were performed outdoors.  Here's a shot of one of the Phase II's mounted in Pauls Br-1 enclosure being tested fig.27.  Note the microphone visible in the foreground.  The mic was moved up from the one meter test distance for the sake of the "photo op" by Pepe himself who, btw, also took the picture

Figure 27. Phase II equipped BR-1 Ready for LMS

Here is a composite curve produced by measuring five stock units Paul had on hand. fig 28 The units were measured individually and then a curve that represents the average response of the individuals was produced.   These tweeters came from the same production run as the ones I modified.  This provides an excellent baseline response curve.

Figure 28. the Average of Five Stock Units

Now, compare the above averaged response with the Phase II's SPL curve figs. 29.  The SPL  is very much smoother with added extension on the lower end with response staying within a + or - 1.5 db window from around 1,250  to 10,000 hz.  Gone is the "sawtooth" resonant behavior showing good resonance control in the rear chamber.  Sadly, the "swag" in the top octave is still present but early reports indicate that the this problem has been substantially improved by use of a different doping compound in the new generation of '070's.  Even with the dip in the top octave response, the overall performance is very good and the sound is detailed and smooth.  

Figure 29. Phase II SPL Curve

Also, as fig. 30 clearly shows, the fs has also been drastically reduced.  The average fs of the five units used to establish the stock baseline is 1,207 hz while the Phase II has a fs of 547 hz, a reduction of more than half.
 

Figure 30. Dramatically Reduced Fs.

With these tests complete and the results favorable, I must admit my spirits were high. However,  there was still one area that was in need of testing especially considering the new found low frequency potential of the '070. That being distortion.

LMS is an absolutely superb testing system, but it doesn't have the ability to test driver distortion.  Things were looking bleak.  Paul and I discussed our options and decided that what "tha' MEN" needed was a good "fire and brimstone" sermon from a man of the speakerbuilding gospel.....who also happens to own a CLIO system.  Enter Rev. Wayne.

Speakerbuilder.net (Wayne's World)

Paul contacted the "Good Rev." Wayne Jaeschke and asked if he was interested in being the next stop on the "World Tour". With characteristic enthusiasm W.J. agreed.  As all who have been around the P.E. tech board or visited his website (Speakerbuilder.net) know, Wayne's a great guy and the most prolific speakerbuilder around.  I don't know the secret to his productivity but I think it has to involve one or more of the following:  sleeplessness, a clone, elves, divine intervention, a genie, lack of kryptonite, lots of coffee, or spanish fly. 

With no further ado, Pepe sent tha' MEN off on their way to church camp.

Fortunately, the WWF wannabe's must've had the day off during this trip for the package arrived at tha' Revs Church of Low Distortion without incident.

Tha' Rev performed a thorough battery of tests on the tweeters (primarily the Phase II's) and came up with some very promising numbers.  As you can see in the graphs,  second order distortion was greatly reduced  fig. 31 (stock '070) and Fig. 32 (Phase II).  The mods achieved an improvement in distortion of 47%.  A very worthwhile gain in performance.  

Figure 31. The Stock 275-070

Figure 32. The Phase IIs step up to the plate

Wayne also performed Energy Time Curve testing on a stock tweeter and a Phase II.  This test measures how a tweeter responds to a sudden pulse of current.  It is a very valuable test in determining the presence of resonance.  The results for the Phase II and the stock unit are shown in figs. 33 and 34, respectively.

Figure 33. Phase II Silk Dome

Figure 34. Unmodified 275-070

The Phase II exhibits slightly more resonant behavior in the .25-2 millisecond range.  The cause of this I'm not sure of.  I suspect that this is caused by the primary pipe resonance produced by the pole vent.  I wish I could see an ETC for a MDT 30 to see if it shares this behavior.  In any case its magnitude is below -40 db and effect is probably negligible.  As many regular posters at the board know David Ralph (dlr) is doing work on these tweeters as well and has the equipment to perform waterfall testing.  Perhaps his results will shed some light on this area.  I will be working to eliminate it in the subsequent Phases that are now in the planning stages.

Summing Up

In summary, I believe that all objectives set forth in the beginning have been met or exceeded.  Technically, the fs has been reduced by over half which will facilitate lower crossover points. Also second order distortion has been cut close to half which accounts for the the sonic improvements. 

Subjectively, I find the sound quality to be much improved with better detail and depth.  Also, in stock form this tweeter can become a little "hard" especially when driven at high levels or crossed low, but the moded tweeter, especially the Phase II stays smooth and retains its composure very well.  Paul commented that when used in the BR-1 the Phase II's sound "held together" better and was more cohesive overall than a stock unit.

While in development,  I listened to these things quite a bit.  One of the recordings that I used extensively was the DVD "Steely Dan's Two Against Nature".  As the cover so aptly states its a "Plush TV Jazz-Rock Party in Sensuous Surround Sound".  Its a "live" recording and  the stereo tracks are very well recorded.  I was struck by the wide sound stage throw by the moded drivers and,  keep in mind, I was listening to the tweeters alone.  The applause between songs was especially impressive with the image floating in the correct position and extending far beyond the width of the drivers.  The dispersion is excellent.  I can't wait to hear what they sound like in a full system.  Stay tuned for the first "Official Tweetertown System" which will feature these tweeters coming this fall to a tech board near you!

I must say I'm really satisfied with how this adventure in tweaking has turned out,  and the story is long from over.  Those of you out there that decide to give this a try are also encouraged to write your own chapter beyond what has been penned so far.  All it takes is a few dollars in parts,  patience, and a thirst for fidelity.

Need Assistance?

If you have any questions about any of the mods or anything else, feel free to e-mail me at Rusty@speakerbuilder.net or post you question to my attention on the P.E. Tech Talk board.

Acknowledgements

This adventure,  from the beginning,  has been all about collaboration,  and the advances made belong to all involved as a group.  As Monsieur L'Mayor of Tweetertown I'd like to thank the following honorary citizens:  Ken Perkins, Nelson Wood, dlr, John Serbanie, Ratch, Dan Carroll, Mike Keenan, Rudy,and Mathew Olson. 

Okay, now.....who else?......Oh yeah, I remember......and a SPECIAL thanks to Paul Holsopple and Wayne Jaeschke.

 I apologize in advance to anyone I left out and extend my acknowledgement to you as well.  Please attribute any oversights, not to willful disregard,  but instead "high mileage" graymatter.  Thanks again everyone.  Its been a blast.

Add Your Project     Other Home Audio Projects

---

DISCLAIMER - Parts Express posts the "Project Showcase" as a community service. Projects displayed have not been tested or evaluated by our staff. We make no warranties whatsoever in regard to the projects' fitness for a particular purpose. Claims of performance or fitness are those of the contributor and not of Parts Express. Project re-production for commercial resale may be a violation of United States patent or trademark laws.

Copyright 2000 Parts Express