No, the first rule In 2-way speakers Is "You do not talk about 2-way speakers".
How do you feel about crossing it at 1.75kHz with the D3004/662000 iluminator tweeter?
@ dot and marley, I'd vote for the first rule to be to simulate the speakers first.
@akarma: this simulation shows that SS 18W4531G00 can be low passed sufficiently good at about 2 kHz with just a simple 2nd order electrical filter. That it will sound very good with any tweeter of equivalent quality, high passed in the same manner, goes without saying even though newby DIYers may have trouble believing it.
@akarma: this simulation shows that SS 18W4531G00 can be low passed sufficiently good at about 2 kHz with just a simple 2nd order electrical filter. That it will sound very good with any tweeter of equivalent quality, high passed in the same manner, goes without saying even though newby DIYers may have trouble believing it.
Attachments
HaHa! 😀
The enthusiasts always follow these sort of threads, perhaps because we are trying to fix the wounds we still carry from grappling with the horrible 6" plus 1" speaker. 😱
Why is it a horror?
* Worst time alignment problems of any two way.
* The 4-6kHz cone breakup.
* The tendency to needing low crossover with all the usual tweeter problems.
Now add the 6" scanspeak horrible frequency response, and we have a complex crossover and maybe a difficult cabinet construction to try and fix the phase.
This is entering a World of Hurt and constructional expense, IMO. 😱
I won't spend time on the Zaph ZRT, because it's an 8 ohm driver and off-topic.
Zaph|Audio - ZRT - Revelator Tower
For the (IMO) horrible 18WU/4741-T00, we might consider Troels' Illumina-66 with the ring radiator:
Illumina-66
And the flat baffle NADA 2-way with beryllium tweeter:
NADA 2-Way Klang + Ton Kit Pair using Illuminator Drivers
Crossover enthusiasts will understand these two designs immediately. Both crossing around 2.7kHz. A lot to like in both of them in engineering terms. But would I take them on? No. I think there are much easier and more cost effective ways to build a good speaker.
The enthusiasts always follow these sort of threads, perhaps because we are trying to fix the wounds we still carry from grappling with the horrible 6" plus 1" speaker. 😱
Why is it a horror?
* Worst time alignment problems of any two way.
* The 4-6kHz cone breakup.
* The tendency to needing low crossover with all the usual tweeter problems.
Now add the 6" scanspeak horrible frequency response, and we have a complex crossover and maybe a difficult cabinet construction to try and fix the phase.
This is entering a World of Hurt and constructional expense, IMO. 😱
I won't spend time on the Zaph ZRT, because it's an 8 ohm driver and off-topic.
Zaph|Audio - ZRT - Revelator Tower
For the (IMO) horrible 18WU/4741-T00, we might consider Troels' Illumina-66 with the ring radiator:
Illumina-66
And the flat baffle NADA 2-way with beryllium tweeter:
NADA 2-Way Klang + Ton Kit Pair using Illuminator Drivers
Crossover enthusiasts will understand these two designs immediately. Both crossing around 2.7kHz. A lot to like in both of them in engineering terms. But would I take them on? No. I think there are much easier and more cost effective ways to build a good speaker.
Attachments
Looking at the Illumina 66 I believe that the chosen woofer (the illuminator 18WU/4747T00) looks worst than the revelator 18W/4531G00 that I choice if combined togheter with the 662000 tweeter. I would go for these two drivers
There's no problems with linear response. It just sounds dull and boring. MW16P has some transparent and more sparkle midrange. SS18W puts out good bass for its size and that's all. It's my opinion after comparison.
And SS18 is very mediocre on high side. XOver must be low enough. At least 1.7-1.5kHz
Exactly!
I'm finding it desperately hard to keep up with all these different Scan woofers and tweeters! 😀
I always think there's the scruffy one that looks like the kids ran sandpaper over it, the one that got slashed by the kids and then stuck back together, and the one with the wavy glue pattern the kids painted onto it. Then, at various expense, you can have underhung voicecoils and sexy looking curvy magnets and a coating of polish to make it look nice. 😱
There's no doubt that 4 ohm 6" bass is easier to deal with than 8 ohm in the filter.
But you do seem extremely confident that you can get this thing up and running. IIRC, Troels usually says he thinks the Scan 6" units are really no use above 2kHz at all and not ideal for a regular two-way. They do great bass, and that's the range where you should use them.
Thing about a problem speaker, is maybe you CAN fix most of the issues, but you end up so constrained that you can't deal with other interesting areas like low distortion, power handling and good impedance and dispersion.
As it goes, I have heard some excellent Scan 6" systems. The scruffy one. But some serious bucks were thrown at it. Including exotic ribbon tweeters and electronic crossovers and subwoofers. And about a years work.
The 6" that works nicely is the polycone, because it has sufficient mechanical damping to save you some effort. SEAS make a good one: H1571-08 U18RNX/P
THAT, I would take on. 😎
Attachments
For the OP, some inspiration. Altho the video hardly goes into the design itself, you can look it up on Parts Expess Tech Talk.
This 2 way rev based MLTL won the hotly contested Midwest audiofest speaker design competition a few weeks back
Speaker Design Competition Winner - Paul Kittinger - Brioso - YouTube
This 2 way rev based MLTL won the hotly contested Midwest audiofest speaker design competition a few weeks back
Speaker Design Competition Winner - Paul Kittinger - Brioso - YouTube
Last edited:
Moving forward
Before buying the drivers I'm simulating using their electric model, zobel and notch compensation, and I ended up with this BW3 crossover, I attach you some pics to get an opinion about that.
I still miss some components to compensate the cabinet, but I'll keep simulating as soon as I decide the model
Before buying the drivers I'm simulating using their electric model, zobel and notch compensation, and I ended up with this BW3 crossover, I attach you some pics to get an opinion about that.
I still miss some components to compensate the cabinet, but I'll keep simulating as soon as I decide the model
Last edited:
Crossed at 800 Hz. I don't think 1" driver can go that low.
Your simulation lacks measured frequency response of drivers.
Your simulation lacks measured frequency response of drivers.
+1 Lojzek, That's what I did, though I did add a Zobel and notch filter. Even so, Madisound has produced at least 2 crossovers with the 4531G in a 2-way at 2kHz. One a 2nd order, one a 3rd.
This driver does just fine. The real issue IMHO is getting a tweeter to match at 2kHz. The 15W would be better options due to higher XO with almost as much bass, which the 18W/4531 is outstanding.
I have zero regrets, except for $$$. 🙂 I have yet to go to an audio show and see a 2-way and go "oh, I wish I had used this mid-woofer instead."
I believe Fritz speakers is also doing quite a bit of work with the 18W Revelators.
Best,
E
This driver does just fine. The real issue IMHO is getting a tweeter to match at 2kHz. The 15W would be better options due to higher XO with almost as much bass, which the 18W/4531 is outstanding.
I have zero regrets, except for $$$. 🙂 I have yet to go to an audio show and see a 2-way and go "oh, I wish I had used this mid-woofer instead."
I believe Fritz speakers is also doing quite a bit of work with the 18W Revelators.
Best,
E
There it is...from datasheet. I can't simulate the frequency response since I don't have bought the drivers yet.
So, here are some ideas.
The 18W/4531 measures much better than spec re: FR at the top end. It's an easy crossover to make at 2 kHz. And you get really awesome bass for a 2-way. Much more impressive than say the Magico S1 Mk II.
This limits your tweeter selection however. Larger AMT's, and dual magnet selections from Peerless and SS are going to be very good choices, but also some of the deep-chamber varieties, which SS has in soft/ring and even Be.
The Peerless XT25BG would be a great choice, though the SS version has lower distortion.
If you stick with SS, the 15W helps you out with a higher XO, but also higher f3.
In either event, the SS demand a big cabinet.
A much easier 2-way mid-woofer is the FaitalPRO 5FE120. That can be crossed over at 3-5kHz easily, with great dispersion characteristics. This really opens up your tweeter choices, and fits in a 0.25 ft3 cabinet. You won't get the huge bass, but it is much easier to design, AND the smaller cabinet easier to live with.
Best,
E
The 18W/4531 measures much better than spec re: FR at the top end. It's an easy crossover to make at 2 kHz. And you get really awesome bass for a 2-way. Much more impressive than say the Magico S1 Mk II.
This limits your tweeter selection however. Larger AMT's, and dual magnet selections from Peerless and SS are going to be very good choices, but also some of the deep-chamber varieties, which SS has in soft/ring and even Be.
The Peerless XT25BG would be a great choice, though the SS version has lower distortion.
If you stick with SS, the 15W helps you out with a higher XO, but also higher f3.
In either event, the SS demand a big cabinet.
A much easier 2-way mid-woofer is the FaitalPRO 5FE120. That can be crossed over at 3-5kHz easily, with great dispersion characteristics. This really opens up your tweeter choices, and fits in a 0.25 ft3 cabinet. You won't get the huge bass, but it is much easier to design, AND the smaller cabinet easier to live with.
Best,
E
Erik what do you think of the 18/4531 and 662000 SS illuminator tweeter crossed at 1.5kHz?
SIM: 18W4531 G00 & D3004-660 000
A contemporary higher quality Scan Speak 2-way speaker is what any DIYer would have gladly taken as a very good compromise between sensitivity and low end output. Normally, this kind of a product could cost retail easily in the vicinity of 5 grand.
A contemporary higher quality Scan Speak 2-way speaker is what any DIYer would have gladly taken as a very good compromise between sensitivity and low end output. Normally, this kind of a product could cost retail easily in the vicinity of 5 grand.
Attachments
- Status
- Not open for further replies.