Need help choosing ScanSpeak either 18W or 22W

[kstich]

So, I have decided that I want to build a large three way speaker using ScanSpeak drivers. I have heard great things about the 18W/8531 from Zaph tests and the designs there and on Troel's site as well.

http://www.zaphaudio.com/ZRT.html

http://www.troelsgravesen.dk/ektagrande.htm

The plan so far is a TMWW with the 6600 and the 15W8530. I chose the 15W because it is coated and has the SD-1 motor and tests a little better than the 12M. The crossover will be an electronic LR4 and the speakers will be tri-amped.

I have also found the 22W/8857 with the Aluminum cone on the ScanSpeak website and am very interested in it as well. However, there is little information available about this driver on the web. Has anyone heard (or tested) this driver in comparison to the 18W? I thought it may be better for crossing with the 15W. Any input would be appreciated.

Thanks guys.

[cotdt]

I would go for the uncoated version of the 15W instead, but the one with the larger "low Qts" motor. The uncoated version just sounds better, and perhaps that's why Scanspeak opted for the uncoated cone for their latest version of the 15W with the rubber magnet boot.

[Jonasz]

I would go for one of the SS 26Ws in a TMW configuration. The SS 26W sounds just awesome! 6600/15W/26W with an active crossover in a box similar to this one should render a really good result. I've heard the Respons loudspeaker a number of times and they sound just plain better than the B&W802D for example.

http://www.responsaudio.com/produkter/grand_d.html

Sorry for the LARGE pic...

[Rick Craig]

So, I have decided that I want to build a large three way speaker using ScanSpeak drivers. I have heard great things about the 18W/8531 from Zaph tests and the designs there and on Troel's site as well.

http://www.zaphaudio.com/ZRT.html

http://www.troelsgravesen.dk/ektagrande.htm

The plan so far is a TMWW with the 6600 and the 15W8530. I chose the 15W because it is coated and has the SD-1 motor and tests a little better than the 12M. The crossover will be an electronic LR4 and the speakers will be tri-amped.

I have also found the 22W/8857 with the Aluminum cone on the ScanSpeak website and am very interested in it as well. However, there is little information available about this driver on the web. Has anyone heard (or tested) this driver in comparison to the 18W? I thought it may be better for crossing with the 15W. Any input would be appreciated.

Thanks guys.

I've used the 22W/8857 and it's an excellent woofer. It will smoke the 18W in terms of extension and impact.

[Paul K.]

I second the 22W/8857

I used the 22W in an ML-TL where Rick Craig designed an excellent crossover (the speaker is named canTiLena and there is an old thread on this forum somewhere with photos). It's a super good-sounding woofer with very well-defined bass and amazing dynamic range, and worked very well in a 3-way with a Morel MDM55 and Fountek NeoCD1.0 ribbon tweeter. The canTiLenas are my "reference" speakers now and for the forseeable future.
Paul

[TacoD]

I am using the Scanspeak 26W with alu cone, and it's very good sounding (I swapped an Eton 11" for the 26W)

[Jed]

A ported Seas Excel W26 will get you bass into the 20s with VERY low distortion as well. I'm using it in a few speakers. The other nice thing about it is it will play LR2 with no issues. I have almost an exact speaker kit like you are looking for. I like the non coated Scan Speak W15. As long as you tame the 840hz peak it is one of the most detailed and smooth sounding mids I've ever heard.

[sdl2112]

I am using the Scanspeak 26W with alu cone, and it's very good sounding (I swapped an Eton 11" for the 26W)

Hello TacoD,

Did you measure the T/S parameters of your 26Ws? If so could you post your numbers. The reason I ask is that I have a pair of 8867s and they are off quite a bit from the published SS data. If this is a trend anyone considering these should take that into account.

[Jed]

Hello TacoD,

Did you measure the T/S parameters of your 26Ws? If so could you post your numbers. The reason I ask is that I have a pair of 8867s and they are off quite a bit from the published SS data. If this is a trend anyone considering these should take that into account.

Did you model your TS numbers in Unibox to see if the actual box size and tuning needs to change? Often time the TS numbers are different, but the end result for box/tuning is the same.

[sdl2112]

Did you model your TS numbers in Unibox to see if the actual box size and tuning needs to change? Often time the TS numbers are different, but the end result for box/tuning is the same.

I don't want to hijack this tread so I'll try to keep it short. Yes I have and I often find what you said to be true. Usually the fs and Qe (thus Qts) a little higher, and the Vas a little lower compared to published specs. These movements of numbers tend to cancel and you end up with about the same response.

I'm at work so I don’t have the numbers in front of me but from memory the Fs was 22.5 vs. 19Hz, the Qts was 0.47 vs. 0.3 and the Vas was 125 vs. 179l respectively for measured vs. published. This was after 80hrs break-in at approx 2/3 xmax. I tried numerous drive levels (I'm using SoundEasy) and even got out my function generator and used the brute force constant voltage and constant current procedures with similar results. In these cases just as you said the numbers were different but the response was similar to the MLS derived numbers from SE. Other checks like MMs were good so I feel it is accurate. BL was a little low so my best guess is that the magnets may not be as charged as were the prototypes used for the published data.

I see similar trends especially in the 15W SS drivers tested by Zaph. One reason I chose this driver was it looked like it would work very well in either a sealed or vented enclosure. I was leaning towards sealed anyway but vented is virtually out of the question unless I want a really big box. At the expense of these drivers I would think the parameters would be more consistent. Other wise these drivers are great. The build quality is top notch and the distortion is very low typical of SS drivers. I don't have them in a design yet so I cannot comment on the most important part, the sound! I said short.....oh well.

[kstich]

Thanks for the many replies already.

It is interesting that there are a couple of postsabout the uncoated version of the 15W. I had chosen the 15W8530K00 because of the slightly better distortion measurments on the Zaph website. Also, I may in the future want to match it to a ZD5 from there as well.

It sounds like the SS woofers are all top notch. I have not heard anyone say something bad about them. I am thinking 2 22Ws in a 2.25 ft3 vented box will give great clean bass.

[cotdt]

It is interesting that there are a couple of posts about the uncoated version of the 15W. I had chosen the 15W8530K00 because of the slightly better distortion measurments on the Zaph website. Also, I may in the future want to match it to a ZD5 from there as well.

There are many different versions of the uncoated 15W. Scanspeak changed the voice coil former material in the newer versions, and they have lower Qts now, and for whatever reason mine has a really smooth response without any peaks (other than the typical Scanspeak shelving used to simplify BSC). Mine is the one with the extra large magnet:



The bass is really nice for a 5.5", and extends quite low even in a small enclosure. The version Zaph uses is the "low Qts" version with a very high Qts and requiring a relatively large enclosure. I didn't find any need for a bass unit, and instead used a subwoofer for the deep bass. The 15W is a really nice sounding unit, one of the best midranges I've heard.

For a dedicated midrange, there is a 15M version that uses a shorter voice coil optimized for a midrange. Not sure if there are any real world advantages over the 15W other than higher efficiency though. But if you're going to use a bass unit, the 15M is the one to go for.

[TacoD]

Hello TacoD,

Did you measure the T/S parameters of your 26Ws? If so could you post your numbers. The reason I ask is that I have a pair of 8867s and they are off quite a bit from the published SS data. If this is a trend anyone considering these should take that into account.

I did not measured T/S, the frequency response is spot on. Like Jed posted earlier, a higher Fs almost always results in a lower Vas -> same tuning. As the surround loosens, the Fs will be lower and Vas will increase. In my opinion it is within measuring & production tolerances.

By the way how did you measured the T/S,
- delta mass or delta volume?
- low signal levels
- driver mounted vertical or horizontal
- temperature of the room
- burn in of the woofer
- used equipment

Given all these variables I see nothing wrong? My pair only had some glue residue on the dustcap. Didn't expect that for such a expensive driver, since the company is in the hands of Tymphany the marketing speak increased but production quality decreased 8O

[sdl2112]

I did not measured T/S, the frequency response is spot on. Like Jed posted earlier, a higher Fs almost always results in a lower Vas -> same tuning. As the surround loosens, the Fs will be lower and Vas will increase. In my opinion it is within measuring & production tolerances.

By the way how did you measured the T/S,
- delta mass or delta volume?
- low signal levels
- driver mounted vertical or horizontal
- temperature of the room
- burn in of the woofer
- used equipment

Given all these variables I see nothing wrong? My pair only had some glue residue on the dustcap. Didn't expect that for such a expensive driver, since the company is in the hands of Tymphany the marketing speak increased but production quality decreased 8O

I'm don't think there is anything wrong either, I just expected the response/box tuning closer. To answer your questions;
- I used the delta volume method
- I tried every signal level possible from 1ma to 5vrms
- I have a sturdy open frame jig that I clamp the driver to mounted vertical (as it would be in an enclosure)
- allowed the drivers to cool after burn-in in a 25C room.
- I measured the drivers periodically out to 80 hrs at 2/3 xmax where the parameters settled in.
-The drivers were bought new and the test equipment I have check with a scope.

Btw, I have some Eton 11" drivers that measure quite close to published data.

[tf1216]

sdl2112,

You have a PM.

[kstich]

That is a beautiful driver, but I can't find any info on the K03 variant. Where did you get it?

[cotdt]

That is a beautiful driver, but I can't find any info on the K03 variant. Where did you get it?

They are custom made by Scanspeak, perfect for my particular application.

[kstich]

Must be nice to have the perfect driver

[TacoD]

I'm don't think there is anything wrong either, I just expected the response/box tuning closer. To answer your questions;
- I used the delta volume method
- I tried every signal level possible from 1ma to 5vrms
- I have a sturdy open frame jig that I clamp the driver to mounted vertical (as it would be in an enclosure)
- allowed the drivers to cool after burn-in in a 25C room.
- I measured the drivers periodically out to 80 hrs at 2/3 xmax where the parameters settled in.
-The drivers were bought new and the test equipment I have check with a scope.

Btw, I have some Eton 11" drivers that measure quite close to published data.

I think you did a proper job, did you already asked Scanspeak if they changed the parameters?

[sdl2112]

I think you did a proper job, did you already asked Scanspeak if they changed the parameters?

I emailed them but received no reply.

[Martijn_H]

I don't understand why you decided to go 4th order electric before you have seen any measurements. With the 15W en 6600 tweeter you have plenty of playground to tune to the best on and/off axis and phase respons.

Why tri amping? Are you not talking about a passive xo network?

I would use the alu or paper 26W. They are one of the best performing 10" woofers in the low range. I know that the 15M mid does worse than the 15W. Offcourse the 22W's are nice also.

EDIT sorry, I now quess that you meant active xo. my bad!

[kstich]

i plan on testing the drivers to make sure that they integrate properly and the crossover frequencies are optimized.

The crossover will be active. The money saved on crossover capacitors can buy a decent seven channel amplifier to tri-amp the speakers.

[TacoD]

i plan on testing the drivers to make sure that they integrate properly and the crossover frequencies are optimized.

The crossover will be active. The money saved on crossover capacitors can buy a decent seven channel amplifier to tri-amp the speakers.

MKP isn't that expensive :rofl: So which 7.1 amplifier can 100 dollars buy you?

[tf1216]

sdl2112,

I sent you another PM.