[Steviewunda]

I have ordered the tweeters in the subject, and would like to order the capacitors I will need for a range of crossover frequencies - say 6khz, 7khz, 8khz and 9 khz

The tweeters can be seen at: https://www.parts-express.com/Aurum-Cantus-AST2560-Aero-Striction-Tweeter-276-440?quantity=1&custcol1=Aurum%20Cantus%20AST2560%20Aero%20Striction%20Tweeter&custcol_ava_item=276-440&custcol_ava_incomeaccount=General&custcol_ava_upccode=844632089312&custcol_ava_pickup=F&custcol_disableshopping=F

Can anyone recommend a good quality capacitor, and the values needed for the frequencies I would like to try?
What considerations do I need to make for the fact that the full-range (Lii F15) drivers I am pairing the tweeters with, are 8ohm and the tweeters are 4 ohm?

Thanks
Steve

[JohnR]

Impedance curve is very flat at 4 ohms so you could just use an online calculator. I'm not sure that first order is a great idea tho. Also, if the specs are to be believed the F15 is more sensitive than the tweeter which might be an issue.

[FullRangeMan]

What considerations do I need to make for the fact that the full-range (Lii F15) drivers I am pairing the tweeters with, are 8ohm and the tweeters are 4 ohm?
Usually they use a small cap to match the 4 ohms impedance tweeter.
https://www.erseaudio.com/First-Order-2-Way
By the factory freq chart the F15 perform nicely til 15kHz it makes sense add a supertweeter, in some cases it dont need a xover.

[DannyBadorine]

What considerations do I need to make for the fact that the full-range (Lii F15) drivers I am pairing the tweeters with, are 8ohm and the tweeters are 4 ohm?
Usually they use a small cap to match the 4 ohms impedance tweeter.
https://www.erseaudio.com/First-Order-2-Way
By the factory freq chart the F15 perform nicely til 15kHz it makes sense add a supertweeter, in some cases it dont need a xover.

I wouldn't worry as much about the impedance difference.  You can always use an L-pad (resistor in series and resistor in parallel with the speaker) to turn the tweeter down if it is too loud.  I would be more concerned with phase issues since the Lii F15 is a full range speaker and then you will be adding a tweeter on top of it. Since both drivers will be producing the same high frequencies but the tweeter will have a crossover filter (I recommend using at least a 2nd order filter for the tweeter) some of the frequencies will be out of phase.  For example, if you use a 2nd order filter at 6k then 6k on the tweeter will be 90 degrees out pf phase from the F15.  This generally works well if the woofer is low passed with a 2nd order filter at 6K because then the drivers are 180 degrees out of phase and switching the polarity on one of them can fix it.
What are you trying to achieve by adding the supertweeter?  It seems to me that it will cause more problems than it solves.  The F15 is smooth to 15k.  A lot of people don't hear above 15k.

[FullRangeMan]

I wouldn't worry as much about the impedance difference.  You can always use an L-pad (resistor in series and resistor in parallel with the speaker) to turn the tweeter down if it is too loud.  I would be more concerned with phase issues since the Lii F15 is a full range speaker and then you will be adding a tweeter on top of it. Since both drivers will be producing the same high frequencies but the tweeter will have a crossover filter (I recommend using at least a 2nd order filter for the tweeter) some of the frequencies will be out of phase.  For example, if you use a 2nd order filter at 6k then 6k on the tweeter will be 90 degrees out pf phase from the F15.  This generally works well if the woofer is low passed with a 2nd order filter at 6K because then the drivers are 180 degrees out of phase and switching the polarity on one of them can fix it.
What are you trying to achieve by adding the supertweeter?  It seems to me that it will cause more problems than it solves.  The F15 is smooth to 15k.  A lot of people don't hear above 15k.

What are you trying to achieve by adding the supertweeter?
IMO nothing good will happen to music adding a tweeter and electric parts, its not comfort but a super at least dont need a cap or resistor, I also feel too bad add a tweeter to a such fine FR driver as the F15.
Maybe the tweeter will firing back.

[Steviewunda]

Thanks for the input guys - i guess the most pressing question is 'what am I trying to achieve?'

I turn 70 this year, and I want to continue with my music-and-its-reproduction hobby, but I also want to reduce my financial investment in hifi gear. To that end I have sold my SS amp and replaced it with a tube amp for one third of the proceeds. These Betsy-style OB's, even with the tweeters, will cost less than one half of the proceeds from the sale of my Maggies. It remains to be seen whether the OB's will compare favourably with the Maggies, and that is the main reason for the addition of the tweeters. I already like the sound of the Betsy's, more than I expected in fact, but the Maggies had more 'air' and a bigger and more varied soundstage. From what I have read in various  media, that sense of air, and the aural cues that create the soundstage, are largely conveyed via high frequency content and harmonics. Here is a quote from a DIY Audio forum member, who used the same tweeter in Betsy OB's:
    "I can now give my take on the implementation with these together with the Betsy (ca 8 ohm) and the Eminence 12``. My amplifier is around 2 x 8 watt (homemade SIT). The Aurum Cantus I connected first with 4,7 microFarad Audun-Cap. This gives ca 9000 Hz crossover point. But later I connected in paralell 1 2,2 microFarad Audiophiler. This lowered the point to a little below 6000 Hz. And everything is beautyfully driven from my Amplifier. And the result is so smooth and revealing - with a lifelike display on good recordings. The Betsy was good alone, but this update raised the sound several degrees."
    Ok it's not apples for apples, but I'm prepared to give it a try (with your help ) I figure that the worst case scenario is having 2 baffles with holes that need to be filled somehow, and having to sell 2 tweeters (or keep them for another project)
   Here is another quote from a member of a New Zealand hifi forum that may be of interest to you (he calls it a quick response, but it's packed with info):
"Just a quick response. Although the Lii F15 driver that you are using is rated as 8 ohms (7 ohms DC resistance) it like many speakers has a rising impedance with frequency and is about 15ohms at the frequencies that you intend to crossover from. Its frequency response is flat to 7 khz and drops a few db until it falls off a cliff at about 12k. So I suggest that you run it full range and use your super-tweeter for above that frequency. The kind of amplifier that you are using is important. With a valve amp the Lii by itself might sound a bit 'toppy' and a 4ohm tweeter whilst probably not ideal might work okay. With a ss amp the Lii won't have this rising response but the 4ohm tweeter will draw more current and therefore be subjectively louder. Because it is a ribbon tweeter the impedance is a flat 4 ohms throughout the operating range. Appropriate crossover values can be obtained from an online calculator (DIYaudioandvideo) and for 6khz this is about 6uf for a first order crossover and a 4ohm tweeter. A 9khz first order crossover needs a 4.4uf capacitor. I suggest that you stick to a first order crossover for the tweeter (only) and consider raising the crossover point to 12khz which all other things being equal would require a 3.3 uf capacitor. However all things aren't equal (valve or ss) and there are some unknowns. For example whilst the Aurum Cantus has an efficiency of 95db there is no information on the Lii. However I suspect that it is also in the high 90s so the main problem is the low tweeter impedance. I suggest that you start with a 1uf capacitor on a valve amp and work up from there until you get the balance of treble that you like. If you start with a 3uf capacitor and this sounds too bright, you could consider adding a 2 or 3 ohm capacitor in series. So I suggest a bit of trial and error using 1,2 or 3 uf capacitors. Ideally the tweeter should be inset and as close to the Lii as possible. However from your photo it doesn't look like the Lii is inset and if so I wouldn't bother with the tweeter. You might like to consider putting a layer of felt around both if they are flush. Mounting the tweeter from behind is a sin."

Sorry for the length of this post! I look forward to your responses

[DannyBadorine]

OK, there's a lot to deal with here.
First, the "sense of air and wide sound stage" involves higher frequencies but more importantly has to be phase coherent.  So, just adding a super tweeter might give more high frequency content, but if it doesn't blend correctly with the Lii then you're going to create more problems than solutions.  Just adding the tweeter with one capacitor in line will cause phase shift at the frequencies the capacitor effects and will make those frequencies out of phase with the Lii (unless you're doing something to compensate for that phase shift).  So, if you're going to add a super tweeter for high frequencies then it's best to add an inductor on the Lii to reduce those frequencies there.  Online calculators can help with this, but ultimately it will probably take a lot of experimentation and you would be best to do this with measuring equipment and a program like REW to get the frequency response flat and in phase. 
So now, let's talk about "sense of air" in high frequencies.  Just adding high frequencies doesn't do this.  The ability of the speaker to settle quickly, be phase coherent and have a flat frequency response will do this.  Also, the listening environment (early reflections, extra reflective surfaces like glass, concrete, marble, etc) have a strong effect on it.  The sense of air in a speaker can be in frequencies as low as 2k, but is generally around 3k-10k.  But again, just adding these frequencies doesn't give the 3 dimensional image that we enjoy, especially if they are out of phase for part or all of them, which is what will happen if you just add a super tweeter with one capacitor. 
Next, most people at 70 years old don't hear above 10k (no offense on this but it's just the truth).  Sometimes it's even lower than that.  I'm 43 and I can still barely hear 16k, but my parents (who are your age) don't hear above 8k.  I think the "air" that you're hearing in speakers is more like 3k-8k so adding a super tweeter to get 15K frequencies might not be helpful.  Of course, if you're only using a 1st order filter at 6k (one 6.63uF capacitor in line) then that tweeter is only 10db down at 2k, and will likely be out of phase there.  So you're adding high frequency content but also adding other content that is out of phase and will create significant amounts of smearing.  A 1st order crossover gives a 45 degree phase shift at the crossover point.  This will make the imaging (depth and location of instrumentation) much worse in the 1k-4k range, which is where the majority of the spatial understanding is.
So, I really think you're going to add more problems than you'll solve with this.  If you really want to add a tweeter then use an online crossover calculator and add a 2nd order Butterworth or Linkwitz-Riley filter to both drivers (then flip the polarity) on one of them.  That way, they are more likely to be in phase with each other.  There are a lot of other considerations on this, but for now this is the way to keep it simple. This is just my input and opinion.

[Steviewunda]

Thanks for the thoughtful response Danny. I have a Deqx DSP/DAC which I use mainly for crossover and time delay correction for the Lii full-range driver and the subs. I could also use it for a crossover to the tweeter at from 48 to 300 db/octave, either linear phase, LinkwitzRiley or Butterworth, if that's any help? I'm not worried about trial and error (time to indulge in experiments like this is one perk of retirement ) and I'm not going to slit my wrists if I can't make an improvement with the tweeters. As you and others have pointed out I might be introducing more problems than improvements, and if that's the case....I'll remove them and move on. And I won't complain that I wasn't warned

[DannyBadorine]

Thanks for the thoughtful response Danny. I have a Deqx DSP/DAC which I use mainly for crossover and time delay correction for the Lii full-range driver and the subs. I could also use it for a crossover to the tweeter at from 48 to 300 db/octave, either linear phase, LinkwitzRiley or Butterworth, if that's any help? I'm not worried about trial and error (time to indulge in experiments like this is one perk of retirement ) and I'm not going to slit my wrists if I can't make an improvement with the tweeters. As you and others have pointed out I might be introducing more problems than improvements, and if that's the case....I'll remove them and move on. And I won't complain that I wasn't warned

Cool.  If you have a DSP to crossover and another amplifier then you can experiment with crossover settings that can work.  You will want to put a 20 or 30uf capacitor on the tweeter while doing this to protect it from low frequencies when the amplifier is turned on an off (ribbons can be easily damaged).  This cap shouldn't effect the crossover at all since you will be crossing it much higher than  where that cap changes things. 
Then experiment with the crossover settings from there.  Once you have what you like with the DSP, I can help you translate that into passive crossover components.  Using DSP in this way is really fun because you get to really hear the differences and it can be immediate (as opposed to changing out passive components which can take some time). 

[Steviewunda]

I'm thinking I'll get a stereo power amp to drive the tweeters, and remove the need for the passive components (apart from the 30uf turn on protection capacitor). The Deqx DSP will be the active crossover and will set gain, delay, EQ, phase, etc. Can you recommend a suitable amp that would drive the 4 ohm tweeters nicely?

[DannyBadorine]

I'm thinking I'll get a stereo power amp to drive the tweeters, and remove the need for the passive components (apart from the 30uf turn on protection capacitor). The Deqx DSP will be the active crossover and will set gain, delay, EQ, phase, etc. Can you recommend a suitable amp that would drive the 4 ohm tweeters nicely?

I think that's a great plan because it will give you an infinite number of ways to test the tweeters and get them to blend with the woofers.  As for an amplifier, you won't need much power since the tweeters are only producing high frequencies.  Pretty much all amplifiers run just fine at 4 ohms so that won't be an issue.  It all comes down to your budget.  Do you plan on eventually converting this to a passive crossover situation?  If that's the case then get an affordable amplifier like a parasound Zamp V3.  They sound good, have plenty of power and won't break the bank.  If you plan on keeping the amplifier and DSP in the chain then I would recommend getting an amplifier similar to whatever you're using to run the woofers.  You won't need as much wattage but keep the characteristics similar (class A, A/B, tube, etc.)

[Steviewunda]

Thanks Danny - I'll be keeping the DSP and amp in the chain. I can't afford an identical amp, but would something like this be suitable - it uses EL34 tubes like my Willsenton R8 https://www.china-hifi-audio.com/en/reisong-tube-amplifier-boyuurange-audio-c-48/reisong-a10-el34-tube-amp-single-end-class-a-integrated-amplifier-boyuurange-p-1300

[Steviewunda]

Forget this thanks Danny - I've ordered a little Tripath amp that will suffice to compare biamping with a single capacitor, high frequency crossover. Thanks for your help - I'll let you know how it goes

[DannyBadorine]

Sounds good. Keep me posted.

[Steviewunda]

Update time - I've tried the tweeter in several configurations - using the Deqx as active crossover and powering the tweeters with either the Tripath or Yamaha amp, at crossover frequencies of 2.5khz, 6khz, 10khz, and 12khz. I've tried with and without Deqx calibration, and with and without DSP correction. I've also tried using a resistor & capacitor crossover (@ 12400hz) - both via the Deqx, and directly through the tube amp without the Deqx. I've also tried using just the full-range driver on its own, both via the Deqx and directly through the amp.

My preferred configuration so far is to just use the full-range drivers without the tweeters, via the Deqx, using the Deqx's calibration & correction, crossover, group delay, and DSP capabilities. The tweeters detract from the coherence of the image and soundstage, and of course I was hoping for the opposite. And I'm not detecting any lack of highs, harmonics, etc without the tweeters in play.

I won't make a final decision until I hear back from my UK Deqx guru, who has my Deqx speaker measurement files to fine-tune. If he works some magic with the measurements that included the tweeters and comes up with a sound that I prefer to full-range only - well and good, I'll go with that. If not, I'll stick with the tweeterless OB's - which are sounding really good just with my inexpert Deqx corrections. And I'll have some tweeters and amps to sell

[DannyBadorine]

Update time - I've tried the tweeter in several configurations - using the Deqx as active crossover and powering the tweeters with either the Tripath or Yamaha amp, at crossover frequencies of 2.5khz, 6khz, 10khz, and 12khz. I've tried with and without Deqx calibration, and with and without DSP correction. I've also tried using a resistor & capacitor crossover (@ 12400hz) - both via the Deqx, and directly through the tube amp without the Deqx. I've also tried using just the full-range driver on its own, both via the Deqx and directly through the amp.

My preferred configuration so far is to just use the full-range drivers without the tweeters, via the Deqx, using the Deqx's calibration & correction, crossover, group delay, and DSP capabilities. The tweeters detract from the coherence of the image and soundstage, and of course I was hoping for the opposite. And I'm not detecting any lack of highs, harmonics, etc without the tweeters in play.

I won't make a final decision until I hear back from my UK Deqx guru, who has my Deqx speaker measurement files to fine-tune. If he works some magic with the measurements that included the tweeters and comes up with a sound that I prefer to full-range only - well and good, I'll go with that. If not, I'll stick with the tweeterless OB's - which are sounding really good just with my inexpert Deqx corrections. And I'll have some tweeters and amps to sell

I appreciate the details in your experimentation.  I think what you're hearing when "the tweeters detract from the coherence of the image and soundstage" is phase issues.  Frequency response is important but phase coherency is what gives us the depth and width while allowing the speakers to disappear.
Did you crossover the full range drivers when you tested with the tweeter or were they always run full range?  It might be useful to put a low pass filter on them at the same crossover point where you add the tweeter.  I recommend a 2nd order crossover (Linkwitz-riley or Butterworth) for both and then flip the polarity of the tweeter.  This might give better results.

[Steviewunda]

Hi Danny - yes I crossed-over the fullrange drivers at the same frequency as the tweeters, but I didn't flip the polarity. Something I can try...

[DannyBadorine]

In general, flipping the polarity works best when both are crossed over with 2nd order slopes.  Keep me posted!