[pstrisik]

I'm going to start a new thread with my supertweeter project as the old thread was more focused on RAAL in particular.  Hoping for general discussion with ideas, experience, and mutual support.  I'm going a DIY route and will need some expertise.  I may need to post over at the DIY forum though, we'll see.

I have Super 7 XRS with RS7A driver.  I love it, but looking to add a bit of sparkle and air.  I'm thinking of a Supertweeter like a subwoofer.  The most important aspect of the single driver is the lack of a xover in the most important range.  In my case, that would be ~90Hz - 15kHz.  This preserves most, if not almost all, of the imaging and coherence that the Omegas are known for.  In reality, the sub rolls off starting at about 90 and the supertweeter rolls off starting at about 17kHz, so there is some overlap.  But it can be best of both worlds, extension on both ends, almost like a three way, but without a xover anywhere in the mid-bass, mid-range, or lower treble range.

The subs have their own amps.  I have a 50wpc chip amp for the ST's thanks to one of our community here.  I am finding, however, that the ST's I have sound better when wired parallel with the Omegas rather than bi-amped.  This may be due to the much greater gain of the chip amp when compared to my tube amp.  I will be experimenting further with this.

I picked up a pair of used ST's made with Vifa dual concentric ring tweeters.  They have selectable high pass filter for 15, 17.5 or 20kHz and level selector of 0, -3 or -6db. 

I have them set for 17.5kHz hi pass and -6db.


 

Bottom line is that they do what I had hoped, though I would like more control and refinement.  So, I will post my planned project in a reply to this message later today.

There is that bit of sparkle with symbols, etc.  And the quality I call "texture," evident in strings, breath, etc., is enhanced.  I anything, I would say imaging is improved.  Definitely not degraded.  I found this with subs as well, even though they (two) run in mono.  Don't ask me to explain how that works! 

More to come..........   

preview:  I have on the way a pair of vintage Yamaha beryllium dome tweeters and a pair of vintage Pioneer (TAD) ribbon tweeters that will be my DIY starting points for a higher quality ST (hopefully!   ).  The plan is to make separate boxes for the tweeters and for the hi pass filter/L-pad level control.  If things go well, it would be less risk to go for some current high end tweeters.  I will just need to swap out the tweeter in that box and possibly change a cap and/or inductor in the filter.  So.....  my questions will come later today.

I love this hobby!   

.........Peter

[pstrisik]

So, here are the issues as I see them.  I'm far from a speaker design expert by the way.

1) If I bi-amp, I can avoid issues of impedance.  Wiring the ST's parallel I think will affect the impedance the amp sees. 

2) The hi pass point is the simple part.  It is the roll off slope that looks like it will complicate things.

3) I don't think I have to worry much about sensitivity if I incorporate LPads.  Just have to make sure they are sensitive enough.  I don't think that will be a problem given my experience thus far with the Vifas.  I have them at -6db - they are rated at 92db.

Ok,

Impedance.  Running parallel with my existing Super 7 XRS isn't an issue.  The Vifas are 6 ohm, the Omegas are 8 ohm.  I will measure the Yamaha and Pioneer tweeters when they come, but I don't think either is lower than 6 ohm.  Using:  Total impedance is the impedance of the lowest impedance driver divided by the total number of drivers.  Parallel with 6 ohm gives me 6 / 2 = 3+ ohms.  No problem.  Does this formula apply for tweeters?  They are such little things!     

But, I may move to Omega 1.5's at some point.  I think Louis is running the two woofers parallel as he rates them at 4 ohms.  If I consider the Omega 1.5 one 4 ohm speaker, it is 4 / 2 = 2+ ohms.  If I consider it to be two 8 ohm speakers, that means impedance would be 6 / 3 = 2+ ohms.  Same either way.  Anyone know enough to say whether this is all correct or not?  I believe my amp (Dennis Had Inspire) to be capable of driving at 2 ohms.  I will double check with him.  Maybe one or both of the tweeters coming are greater than 6 ohms which would help unless the calculation for the 1.5's has to be as one 4 ohm speaker (it would stay the same) rather than two 8 ohm drivers (would allow overall impedance to increase).

Slope.  What I am seeing about filter design suggests that I can get away with a single capacitor, but that gives me a 6db/octave slope.  I would like it to be steeper so there isn't too much interaction with the Omega's strong range.  It looks like this means adding an inductor and possibly a second capacitor.  More complicated, particularly since I will likely experiment with different hi pass points and maybe different slopes.  I don't want to have to buy a ton of caps and inductors to do this.

But... the Vifa's use a 6db slope, and the integration seems good, so maybe that will work ok.

Anyway, that's where I'm at.  A fair amount of thinking out loud at this point.  Comments welcome.

Here are photos of the coming tweeters:

On the left (top) is the Pioneer (TAD) Beryllium Ribbon Tweeters PT-R14A from S-910 speaker.
On the right (bottom) is the YAMAHA Beryllium Dome Tweeter JA-0522 from NS-590 speaker.

.

 

.

[yyz67]

Interesting project... sounds like you might ultimately end up with a partial 3-way or a .5+1+.5 way... that is, the Omega wideband driver running in it's full range but low pass on sub/woofer and high pass on ST, correct?

I'm no expert either, but I think combining the "nominal" or average impedance of each driver using any formula is not necessarily correct. As I understand it impedance is a function of frequency, so it might dip in the overlap region between two drivers (which I think also depends on relative phase), but in the region where each driver is operating independently, it will be dominated by that driver's impedance -- of course ignoring cabinet and other interaction effects.  I think the way to really know the impedance is to measure the speaker-crossover system as a function of frequency. Then there will be no doubt as to its real value and possible problem areas.  Also, I believe the impedance in the low frequencies is generally the problem area for amps having to drive more current, but I could be mistaken.

Anyone, please feel free to correct this if wrong.

[pstrisik]

Thanks for the reply yyz67.  You are correct; I'm looking at the Omega running full range as it was designed.  Nothing added to its signal path.  And adding Sub with lo pass and ST with hi pass.  I don't know if anyone has figured a naming convention for this.  Maybe 1.5.5 way?

I don't know if I will get to the kind of measurement you are talking about.  I probably should at least pick up a proper mic to use with REW and figure that out.  I will keep an eye out for more about interactions that affect impedance.  It may be that the addition of a tweeter in this scenario is more benign than it would be adding another full range or a woofer.  Maybe the tweeter doesn't have or cause significant impedance dips below nominal.  More info to gather.

......Peter

[DaveC113]

Looks like fun, I think you'll have to decide if you want to proceed using trial and error or whether you want to model the system on a computer and then measure the results.

Trial and error could be a good way to go using something like a Behringer digital eq/xo unit, once you find a good spot you can implement it passively.

If you measure the speaker response you'll need a calibrated mike, DAC/ADC with mic pre/phantom power, a compatible sound card, cables and you can use freeware like REW or HolmImpulse. Or you can get the Dayton system that uses a USB mic. It's a little pricier but plug-and-play. You can also use the measurement system for room eq or to see what effect changes in your system or room are having.

[yyz67]

To avoid hassles of phantom power etc, I recommend a USB mic. I got the UMIK-1 with calibration files from cross spectrum though they also have a USB and non-USB Dayton model.

I've only done some basic FR and distortion measurements using REW. Seemed to work well. I'm haven't tried measuring impedance as it seems to require wiring a sense resistor in the speaker circuit - not hard but I haven't tried it:

   www.roomeqwizard.com/wizardhelpv5/help_en-GB/html/impedancemeasurement.html

[pstrisik]

I don't think I'll be doing the impedance measuring.  Only if it comes down to necessary.  I'm hoping to gather some info about how the impedance curve would be affected by adding a tweeter in this way - and hoping it is not much of a concern.

As far as freq measurements.  I do have a TASCAM US-366 USB audio interface.  It is for ripping vinyl to the computer.  But it has XLR mic input and USB out.  I am assuming I can use this with a non usb mic.  I have a laptop with nothing fancy for sound.  Any advantage of going with this vs. a USB mic other than cost?

BTW, I received the Yamaha beryllium tweeters.  Very nicely made - heavy.  Nice mesh screen and dome very well protected.  I measured nominal impedance and got 6.3/6.6 ohm. 

[yyz67]

Any advantage of going with this vs. a USB mic other than cost?

Not that I know of... just recommending a USB mic if it made sense for you.

Have fun and let us know how it goes.

[pstrisik]

Not that I know of... just recommending a USB mic if it made sense for you.

Have fun and let us know how it goes.

I sure will, though this is a longer term project.  Still waiting for the Pioneer ribbons to come from the Ukraine and I ordered some connectors (binding posts, spades, etc.) from China to keep this project affordable and that will be at least a couple of weeks.  But stay tuned!

........Peter

[pstrisik]

I received information privately from someone here about impedance.  Much simpler than we thought.  Easy to just measure nominal impedance with a multimeter directly.  Without powering the speakers, take a reading on the main speaker without the supertweeter and then one with it connected.   

I'll try that tonight with the Vifas.  I've bought a few components for making an enclosure for the tweeter while waiting for things to come in the mail.  Will work on that this weekend.

.......Peter

[pstrisik]

Back again with some progress!

I built a pair of supertweeters from the Yamaha beryllium drivers.  Amateur and odd looking job, but functional.   If I love these long term, I may try to make more professional looking enclosures. 

Simple design:  Yamaha JA-0522 beryllium tweeters from the 1980's, the enclosure you see, negative wired to binding post, positive with 1.03 mf capacitors in line to binding post.  Voila!  This gives me a high pass at over 23kHz, first order which means 6db/octave roll off.  I have a 50wpc chip amp and Schiit SYS volume control so the tweeters are powered separately from the Omegas.

Louis cautioned me that any supertweeter, even when the high pass is so high, can interfere with the point source benefits of the single drivers.  He said the way around this is to put them behind the main speakers pointed up.  So I will be seeing about this aspect as well.

I'm using a variety of cheap electrolytic caps so I can determine the best values/best high pass frequency.  Once I determine this, I will order better quality caps.  The internal positive wire has alligator clips to hold the cap, so they are easy to change out.

These Yamaha tweeters are of impressive build quality.  Very heavy magnets and metal face.  The beryllium dome is completely protected.  It makes these little cubes heavy and stable.  Speaker cable with bananas don't move them at all.

23kHz sounds high, but the roll off is relatively gradual, so it is audible.  I just want to add a bit of sparkle and air, not augment the 10-15kHz range, so this may work out.  I'll do some listening today and see how they sound.  They are playing now, but have to wait until I have the house to myself in an hour or so to really give them a go.

Havin' fun!


      

[opnly bafld]

You are also going to try them facing the "wrong" way right?
aka rear firing.

[DaveC113]

I agree rear firing is a good way to go but they need a few feet between the tweeter and back wall, another way to go would be up-firing, either from the top of the speaker or behind the speaker with the tweeter on the floor. Having the tweeter on the floor would increase distance from tweeter to ceiling, which may be better than on top of the speaker.

[pstrisik]

Sure......  from my post above:

Louis cautioned me that any supertweeter, even when the high pass is so high, can interfere with the point source benefits of the single drivers.  He said the way around this is to put them behind the main speakers pointed up.  So I will be seeing about this aspect as well.

[pstrisik]

Initial results are promising.  Nothing near conclusive at this early stage tho'.  Behind facing up seems like the best approach so far.  I have the volume all the way up though, which tells me I need to high pass a bit lower.  That chip amp has much more gain than my single ended 5 watt tuber.  It would be nice to use a single cap.  There are 1.2mf caps around.  Not 1.1 though; I was going to have to use a 1.0 and a 0.1 in parallel.  1.2 would give me 20,400kHz instead of the 22,300kHz with 1.1mf.  Saves some $ too.