SS18w8545/ focal tc120tdx two way

Hi folks, just wondering if anyone else out there has used these drivers to make a two way only system; i.e. not a watt/puppy clone. I have found documented designs on the web, but only for two and a half way or sub/sats like the wilson clones. I have used these units myself in a 16 litre-ish box with second order electrical xovers and am fairly happy with the sound, but would like to hear from any others who feel they might have something good.
Personally, I reckon the scan carbon woofer is a star; it sounds open and dynamic, and has bass beyond its size; the focal tweeter seems to blend well, with a complementary tonality and speed- percussive cymbal hits for example seem more realistic than soft- domes. I guess that I've had these for five years now and I'm still looking to play a bit more, so any and all comments are welcome.
I'm new to these fora, so hello to all; this community has created a truly amazing resource- kudos to each and every one of you.

Thanks in advance
 
I have no experience of the SS woofer but the TC120tdx (mine is tdx2) was my first venture into 'high end' drive units back in Uni. I couldn't get over the energy and detail at the time but now find it a bit tiring. I too am still waiting to see if I can tweak things a little more to get the best out of it, as it's still a quality unit capable of audio magic.

I'm afraid I went with a 3 or 2.5-way design as well, but I'm a little confused as to exactly what help you are asking for, as you seem very happy with the units themselves. Do you just want ideas for new crossover alignments? If you do, and you may well have seen this already, but your speakers are basically identical to the top half of Tony Gee's Andromeda speaker here. He has made a few alterations to the crossover it seems, but is still running it as a 2-way with sub-woofer. Trusting his ears and experience with these units, employed in one of his most impressive designs, I see no reason not to try out his crossover. Then send him a message of thanks for putting a smile on your face!

I've only just arrived on the boards myself (appreciate your classical education btw ;)), having somehow survived for a decade of building on trial and error and a load of other websites! It was an interest in building a transmission line that led me here and I too am bowled over by the breadth and depth of knowledge, as well as the size and activity of the community. There was me, wondering who all these driver manufacturers were selling their other units to! :)
 
First off, welcome to diyaudio, float.

Like Noksukau, I have no experience with the Scan woofer and am using the slightly different TC120TD5 in my latest favorite system with a Focal 6w4254. You didn't mention your crossover freq, but I use a 2400 Hz, 4thorder LR active XO.

Looking at the response Tony Gee posted in his andromeda, I suspect that there is some breakup in the Scan cone starting as low as 3KHz. You may find the results cleaner crossing over below that with a higher order filter and/or using a notch filter to tame the breakup modes.

Using a 4th order XO at 2400 Hz my tweeters show no sign of stress at any lisening level that I can stand for more than a few seconds. I had used a second order filter, which seemed a little harsh. I don't know if it was due to overstressing the tweeter, exciting the woofer's breakup modes or both, but it cleared up with a 4th order filter and a notch filter on the woofer.

If you want to get really fussy, you might try a notch filter on the tweeter to tame that small peak in response at around 3-4K. I plan to try it on my next version of the crossover. Not that I hear a need, but the Moxlite boards have eq in the tweeter section so why not?

I am a big fan of active crossovers. Besides being easier to impliment baffle step compensation, EQ and phase compensation, they are less expensive to tweak (meaning I am more likely to keep optimizing) it also gives you an excuse for another project if you need extra amp channels. ;)
 
Hi Noksukau, and thanks for answering. I read my post again and it is a bit hard to find the question!
Basically, I am looking for alternative crossover schematics, but with full baffle step correction. When I first put these drivers in a box it was with essentially the same values as the andromeda and I was very impressed with the detail, openness,etc. but with the 0.4mH inductor on the midbass there was of course nobass! Adding a woofer underneath is one solution, but I don't really need the volume level so I am happy to trade off sensitivity and stick with a two way.
I am using a 1.8mH on the bassmid to give BSC and a flat response ( although this reduces system sensitivity to around 83dB/watt) and this crosses at about 2300hz to the focal.
As I think I suggested before, I am happy but I'd be very surprised if I can't do better so I was hoping others may have been playing with the same ingredients and wanted to share experience.
I must say the focal "sting" has never bothered me- I cant hear much over 16khz anyway, although my tweets do seem maybe a little hot around 4-5 khz .
 
I'm using 8545K in my speakers. It is an excellent driver, as you may suppose. Yes, it has some peaks that you may wish to deal with. I'm going to try various serial and some parallel passive filters and will decide which sounds best. Moreover, the driver behaves well in higher volume enclosures, I use it in 22litres with a 29Hz port tune and for now it is the best sounding version, considering the room response.
 
Cheers Bob, your post must have appeared as I was typing my last response. My crossover slopes are about 3rd order acoustic, 2300hz, and seem pretty clean in terms of distortion. It probably helps that the tweeter is padded by around 9db so it never really gets driven that hard.
Active does sound appealing, but at the moment doubling my amp expenditure is not possible. Anyways, I've always thought that if passive is good enough for the likes of wilson and jm lab then its probably good enough for me:D
How do you rate the focal w-cones against the competition?

All the best
 
Hi Machinow, the peak on the scan that seems most troublesome is the one around 600-800hz of around 4db or so. I've not tried notching it out because listening slightly off-axis it seems to go away, particularly with a lowish crossover point (it seemed more noticable when I was using a 3.2khz xover). I have wondered if notch filters take away some of the magic:)
 
Hi float,

I did a bookshelf 2-way BR with 18w8545 and Revelator 9900 crossover
In my expirience SS8545 cannot be used with simple 2. order crossover because of it's riseing and far reaching response. Using it that way (or even worse with 1. order xover) attenuation slope above xover point is to slow. If one uses larger inductor to overcome this the crossover point gets very low. With your 1,8mH you get very wide overleap with the tweeter and it can make the highs a bit aggressive (even more so with 1,5mH - machinow ).
And, of course, you can't use Andromeda crossover because it's intended for a 3-way so you would have very bright sounding speaker without help of the bass driver. ;)
 
My last few systems have used Focal drivers so I don't have any experience designing with the competition. Before trying the 6W cone, I was using a doped paper 7C for 5-6 years. The W cone trounces the paper, much more detailed and smooth. With the W I hear nuances that were completely obliterated with the paper cone. I noticed a difference between my Leach Amp and my LM4780 amp that I didn't with the older speakers.

The W cone doesn't have the harshness of some of the kevlar and magnesium drivers I have heard in commercial systems when driven hard. Looking at the breakup response of the excel drivers for example, I think the W cone is a lot easier to tame - the peaks are much smaller. I fiddled around with the notch filter, trying with and without. I am not sure that I heard a difference, although it did show up (barely) in measurement. I guess 4th order electrical an octave below the peak may be enough to make a 7 db peak inaudible.

Sounds like you may be hearing that little peak in the tweeter's response. Notch filter time. :)

The "big guys" have to face commercial realities pushing them to passive XO. If they went active the customer now has to buy an additional high end amp. If the customer then decides to change an amp to one with different gain than the crossover was set for, then the system response changes. Not to mention the "warranty issues" when the customer hooks up the woofer output of the XO to the tweeter amp and cranks up Stanley Clarke.

If they made the levels user adjustable then they subject their reputation to the whims of their not necessarily golden eared customer. When my daughter was at home she'd always crank my subwoofer level all the way up. You'd think my system was pretty lousy, with no top end. Back the sub down to the proper level and it was decent sounding.

True, you need more amplifier channels to go active and that can add up. (this drove my brief foray into chip amps, 4 channels of LM4780 cost about $20 with junk box parts, but it couldn't hold a candle to my Leach so that amp was a giveaway) But, now that I have built more quality amplifier channels than I will likely ever need, I really like the tweakability of an active XO. I can load it with good quality caps for less than the cost of one of some speaker level caps. Shift the crossover point, phase correction or baffle step by changing out a few cents in resistors. Tweaker heaven! Also, for some reason I find the component calculations easier for active filters, but that's just me.
 
Jeeeeesus Vuki, thats some crossover:eek: How long did it take you to come up with that!!??
My experiences tally with yours exactly regarding first orders and so on.
My low pass actually consists of a 1.8mH with a 9.4parallel cap( second order electrical) and a zobel of a 3.9 cap and 8.2ohm resistor. I altered the zobel from the values for flat impedance to shape the top end response a bit.
Measurements with a radioshack spl meter show a good third order acoustic rolloff, but l have a slight loss of level between say 1.25khz and 2khz. It looks from your FR plot that you may have the same thing?

Bob, I can generously offer a free storage facility for any excess quality amp channels you have:D
 
BobEllis said:
If you want to get really fussy, you might try a notch filter on the tweeter to tame that small peak in response at around 3-4K. I plan to try it on my next version of the crossover. Not that I hear a need, but the Moxlite boards have eq in the tweeter section so why not?

I'm not sure about the original TC120tdx (if that is what you have float), but the tdx2 like mine does not have this peak (at all), it's quite a different beast to the newer TD5. Its response is smoother and is about 3dB more efficient in the usable range, however it starts rolling off a lot sooner at the low-end and has a higher Fs. A 2.4kHz crossover would barely be 1 octave above resonance for the tdx2 but the it does have a far smaller resonance peak than the TD5, its impedance is almost flat across the board, so it might sound just fine. Only one way to find out! :)

vuki said:
And, of course, you can't use Andromeda crossover because it's intended for a 3-way so you would have very bright sounding speaker without help of the bass driver. ;)

Apologies! :eek: I hadn't realised Mr Gee had tweaked them that much, the Andromeda used to be a pure 2-way with sub.

I never finished my speakers with the TC120tdx2, but I think I'll try some tweaking as well, sadly via the arduous non-active route. One more thing to add to the list of future projects!

Just done a quick search for the parameters of the TC120tdx and its Fs is even lower than the TD5, just under 700Hz. Couldn't find any response plots though, but it seems it may be as different again as the tdx2 and TD5 are from each other. You'll just have to play around float, but a low cross-over point is definitely back on the cards and will flatter the woofer too.
 
Here is the fr measurement of ss18w8545 at 1m in his box. One can see a large and broad peak from ca. 600-1100Hz. It's this driver's feature and you can see it in almost every commercial speaker using it. It ads to some of midrange "warmth and presence" unique to that driver. To get rid of it (that's - if you want to) you have to use more complex filters than simple textbook crossover filters. Otherwise (to get rid of it) you either have to use larger inductor and than have sudden drop after 1,1kHz ( float ) or use smaller inductor and live with it and have somewhat agressive midrange/highs.
 

Attachments

  • rawbas.jpg
    rawbas.jpg
    50.7 KB · Views: 393
Noksukau said:



Just done a quick search for the parameters of the TC120tdx and its Fs is even lower than the TD5, just under 700Hz. Couldn't find any response plots though, but it seems it may be as different again as the tdx2 and TD5 are from each other. You'll just have to play around float, but a low cross-over point is definitely back on the cards and will flatter the woofer too.


Resonance on the tc120tdx is quoted as 694hz with a Zmax of 25 ohms. I wonder if anyone considers the successive tc120 models to be improved? It seems that the parameters have changed for the worse from tdx - tdx2 - td5.

I have used a simple second order of 4.7uF cap and 720mH coil (and L-pad of 5.6ohm series, 4.1ohm parallel) and this gets -6db down at 2500hz and 34db down at tweeter resonance; the shape of the rolloff isn't textbook-perfect, but it's not that far off.

The aspect of the sound that troubles me most is a slight lack of solidity in voices - e.g. Norah and Diana Krall sound great, but theres a hint of wispiness or phasiness that prevents you believing that they're in the room with you.
Its that rock-solid image that I want.
Maybe I'm expecting too much of the system- cd63 ki-sig and audiolab 8000s power with a shunt-passive pre.:cannotbe:
Or is the lack of energy (maybe 3db) from 1250hz to 2000hz a likely culprit?
 
higher crossover pt?

this might sound crazy, but i think the 2-3 khz area is better done by woofers than tweeters, no matter how expensive tha latter. i've got the scanspeak 8" 8554 kevlar and 2905/9700 crossed over somewhere between 3 and 4k witha simple 2nd order and it's better this way than lower, but this is just preference.

carbon is way more critical here.

i would be most concerned about the 3k peak on this woofer - look at the time delay graph if you can find it but if it's not too nasty give this suggestion a try.
 

tktran

Disabled Account
2003-03-17 4:30 am
Perth
kram0.com
the 9700 is one of the very best tweeters, but measurements show that suffers from higher distortion below 2Khz, particularly horrible at 1.5Khz:

http://206.13.113.199/ncdiyaudio/mark/9x00_comparison/scan_speak_9300_and_9700.htm

This seems consistent with your observations.
The 9700 may measure flat down to 800Hz, but will sound best crossed steeply (eg. acoustic LR4 @ 2+Khz) or higher if using shallow slopes.