[chris porada]

A year or so ago I posted about adding a DIY supertweeter to my ACI Alpha limited DIY speakers, and was very happy at the addition.  However, after extended listeing, it became apparent that the intial crossover was way too low. After a couple of weeks, things went from sounding really exciting to sounding somewhat fatiguing. I removed the 1.1uF caps and replaced them with two 0.22uF North Creek caps and a 0.01uF Infinicap bypass (0.45uF total), which, according to the crossover webpage I found, should have me at -3db at somewhere around 39kHz, but with such a gradual slope (6db per octave i think) that there is still some very high frequency audible information coming through, i.e., if I listen to test tones, I can hear some of the highest tones up near 20kHz coming from the supertweeter. The sound now is much better, and not at all fatiguing, just a hint more "sparkle" and much better resolution of microdynamics and very subtle details in the music. The SACD of Pink Floyd's DSOTM is unbelievable now on the tracks that involve cymbols/high hats. It sounds like the drum set is right in the room. I still haven't figured out how to get the precise time-alignment, though, especially given the short wavelength we're dealing with at this high frequency range. I don't have any type of fancy test equipment. I've just tried to align the super tweeters with the tweeter on the main speakers by eye, and angled them down to my listening position using a laser pointer. Any suggestions on how to get time-alignment without fancy equipment would be great if that's possible.  I'm just curious if there's some way to sqeeze even a  bit more from this already rewarding DIY project.
Thanks for all the help,
Chris

[SET Man]

Hey!

    Yup, adding super tweeter really do make the system sound better.... at least for my system is help.

   As for the xover cap. I went through the same thing as you did when I first added Fostex FT-17H to my 6" Fostex FE167E in TQWT. I did calculation with the online site.... maybe the same that you used and it show that I should use 1uf also the Fostex paper recommended .68uf-.33uf. I went with 1uf Solen and like you said the was too much. Now I'm using .33uf.... but sometime I switch to .47uf. I'm still torn between the 2

   After trying the 1uf and it didn't work make me realized that I can't only use the ohm to calculate the xover but I also have to take the sensitivity of the super tweeter and the main driver in account.  My main driver is rated 95dB and the tweeter rated at 98dB. I did look at the respond graph of the FT-17H to see how the roll off with xover would be and the higher cap work just fine. The FT-17H is rated up to 50khz

Anyway, as for the sound. Yes, definately a big improvement. Eventhough the super tweeters don't call attention  to themsleve but the improvement from top to bottom is undeniable I love it when I lisnten to LPs and SACD. But I also notice some improvment on CD but not much just some more top end. Well, you know that CD cut off at max 22khz 

I don't know about the timealigment thing but I do set my supter tweeter about .75" back from the front of the main driver. I guess you could run the tweeter at lower range with the main using pink noise for FM hiss than start move it back until you here one voice without the tweeter calling attention to itself. Might work

Here is link to my Fostex FT-17H that I'm using for supertweeter duty. Not bad for $36 ea.

http://www.madisound.com/cgi-bin/index.cgi?cart_id=200162.11440&pid=342

If you play LPs and SACD I highly recomemd trying a supertweeter even cheap one like the Fostex I have

Take care,
Buddy

[chris porada]

Hi Buddy,
Thanks for the information.  Actually, that's the same Fostex I'm using, and I love it.  I also notice an improvement with regular CD's, which I think is curious, especially with such a high crossover from the .45uF caps.  How do you equalize the efficiency of the super tweeter to the main speakers?  My alphas are around 93db and the Fostex is at 98db.  I'm using the Fostex R80B 100W L-pad set at -6db, and that seems to blend pretty well with my main speakers, but I don't know enough about electronics to know whether setting the efficiency to -5db on the L-pad will really make the 98db fostex equal in output to my 93db alphas, especially since the alphas are a 4ohm speaker and the fostex is 8ohm.  Could you please help me to figure this out?
Thanks for all your help
Chris

[SET Man]

Hey!

   No, problem chris. Funny isn't it that we both are using the same tweeter and end up using similar cap value for xover

    The improvement is undeniable as I said before. As for the improvement with CD. I think I have an explanation on this.... Most speaker rated up to 20kHz+ are not totally flat out to the top. Example speaker that say 40Hz-20kHz +/- 3 are likely to have a 20kHz at -3 dB point or lower if listen off axis. When we added the super tweeter even thou xover much higher than 20kHz point we still get some boot at the 20kHz and lower. So, this will fill and lift that sound around that point a bit. And this I think is why you and I find that there is some improvement with CD even though CD cut off point is 22kHz or was is 21.5kHz?

Okay now the balancing act between the super tweeter and your main speaker. I actually run my tweeter wide open without any L-pad. This seem to match my main Fostex 167E at 95dB well. Yes, the FT-17H is 3.5dB higher than my main. But if you look at the respond graph on the FT17H you will see that the over all the respond is flat around 98 up to about 30KHz than it start to roll off to around 96dB. So, if I xover the tweeter at 30kHz+ of which they are... the SPL is matched up well. And also we have to take account of the off axis reason too of which will be lower in dB at the 30kHz point. Well, I'm pretty lucky that I don't have to pad it down... cleaner sound without L-pad

I see that your speaker is rated at 93dB and 4 ohm. The FT-17H is rated at 98.5dB at 8ohm. 5.5dB higher than your main speaker and seem like it. But if my memory served me right to get that sensitivity number with 4 ohm speaker is that you will need 2 watts at one meter. So, I guess you could call your speaker 90dB in this case. According to this you will need to pad down your tweeter 8dB. I'm pretty sure that if run your tweeter wide open  it will be louder than your main.

Since you already have L-pads for them. The number on L-pad are not accurate. What you could do to balance them with your main is to use test tone CD and a SPL meter.... I use Radio Shack analog one. Use let's say 8kHz test tone.... you will need to take out the xover you have now. Set your SPL meter on a tripod with meter leveled and point on axis with the tweeter and 1 meter away from the tweeter. One channel at a time. Turn on the test tone and set meter to 80dB or 90dB.  Than turn the L-pan down.... I would start with 5dB down first with your speaker.... until the meter read -5dB from 0 middle point. Do, the same on the other side. Listen and see if is the sound is too dull or too bright. If is too bright than do it again but with 2dB down more than before. But this is very dependent on your taste, front end, music type, room and on axis or off axis listening position. Oh! don't forget to use earplug when you do SPL testing and don't play the tweeter with low fequency too loud it might burn out.

I used this method when I first use the FT17H with 1uf until I realized that I could use .47uf or .33uf for higher xover point wide open! Well, I just listen to my system and I had them connected with .47uf and I think I will keep it that way because overall is very well balanced.

Well, I hope this help. Don't go crazy over it with the measurement and the mathematical calculation. You know your best measurement tool you have is still your own ears

Take care,
Buddy

[cporada]

Hi Buddy
Thanks for all your helpful information.  I'm still curious if you, or anyone else on this forum, may have an idea how to get the super tweeter time-aligned, though.  From what I've read (not a great deal, but some at least), people suggest that at these short wavelengths it is easily possible to be "in phase" but out of time alignment, such that I could have the super tweeter in phase, but at a distance of several wavelengths from the main drivers.  Someone on another forum suggested that using pulse measurement would be the best way to time align, but I don;t know exactly what this means or how one does this.  Can anyone here help me with this?
Thanks again
Chris

[SET Man]

Hi Buddy
Thanks for all your helpful information.  I'm still curious if you, or anyone else on this forum, may have an idea how to get the super tweeter time-aligned, though.  From what I've read (not a great deal, but some at least), people suggest that at these short wavelengths it is easily possible to be "in phase" but out of time alignment, such that I could have the super tweeter in phase, but at a distance of several wavelengths from the main drivers.  Someone on another forum suggested that using pulse measurement would be the best way to time align, but I don;t know exactly what this means or how one does this.  Can anyone here help me with this?
Thanks again
Chris

Hey!

   As for the time alignment.... using pulse measurement would be ideal but I have no idea what equipment we need for this. I subscript to Stereophile and they do use this pulse measurement to test if the speaker are time aligned I think.... but I can't remember the last time I see a pair of speaker that they say are perfect time aligned.

   Well, like I said before... you could use pink noise or FM hiss, with lower xover on the supertweeter then move your supertweeters a few millimeter at a time and see if the image locked in and the supertweeter total disappeared at you listening seat. Ain't easy.

  But... Chris I don't go crazy over this. Remember that your main speaker probably not time aligned themselves.... this is worst at lower frequency and they still sound good  and you still like them right? 

   Well, have fun. But don't forget to sit back and just enjoy the music

Take care,
Buddy

[cporada]

You're right of course, enjoying the music is what this is all about after all!

Perhaps Mike D can help out here, but I'm pretty sure my ACI Alpha's are time-aligned at least to some degree, since the tweeter is set back from the midranges.  That's why I'm kind of keen on trying to ensure I'm not wrecking the imaging by having the super tweeters out of time-alignment with the rest of the array of drivers.  Am I right Mike?   Is there an easy way to do this?
Thanks again Buddy and anyone else who may be willing to help.  The quest for audio Nirvana is never ending!
Chris

[bbchem]

Hi, I have a pair of 89 sensitivity Omega single point source, can I add the fostex ft-17h directly to the binding posts? sorry I am a complete novice

bill

[cporada]

Hi Bill
I'm also no expert.  I relied heavily on the advice I received from others on this group and from trial and error when I set up my supertweeters.  My understanding is that you have to have some kind of crossover in the form of a capacitor, otherwise, you're going to be asking the tweeters to reproduce all the frequencies of sound coming to your main speakers, and this is not something they will like.  First, I would guess you will have a lot of distortion from trying to get the fostex to reproduce frequencies that are too low for them, and secondly, I think your system will sound really nright and harsh to to reinforcement of a wide range of frquencies that your main speakers are already producing.  Also, if your main speakers are at 89db, I would expect you need some kind of L-pad or resistor to drop the output of the fostex, since at 98db, they're a lot more sensitive, and hence, a lot louder per watt of input, tha your main speakers.  So again, you would have way too much treble.  Maybe someone on this forum with experience in crossover design can help you with the specifics of what you need.  I based my guesses on the calculator I found at:
http://www.carstereo.com/help/Articles.cfm?id=1
I hope this helps
Chris