Tweeter Recommendation for Woofer

[mike7877]

Peerless D27TG-35-06

How did I miss your post?!

That tweeter looks excellent! It does need a notch filter (like I said I was trying to avoid if at all possible), but other than that, it's perfect! (honestly they all need a notch filter so this doesn't matter)

Because of the way it performs and the way I like to use drivers, it does need to be crossed at 2kHz though.. This isn't a problem for me - the midbass being 85cm2 and moving 4.4mm peak to peak (it moves further, but that's linear), this is only going to be a ~25 watt speaker, so the tweeter shouldn't be overpowered

The tweeter just needs a first order cap at 2000Hz, and the woofer just needs a cap and coil!

The tweeter's QMS is an amazing 0.9 (lowest I've ever seen!), essentially making it so almost no matter what electronics are before it, Qts will not vary from 0.5 +/- 0.1

It's a lower point than the other ones (not that I couldn't cross it at 4kHz too), but it'd be a shame not to use it in the 2-4kHz range instead of the woofer (which will have doppler from the bass and potentially other IMD too). I save on parts! The wins keep stacking up...
Xmax is lower (0.9nmm p-p), but that should be more than enough to match the RS150P-4A's output when crossed at 2kHz second order (no recommended crossover point is given, but a 100W "system power" is implied, so even if that's 2nd order at 3kHz (unreasonably high), 2nd order 2kHz would be 20W

 

[gfinlays]

Oh, I haven't settled on anything yet - I was moving toward using Equalizer APO, but I'm open to anything. Suggestions?

I'm not familiar with Equalizer APO - from what I've read of it's capabilities, it seems to be a fairly basic parametric equalizer?

If your ultimate aim is to go active with some ATCs, then you'll want something more sophisticated.

I use Acourate, but it's a deep rabbit hole. It generates linear phase filters, provides full driver amplitude and phase linearisation, zero delay plane time alignment and then adds room correction on top.

Compensating for the effect of a tweeter protection capacitor is trivial. You do the driver linearisation measurements with the cap in place and Acourate automatically compensates for the amplitude and phase effects when it convolves the inversion into your target crossover.

You've not heard what a speaker can really do until you've heard linear phase filters.

 

[LSPhil]

I'm not familiar with Equalizer APO - from what I've read of it's capabilities, it seems to be a fairly basic parametric equalizer?

If your ultimate aim is to go active with some ATCs, then you'll want something more sophisticated.

I use Acourate, but it's a deep rabbit hole. It generates linear phase filters, provides full driver amplitude and phase linearisation, zero delay plane time alignment and then adds room correction on top.

Compensating for the effect of a tweeter protection capacitor is trivial. You do the driver linearisation measurements with the cap in place and Acourate automatically compensates for the amplitude and phase effects when it convolves the inversion into your target crossover.

You've not heard what a speaker can really do until you've heard linear phase filters.

In another thread you mentioned that you are activating SEAS speakers. I would like to ask about the progress you have made on this. I am also a fan of steep slope filtering, but what your program achieves really amazes me. Already at 1500 Hz it attenuates by -24 dB/oct and then even faster. That is why phase correction is so important, because otherwise it would not work well.
I recently used a 48 dB/oct Bessel filter and Linkwitz transformation, but unfortunately I am still a fan of the 7th order Butterworth filter. I can safely say that I understand how well designed crossovers work. Try to get a better Chassis, for example the SEAS U16 series and their wonderful but expensive SEAS Excel T29CF002 tweeter. The front panel is similar to the Peerless 27, but the drive is achieved by a magnetic gap saturated to the maximum. The Peerless uses a copper ring for this purpose, which gives a similar effect.

 

[MAB]

Very true - I'll be using a cap, but since it's going to be in series with a SOTA tweeter driven by Topping LA90D and a DAC more transparent than even the amp, I really have to choose the right one. Well, I have to be sure the one I choose, I'm not going to learn 32 days later it's fundamentally flawed for the purpose lol.

I've got a 10 or 20 uf cap I can use in the mean time - I bought it to use in a crossover that I never made. It was $20 and metalized polypropylene

You got some really good advice in another thread you opened on the protection cap, and caps in general.

Best DC blocking, AC transparent capacitor for tweeter protection (not crossover)

The capacitor is especially important because it will be the only thing between the tweeter and amplifier. I'm in the final stages of acquiring everything necessary for my passive to active speaker conversion - exciting! Before recommendations, you need to know the capacitor's performance is...

www.audiosciencereview.com

Are the criteria still unclear? This protection cap is about the least critical aspect of your project, except that you should use one. I'm glad it was pointed out the need for you to implement a blocking cap. Now is the time to just get a capacitor or two and get over the FUD and FOMO that you will somehow get the wrong capacitor. This is how people get all twisted up and go and buy exactly the wrong item for their purprose. You have some good advice from members, just take it.

I think it was also pointed out you can use a cap as part of the final crossover filter, a really nice way to go. Either way, I don't see much risk in the choice. The type of tweeter, SOTA or not, isn't relevant. Implementation matters.

Let's revisit some of that capacitor advice that you got previously, with measurements. You are worried to the point of indecision about the qualities or specs of various capacitors. It was advised that ordering products from companies like Panasonic and WIMA offer the best performance, quality, and reliability, and at reasonable prices. I have a collection of capacitors, all are appropriate for a speaker crossover. Here they are pictured on a book I also think you should read as a start.

Top to bottom:

1. WIMA MKP 8.0 uF
2. Solen 5.0 uF
3. Dayton 4.7 uF
4. Crescendo 6.0 uF
5. Auricap 8.2 uF
6. AEON 8.2 uF
7. Hovland 9.0 uF
8. Panasonic ECWFG 8.0 uF

The Hovland is the most expensive. It has an interesting look, somewhat lumpy, as if rolled by a bunch of slackers in a garage using one of their uncle's joint-twisting machines. Let's see if that roll-your-own ethos results in a better capacitor. As far as looks go, I like the WIMA and Panasonic with their precise potting. The Panasonic is useable for automotive purposes with a flame-retardant case, shock and heat resistant potting, corrosion resistance, high voltage rating. Both the WIMA and Panasonic come with spec sheets that include lots of data. Here is the Panasonic ECWFG datasheet.

https://industrial.panasonic.com/cdbs/www-data/pdf/RDL0000/RDL0000C245.pdf

Panasonic provides graphs of all the parameters you were asking about in all of those threads where you were wondering about the very-best component to use. For instance, capacitance change vs. frequency for the exact 8.0 uF part I have.

I have an LCR meter with a test fixture to measure this, let's see if Panasonic's capacitor meets the spec!

It does indeed meet the published spec across frequency, no surprises. The WIMA does too. How about the other capacitors? The companies that make the other capacitors don't bother to publish meaningful data as already explained to you in those other threads. But let's see how they actually compare to performance of the Panasonic and WIMA caps:

Good grief, the Hovland is the worst by far for capacitance change vs. frequency, the Panasonic and WIMA are both best at the very bottom of the graph. Let's zoom in:

The Panasonic and WIMA are an order of magnitude better for cap vs. freq, might be audible if we had hearing well above 10kHz, but we don't so likely inaudible differences. But better performance as measured by a fairly sensitive LCR meter. Any of these will be fine with your tweeters, both your fancy tweeters and the cheap ones you are debating buying. This does illustrate how much better Panasonic's and WIMA's capacitor design and manufacturing process are compared to the woo woo caps. Not to mention Panasonic sells for $7, WIMA for $4, the other caps range from similar prices to the Hovland which sold for $70 if I recall what the list price was 20 years ago. The Hovland adheres to the typical HiFi price correlation: price increases as the inverse of performance.

Dissipation Factor can also be compared:

The Panasonic is best of all of the 8 to 9 uF caps, the WIMA is best of all of the ~5 uF caps. The Hovland is solidifying it's status as worst. All of these are just fine for audio. So my use of best and worst is a bit dramatic. In fact, these words are problematic to apply here.

ESR can be measured, the test fixture I am using isn't ideal and I'm too lazy to swap test modules to measure the capacitor's intrinsic ESR, but this is a good measurement of how it will perform in an audio crossover with parasitic layout effects. As such, these measurements will differ from the specified values.

All of these capacitors have low ESR, and will have negligible differences in a passive crossover spanning the entire audio spectrum, even for the few applications where ESR is important.

Some might say that's picky, but especially with a passive crossover I just wouldn't want to use it unless >4kHz with 3rd order (and I'd still feel that was a compromise)

I think you are being counterproductively picky. And the above statement is not correct. With a 3rd or 4th order you can use that Dayton tweeter well below 2kHz no problem, 1400Hz as @ktaco said to you. Your criteria for tweeter crossover and 'high fidelity' are not making sense. Lots of other good tweeters beside the Dayton also recommended. Just the specs alone won't tell you which one is best.

You should get just about any of the tweeters mentioned. Measure the TS parameters for the drivers. Get a cheap set of blocking caps from a company that specializes in making good capacitors. Figure out the baffle design for your inexpensive woofer and tweeter. Figure out the woofer alignment. Build the boxes. Confirm the result in-box aligns to your design. Measure on and off axis performance of the drivers plus boxes, close field, etc. Experiment with filters. Measure on and off-axis each experiment. Find the best frequency and filter types to merge tweeter and woofer for the arrangement you have constructed. At this point consider if you want blocking cap on tweeter or if you want to integrate the cap into the actual crossover filter as some do. At that point, cautiously move on to your expensive kit.

edit: typos

 

[gfinlays]

In another thread you mentioned that you are activating SEAS speakers. I would like to ask about the progress you have made on this. I am also a fan of steep slope filtering, but what your program achieves really amazes me. Already at 1500 Hz it attenuates by -24 dB/oct and then even faster. That is why phase correction is so important, because otherwise it would not work well.
I recently used a 48 dB/oct Bessel filter and Linkwitz transformation, but unfortunately I am still a fan of the 7th order Butterworth filter. I can safely say that I understand how well designed crossovers work. Try to get a better Chassis, for example the SEAS U16 series and their wonderful but expensive SEAS Excel T29CF002 tweeter. The front panel is similar to the Peerless 27, but the drive is achieved by a magnetic gap saturated to the maximum. The Peerless uses a copper ring for this purpose, which gives a similar effect.

Yes, with some help, I've finally reached the stage where I have a fully working active crossover setup. Linearised drivers (magnitude and phase), linear phase filters and perfect time alignment. The sound is like nothing I've ever heard before. Soundstage is immense and everything is precisely placed within it. It presents sound in a way that no passive loudspeaker ever could.

 

[LSPhil]

Yes, with some help, I've finally reached the stage where I have a fully working active crossover setup. Linearised drivers (magnitude and phase), linear phase filters and perfect time alignment. The sound is like nothing I've ever heard before. Soundstage is immense and everything is precisely placed within it. It presents sound in a way that no passive loudspeaker ever could.

Thank you for the information. As you can see, using filters with a steep slope ensures that the woofer does not interfere with its unsuccessful attempts, as is the case with a 500Hz crossover with an LR2 filter at 1kHz, which only attenuates by 12dB, equivalent to only half the volume. This disrupts the performance of the midrange driver. In your case, it is completely disabled. One looks in awe at the figure of 300dB, but it is purely theoretical, as there is no device with a dynamic range exceeding 120dB. I use filters of 42 dB/oct, which is only 4 times quieter, but better than LR2, which is noticeable in the imaging. I am curious about the delays between the woofer and the midrange, and between the midrange and the tweeter. Another matter that interests me is the crossover point, whether it is at -6dB or -3dB.

 

[gfinlays]

One looks in awe at the figure of 300dB, but it is purely theoretical, as there is no device with a dynamic range exceeding 120dB.

This is true. The slopes are still incredibly steep beyond the crossover point:

As you can see, crossover points are at -6 dB.

Time delays between drivers were actually quite small. Their positioning on the baffle and a 5° backtilt helps with physical alignment, although ZDP alignment rarely coincides with physical alignment. I was only 6 samples at 48 kHz between bass and tweeter and 2 samples between mid and tweeter. Acourate uses the tweeter as the reference. In my case, time alignment in Acourate was probably completely superfluous, but it's nice that Uli offers it as a feature and it really comes into its own with subwoofers.

 

[LSPhil]

This is true. The slopes are still incredibly steep beyond the crossover point:

View attachment 427076

As you can see, crossover points are at -6 dB.

Time delays between drivers were actually quite small. Their positioning on the baffle and a 5° backtilt helps with physical alignment, although ZDP alignment rarely coincides with physical alignment. I was only 6 samples at 48 kHz between bass and tweeter and 2 samples between mid and tweeter. Acourate uses the tweeter as the reference. In my case, time alignment in Acourate was probably completely superfluous, but it's nice that Uli offers it as a feature and it really comes into its own with subwoofers.

Through our experience with filters with a large slope (>30 dB/Oct), we are setting new standards that refute the general claim that LR2 filters are acoustically superior. This is definitely not the case. I was already surprised by the acoustic performance of LR6 more than five years ago. I only abandoned them because they lack -3 dB in power at the crossover point. Voltage-wise, they are acceptable. That is why I turned to odd-order Butterworth filters. Troels Gravesen promotes this acoustically weak filtering. I would like to inquire whether your tweeter still lacks a waveguide. Finally, I would like to persuade you to use woofers with copper or aluminium rings. Scan-Speak offers a very successful series with NRSC Fiberglass Cones, such as the 15M/4624G00 and the 22W/4534G00. I sold the 15M because it was too large and I use Audax HM130Z12, and now I think I will return to Scan-Speak.

 

[gfinlays]

Through our experience with filters with a large slope (>30 dB/Oct), we are setting new standards that refute the general claim that LR2 filters are acoustically superior. This is definitely not the case. I was already surprised by the acoustic performance of LR6 more than five years ago. I only abandoned them because they lack -3 dB in power at the crossover point. Voltage-wise, they are acceptable. That is why I turned to odd-order Butterworth filters. Troels Gravesen promotes this acoustically weak filtering. I would like to inquire whether your tweeter still lacks a waveguide. Finally, I would like to persuade you to use woofers with copper or aluminium rings. Scan-Speak offers a very successful series with NRSC Fiberglass Cones, such as the 15M/4624G00 and the 22W/4534G00. I sold the 15M because it was too large and I use Audax HM130Z12, and now I think I will return to Scan-Speak.

I'm not entirely sure where the notion came from that low order crossovers are best, but it seems to have urban legend status now.

My tweeter does stil lack a waveguide and I'm still pondering whether to fit one - it's a one-way ticket with my carbon fibre skinned cabinets, though I suspect it's probably not a bad idea. It would better match the midrange directivity at the crossover frequency and reduce dispersion.

Thhe Visaton WG148R waveguide is a very god fit on the 27TFFC and my initial measurements are encouraging. I also have a pair of Visaton G25FFL tweeters which the WG148R was designed to fit. I'll fit them together and measure them on my test baffle this week. There are lots of driver options to play with in 22 and 15 cm sizes. I have also considered the W22EX001 and the W15CY001 from Seas, though the prices are not inconsiderable!

 

[LSPhil]

I had the W12 as a midrange with a magnesium membrane, and it was too gentle for me, although very musical. If you like the British style, you'll be pleasantly surprised. A few days ago, out of curiosity, I performed a filtering of the low-frequency with the midrange using Butterworth 78dB/Oct. I must say, my low-frequency driver caused a lot of chaos in the music. Now, just like the two of us and Mr. Gauder in the Isophon Berlina, we use filters (Isophon 60dB/Oct) with very good results, showing what a good crossover can achieve. It's like wiping your glasses and seeing a beautiful picture anew.

 

[Wolf11Man]

To the thread starter, I think there is something wrong with how you value your "expensive ATC drivers" as something special. The ATC SCM20 Pro PSL Mk2 may have a ridiculous high retail price, but the value of the OEM drivers they use is only a tiny fraction of this. You will find similar quality for less than 300$ (a woofer and a tweeter), of course a speaker manufacturer doesn't buy at the inflated DIYS prices you pay. Not even half of that.
If you expect to get something better from the ATC drivers than from a Dayton and SB pair as mentioned, you may be in for a surprise.

This doesn't take into consideration personal skills, knowledge, experience, instruments and software. Not to speak about learning what of ones knowledge is plain nonsense. That can take quite sometime. Also, one doesn't learn in "a week or so" how to do a hiend active speaker.