Tweeter Recommendation for Woofer

[mike7877]

So I have a project - transforming my (pretty expensive) 2 ways from passives to actives (ATC SCM20 Pro PSL Mk2), and I had the (probably wise) thought: I should probably practice on cheaper drivers... until, at least, I get very familiar with amp behaviours (like at power on) and DAC behaviours, and software behaviours (basically everything...). I don't want to scorch [especially] the tweeters!
(the amp for the tweeters will do about 60W RMS into 6 ohms before distortion begins to rise (which is about the same as the amp in the active version of those speakers), but due to their design, accidental LF tones could actually result in the dome being forcefully ejected (kinda like that 6' (foot) woofer in the guitar amp in Back to the Future), and I don't want that!!

So the woofers I have, are Dayton Audio RS150P-4A

They are cast baskets, phase plug, 150mm. They're a small 150mm, though, because of two things: first, it's small because it just is, and second, the phase plug (takes some area away in the middle) . Its surface area is 85 sq. cm, a 150mm cone would be 176 sq. cm - big difference!

Paper, Kevlar, Glass composite cone
Aluminum phase plug
Cast basket with 6 points + large openings
I believe I read somewhere there's a copper shorting ring in the motor

It's their "reference" range, for what that's worth - I think it's a pretty good woofer. I don't remember what I paid because it's been a few years now, but they were at least 100 bucks a piece. Their xmax is 4.4mm

The frequency response chart above shows this driver can be crossed at up to 4kHz (with high fidelity results) no problem. The slight impedance bump around 1.5kHz implies the driver begins becoming directional at approximately that frequency. A trick I like to do is aim to cross at about twice that - avoiding both the baffle and driver beaming

For the active practice, the crossover point can be made to be pretty much anywhere between 2kHz and 5kHz, but when I'm done messing around with active, I'd like to make it function as passive. I'm sure I'll find an exact frequency I think works best, but since I can't know that in advance, I'm looking for a tweeter that is matched (re: quality), that can be crossed easily (without a notch) second order at 3kHz. If it's 8 ohms, it needs to be at least 92dB/W so I can pad it down a little (and in case the woofer is more efficient or the tweeter is less efficient than datasheets)

Anyone have a tweeter suggestion? I don't want to spend more than $50 each, but I'm not opposed to $80 each if the value is there. I have no idea about raw drivers lately

 

[Devnull]

Dayton RST28 in either fabric or AL dome would be a good match. They've been used in a lot of DIY projects so there should be a bunch of people around if you've got questions.

 

[LSPhil]

Dome tweeter Audax TW025A14, 4 ohm, 1-inch voice coil, 3.94 inch front
Dome tweeter SB Acoustics SB26ADC-C000-4, impedance 4 ohm, voice coil 26 mm

 

[DVDdoug]

but due to their design, accidental LF tones could actually result in the dome being forcefully ejected (kinda like that 6' (foot) woofer in the guitar amp in Back to the Future), and I don't want that!!

Use a capacitor as a "protection filter". The cutoff should be lower than the crossover frequency so it doesn't interfere. That's always "good practice in case "something bad" happens with the amplifier or the crossover. (High-Pass Filter Calculator)

 

[mike7877]

Dayton RST28 in either fabric or AL dome would be a good match. They've been used in a lot of DIY projects so there should be a bunch of people around if you've got questions.

I can't do the RST28 - it looks like a decent tweeter at a pretty good price, but I couldn't cross it any lower than 3kHz (and I might want to do that)

Its Qts is almost 1 and with its fs at 750Hz, two octaves above Fs is the minimum crossover frequency (and still not ideal)
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)

 

[mike7877]

Dome tweeter Audax TW025A14, 4 ohm, 1-inch voice coil, 3.94 inch front
Dome tweeter SB Acoustics SB26ADC-C000-4, impedance 4 ohm, voice coil 26 mm

The Audax is almost compatible - would need a notch but maybe I can't get around that.
The SB Acoustics Qts is 1.1 so would force a really high crossover as well. The Audax 0.7 is close enough, I think, that it wouldn't be an issue.

Thanks for the suggestions - I'll keep the Audax in mind

 

[mike7877]

Use a capacitor as a "protection filter". The cutoff should be lower than the crossover frequency so it doesn't interfere. That's always "good practice in case "something bad" happens with the amplifier or the crossover. (High-Pass Filter Calculator)

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

 

[LSPhil]

Peerless D27TG-35-06

 

[Doenerkunde]

SB Acoustics SB26ADC-C000-4 / SB26CDC-C000-4 are probably the tweeters of choice around 50$/pc. They are employed in some very impressive loudspeakers like…

Revel Performa3 series
E.g. Revel M106: https://www.audiosciencereview.com/...ds/revel-m106-bookshelf-speaker-review.14363/

ASCI Labs
E.g. ASCI Lab C5B: https://www.audiosciencereview.com/forum/index.php?threads/ascilab-c5b-review.58753/

Mesanovic CDM65: https://www.audiosciencereview.com/.../mesanovic-cdm65-studio-monitor-review.59353/

Audio First Fidelia:

 

[FERCSA]

+1 for the SB26CDC but I warn you here there is a chance you'll fall in love with it.
Very transparent, low distortion and will open your soundstage you never heard before especially with a sota amp like la90 or b100.

 

[LSPhil]

The capacitor won’t solve anything;it will only complicate the situation. This is because it acts as a6dB/Oct filter that adds to your DSP filter, and if you’re not ableto manage this setup properly, it’s better not to use it. Ifsomething goes wrong, the capacitor won’t prevent it.
For example, a friend of mine had anissue with his amplifier where there was a poor connection betweenthe plug and the amplifier input. As a result, there was a whistlingsound around 7 kHz, and his Scanspeak tweeter (worth 600 Euros)burned out. Here’s a basic rule to follow:
1. Set all levels to 0 before turningon the amplifier.
2. Connect the speaker, start playingmusic, and gradually increase the gain.
The same applies to headphones:start playing music first, and only then put them on your head.
By the way, there are no good tweetersthat perform well below 2.5 kHz. The idea that tweeters can handlefrequencies from 2 kHz is more of a fairy tale than reality.
Regarding drivers, the Dayton AudioRS150 is an excellent choice. I own their RS270-4 and RSS315HF-4, andthey are truly outstanding. If you want the best quality, I highlyrecommend ScanSpeak. I recently purchased a used ScanSpeak 18M/4631,and I can confidently say it’s suitable for a two-way speakerproject.

 

[gfinlays]

I can't do the RST28 - it looks like a decent tweeter at a pretty good price, but I couldn't cross it any lower than 3kHz (and I might want to do that)

Its Qts is almost 1 and with its fs at 750Hz, two octaves above Fs is the minimum crossover frequency (and still not ideal)
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)

What are you doing front-end for DSP? 2 octaves above Fs is a good guide for passive xovers, but with actives, you have a lot more choice with crossover slopes. I've recently started using Acourate - Uli's proprietary UB11jPol filters are 300 dB/octave (yes you read that right), so tweeter crossover points needn't be confined by conventional assumptions. Mine are crossed at 2 kHz.

This is what a 2 kHz high pass UB11jPol filter looks like compared to LR2 and LR4:

 

[BenB]

The Audax is almost compatible - would need a notch but maybe I can't get around that.
The SB Acoustics Qts is 1.1 so would force a really high crossover as well. The Audax 0.7 is close enough, I think, that it wouldn't be an issue.

Thanks for the suggestions - I'll keep the Audax in mind

What's your reasoning for associating a Qts of 1.1 with a really high crossover?

 

[mike7877]

What are you doing front-end for DSP? 2 octaves above Fs is a good guide for passive xovers, but with actives, you have a lot more choice with crossover slopes. I've recently started using Acourate - Uli's proprietary UB11jPol filters are 300 dB/octave (yes you read that right), so tweeter crossover points needn't be confined by conventional assumptions. Mine are crossed at 2 kHz.

This is what a 2 kHz high pass UB11jPol filter looks like compared to LR2 and LR4:

View attachment 426534

I think you just misread my OP - I'll be using these two drivers as the trial run for my drivers that cost 10-20x more. I'm going to learn the software and ensure the hardware is stable long term (no pops etc). Once I've got things configured and the drivers are perfectly integrated, I'll know I'm good to move to the next drivers. Presumably this will take me at least a few evenings - this will be the confirmation that no unwanted behaviours are observed (I'm going to have to observe DAC directly into power amp behaviour, specifically).

Imagine buying a Ferrari and sucking at stick - you don't then buy a $500 beater to learn - something in the 7-12 grand range would be a helpful though.. I feel these drivers are the 10k learner car lol

 

[mike7877]

The capacitor won’t solve anything;it will only complicate the situation. This is because it acts as a6dB/Oct filter that adds to your DSP filter, and if you’re not ableto manage this setup properly, it’s better not to use it. Ifsomething goes wrong, the capacitor won’t prevent it.
For example, a friend of mine had anissue with his amplifier where there was a poor connection betweenthe plug and the amplifier input. As a result, there was a whistlingsound around 7 kHz, and his Scanspeak tweeter (worth 600 Euros)burned out. Here’s a basic rule to follow:
1. Set all levels to 0 before turningon the amplifier.
2. Connect the speaker, start playingmusic, and gradually increase the gain.
The same applies to headphones:start playing music first, and only then put them on your head.
By the way, there are no good tweetersthat perform well below 2.5 kHz. The idea that tweeters can handlefrequencies from 2 kHz is more of a fairy tale than reality.
Regarding drivers, the Dayton AudioRS150 is an excellent choice. I own their RS270-4 and RSS315HF-4, andthey are truly outstanding. If you want the best quality, I highlyrecommend ScanSpeak. I recently purchased a used ScanSpeak 18M/4631,and I can confidently say it’s suitable for a two-way speakerproject.

I'm putting the cap in series with the tweeter, not across it - for DC blocking vs. 1st order crossover.

The tweeter I have is rated to work down to 1.5kHz - if you've ever seen ATC's 3 inch dome, it's a mini version of that. It's got a dual suspension so no rocking modes (its linear travel is literally 1cm peak to peak (and it will move further, I've seen it). It's underhung design, the magnet used is just a step or two below lab strength (n48 or 52 or something) and the gap is 3mm. The gap itself is completely saturated (so completely uniform) which is a large reason it keeps 3rd harmonic distortion down between mostly -80 to -90dB and never higher than -70 from 1.5kHz to 20kHz. The width of the gap for the voice coil is tiny because the alignment is second to none (dual suspension!!). The VC is a rectangular copper clad aluminum, wound with no space and it's only wound to two layers. There's no ferrofluid to dry out and screw things up. And even though its moving mass probably double that of any fabric dome made since 1971, its -3dB point is 22kHz, -2dB is 20kHz. The magnetic circuit was FEA optimized, the dome itself is woven and is damped like a CLD woofer but different. This is the best tweeter, I'm already sure of it. There's more, but I think I've said enough.
And even if there's a tweeter that's slightly better, I already have these ones lol

edit: to add, in active form (this tweeter included in an active model), I think I already said this, but it gets its power from a 50W RMS class ab amp, and the amp I'll be connecting to it provides about the same. I'll be setting it to low gain in the beginning to be extra safe!
I might actually use a headphone amp (which has the same gain) in its place. It can only put out 1-2 watts into 6 ohms, but at 91.5dB/2.83V (it's 6 ohms, that's why the rating is given this way) that's enough power to play around with in the beginning. The HP amp I'm considering for tweeter safety in the beginning is my L30 II

 

[mike7877]

What are you doing front-end for DSP?

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

 

[mike7877]

What's your reasoning for associating a Qts of 1.1 with a really high crossover?

The way I look at tweeters, if you want to be able to use it somewhat low (ie. from 0.5 to 1.5Fs especially) and you want high fidelity, Qts needs to be between 0.5 and 0.65, maybe up to 0.75 if it's really nice otherwise. Reason being, in the two octaves above resonance, there will be added ringing from the tweeter. It won't be a lot (unless Q is really high, and it can be - some tweeters I've seen have Qts of over 4), but it will colour the sound. The way those octaves interact with passive crossover parts also tends to mess with amplitude - even with a notch filter: this screws up driver integration. If you wanted a passive crossover on a tweeter with Qts of 1.8 (very common) to cross at 1 octave above Fs, I don't think you could do a good job without 40 parts (and a bunch of talent, time, and patience, of course).

Most simply:

Primary affected: Frequency response
Secondary: Causing poor driver integration (from frequency response from altered amplitude and late + hanging tones)

I don't even think you could actively cross a tweeter with Fs 900Hz at 2.1kHz (2nd order) if its Q was 1.1.
Well, you could, but there would be problems: if you wanted to do the best job you could, you'd have to equalize the bump in response before the 12dB/oct roll off first, and then you're done (not entirely, obviously, but for this aspect). And you're still stuck with the ringing. In most cases it won't be an audibly bad result, but in this case, these tweeters (and the rest of the system) are very transparent, so it's especially important

I don't know if you have any practice with sealed woofers with their 12dB/oct natural rolloff. If you do, and you've known their drivers' Qs, and you'd equalized away the frequency response bumps from their Qs of >0.9, and you've had exposure to truly full-range systems or headphones which extend low and are well damped, the difference you hear between the truly high-fidelity full range systems' 50-150Hz, and the sealed+equalized 50-150Hz, is the ringing. With bass, the way I describe it is it's kind of.. soft. It doesn't start exactly when it should and nor does it stop, and this is what causes the "soft". To describe this at higher frequencies is not so straightforward, though, and the way that a problem manifests depends a lot on the spectral content of the music you're listening to. Since a lot of sounds are complex with components all over the spectrum, it's often a surprise when a problem arises (or you don't attribute the characteristic to the cause). I don't want to be any more vague than I'm already being, but I don't think I can be more descriptive unfortunately!

 

[Ktacos]

I can't do the RST28 - it looks like a decent tweeter at a pretty good price, but I couldn't cross it any lower than 3kHz (and I might want to do that)

Its Qts is almost 1 and with its fs at 750Hz, two octaves above Fs is the minimum crossover frequency (and still not ideal)
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)

Um the rst28 can easily cross around 1400hz.

Sure you're ready to embark on such a task?

 

[mike7877]

SB Acoustics SB26ADC-C000-4 / SB26CDC-C000-4 are probably the tweeters of choice around 50$/pc. They are employed in some very impressive loudspeakers like…

Revel Performa3 series
E.g. Revel M106: https://www.audiosciencereview.com/...ds/revel-m106-bookshelf-speaker-review.14363/

ASCI Labs
E.g. ASCI Lab C5B: https://www.audiosciencereview.com/forum/index.php?threads/ascilab-c5b-review.58753/

Mesanovic CDM65: https://www.audiosciencereview.com/.../mesanovic-cdm65-studio-monitor-review.59353/

Audio First Fidelia:

The SB Acoustics does look like a nice driver. If people love its sound, it is compatible with the RS150P-4A, very well!
Due to its Q and the fact I'd be doing this passive, I think a 4-4.5kHz cross is where I'd end up (not a problem for what's effectively a 2-2.5" driver (a 4-5" driver with phase plug..)

Am I reading its datasheet correctly, it covers up to 18kHz +/- 30 degrees (60 total)? and to over 16kHz +/- 60 degrees (120 total)??
If so, DAYUMMM!
And it sounds nice, too!

I like when things come together well. I'm strongly leaning toward the SB Acoustics SB26ADC now. When practicing active on it I'd be discovering the optimal frequency to cross at 12dB/oct - hopefully it's 4kHz or so!
The 150mm driver having half the surface area of most 150mm drivers and also having a (immovable) aluminum phase plug in the center virtually halving the size of the driver again (as is demonstrated from the frequency response in op: A-mazing) should be ideal for 3-5kHz crossover!

 

[mike7877]

Um the rst28 can easily cross around 1400hz.

Sure you're ready to embark on such a task?

No. If it was going to stay active, I'd listen to it and if I liked it I would definitely go with it no problem. But I'm not an experienced passive crossover designer (and I don't have a shop full of coils and caps to pull from), so because of its Q of 1, it's not ideal for me.

You could call it the desire to keep part count low, the inability to make crossovers with high part counts, or maybe even laziness lol.

I've seen the tweeter before, and heard good things - it just doesn't have the ideal attributes (specifically Q)

edit: to add, it also has ferrofluid - I try to not have ferrofluid in drivers because I want my speakers to be 100% maintenance free. I might give or sell them to someone in the future, and I don't want something from me to have problems - I go so far as to choose rubber surrounds now since my Infinity SM112s foam surround woofers gave out on me in 2017. They were originally my parents - while they did get 27 years of good use, those surrounds eventually went, and so did some proprietary component on the crossover
It sucked about the crossover because it was an age related thing, so buying a replacement wouldn't guarantee the replacement would last long (or even work)
So now I get parts made of things known to last long, and no more parts than necessary (within reason lol)

edit2: It is relatively easy to cross at 4ish kHz too (like the SB26ADC).
So far what the SB has over it is that it's not ferrofluid cooled and its dispersion... Its dispersion actually looks like something truly special...
The SB also has a much longer xmax - not really applicable in this design (at 4kHz), but if I don't end up liking the combination, if I were to use the SB with a different woofer in the future (and buy a different tweeter for the RS150P-4A), its 2.5mm gap + 1.3mm coil + 0.6mm xmax would be better for transients when crossed at lower frequencies.. Though if I were to cross the SB at a lower frequency, the crossover would have to be active and I probably wouldn't be doing that...
I lost focus with edit2 lol. I didn't sleep last night - it's time for that. Thank you for the suggestion