|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
72.188.12.218
I must be doing something wrong with my calculations, so I'm in need of some help from our more diy-oriented members. I have a new pair of Fostex T900a super-tweeters coming sometime next week and most everyone I talk to says I should use a .47uf cap with these tweeters. However whenever I use the online calculators, if I want to use a first order (6dB/octave) crossover on an 8 ohm tweeter and cross it in at 8Khz, it says I should use a 2.48µF cap!
Just for kicks I kept raising the crossover frequency over & over until I arrived at the frequency these online crossover calculators said would require using a .47uf cap and according to these calculators a first order crossover using an 8 ohm tweeter and a .47uf cap would crossover at 42Khz!
Please tell me what the heck I'm doing wrong!!??!! I'll be running the FE208ES-R fullrange and I'll want the T900a somewhere to come in between 8Khz and 10Khz but I'll need to listen and play around to see what sounds best, but I don't even know what value of caps to start with anymore after using these online crossovers...
Listening to Larger Than Life by Skywalk
Thetubeguy1954 (Tom Scata )
Full-range/Wide-range Drivers --- Front & Back-Loaded Horns
Central Florida Audio Society -- SETriodes Group -- Space Coast Audio Society
=================================================================================
"The man that hath no music in himself, nor is not moved with the concord of sweet sounds, is fit for treasons, deceptions, and spoils;
The motions of his spirit are dull as night and his affections dark as Hell. Let no such man be trusted."
- William Shakespeare from The Merchant of Venice; Act V,i
Follow Ups:
My situation is somewhat similar. I'm combining Jensen RP-302 tweeters to Altec 756Bs. There is an approx. 3 - 6db sensitivity mis-match between these drivers.About 6 or 7 years ago I started with Mundorf Silver/Oil .68uf caps and it sounded fine. A few years later I switched to Mundorf Gold/Silver/Oil .56uf caps and it sounded even better. About a year ago I switched again, this time to V-CAP TFTF .47uf caps. This sounds the best.
The RP-302s are 16 ohm, so with the .47uf caps, that puts the roll-off frequency at approx. 21,200hz. This was intentional to compensate for the sensitivity mis-match, but is it wrong? The high frequency response of this system sounds very good to me and everyone else who has heard it.
Gerry
Edits: 02/04/12
Gerry
If it sounds right, then it IS right. However there's no reason to feel 100% complacent until you push the concept to the point where it begins to sound worse. So you could try the next value 0.33, or 0.22, 0.15 or even 0.10 to find the point where things get worse, but you will run out of values with the Mundorfs at 0.10 uF., and hopefully you will have verified the sweet spot before this. Then again a feeling of doubt may arise that a conventional 2 way crossover may be better, or maybe even not as good. To investigate that route you will need to look at the freq. response of the 2 drivers alone without any crossover, and select a crossover point and slope that makes use of the best of what each of the drivers have to offer, while avoiding areas where they are not so good. In the event that you go back to the single "full range" driver mated with a tweeter with a single cap crossed at 6 dB per octave up beyond the normal hearing range, there will be some over lap where both drivers are making the same frequencies for a few octaves at least. Along with this there will be opportunities for phase cancellations and resonances where the distance between the drivers corresponds to some wavelengths or fractions thereof. Also, a lot of luck will be involved in finding the sweet spot where the attenuation of the tweeter is right, and where the 6 dB per octave attenuation of the cap does something useful in mating the tweeter performance and the full ranger together. In addition this also depends on good behavior of the tweeter near its Fs, and that the filter the single cap forms will also be able to deal with (or not need to deal with) any resonant peak there. Too much of a roll of the dice for me, but if it works, then it's right.
Paul
I would wager the HF of that driver is rolling at at least 6dB/oct in the top octave anyway, so you are just compensating for that.
Hello DJK and everyone else who's trying to help me. Above this post is the frequency response curves of both the Fostex FE208ES-R wide-range eight in driver and the Fostex T900a tweeter/super-tweeter. I'm still attempting to digest and understand what's been said in the various responses to my original request for help with a crossover. If I've learned anything from reading all the responses, I believe, when looking at these two charts the FE208ES-R at 100dB at 10Khz and the T900a is at approx 105dB at 10Khz.
Therefore if I decide run the 8 ohm, 99dB, FE208ES-R "fullrange" sans crossover, allowing this driver to crossover mechanically. I should use a first order (6dB/oct) crossover at 20Khz ---{ actually 19Khz when using a 1µF cap }--- on the 8 ohm, 106dB, T900a Fostex tweeter. Doing this will cause the T900a, at one octave below the 20Khz crossover point ---{ at approx 10Khz }--- to be attenuated by 6dB. That in turn would now match the FE208ES-R's 100dB output at that same frequency!
Is my thinking correct on that? Have I chosen the correct XO frequency to use with these two Fostex drivers of different sensitivities? Finally have I chosen the correct value of cap to purchase until I can afford a Fostex R100t attenuating transformers? Or would you do something different?
TIA.
Not listening to anything yet.
Thetubeguy1954 (Tom Scata )
Full-range/Wide-range Drivers --- Front & Back-Loaded Horns
Central Florida Audio Society -- SETriodes Group -- Space Coast Audio Society
=================================================================================
"The man that hath no music in himself, nor is not moved with the concord of sweet sounds, is fit for treasons, deceptions, and spoils;
The motions of his spirit are dull as night and his affections dark as Hell. Let no such man be trusted."
- William Shakespeare from The Merchant of Venice; Act V,i
Here is the PDF from Fostex, They use a 1 uf cap to the 900 tweeter. goto
http://www.madisoundspeakerstore.com/approx-8-fullrange/fostex-fe208ez-8-full-range-sigma-series/
They have a drawing
Tom
It looks like it's time for some Cowboy Engineering: try it, see what happens. You can get good quality audiophile XO caps from MCM or Parts Express for not much money. In spite of the nay saying I have been doing regarding this, one thing which will let you pull it off is the very good impedance behavior of the T900 shown in the graph here. This is typical for a horn loaded driver. As a contrast, imagine trying the same trick on the FE208ES, that is to say using a series cap with the intention of attenuating the driver a few dB so that it can match a subwoofer you happen to have at hand. You will run into that rising impedance mountain that peaks at the 38 Hz Fs and goes way past 64 Ohms, in addition to the rising impedance toward the treble end due to voice coil inductance.
Anyway, let us know how it goes.
Paul
I concur with Paul's suggestions (I'm fortunate to live close by and he has always pointed me in the right direction), but I'd add one suggestion from my experience. Buy smaller cap values than you need and run them in parallel, building up to where you get good results rather than risk overshooting and getting a cap that's too large. I've found that mixing nice caps (Duelund, Mundorf, V-Caps) with less expensive caps carries the majority of the sound elements of the better caps. This of course is more valuable for midrange or mid-bass because those caps can get very expensive.
I added a picture of my 4-way passive crossover. You can see that everything is clipped together to make swapping components easier.
Romy was using similar values with his t350's . He was padding with the cap sise. Searching for his posts could be useful. I think it is a matter of padding down also I am guessing you are working on 8 ohm calculations which your tweeter may not be anywhere near at the given frequency. I love my auto formers that I bought from mass mutter of ebay. But I am sure you could have them wound locally. That ensures a even impedance to the amp I think. Which is really handy with 0 feedback amps. I suspect your tweeter then is less likely you change pitch with frequency but they might cause phase shifts. I linked the guy but he hasn't any listed currently. They where about $200 intact audio has some cool ones too which I am sure you know about. Auto formers make a lot of sense.
Don't forget you need to reduce the cap every time you move to a lower tap on the autoformer.
I havent. I must have just got lucky sofar. I have the mid filter before the auto former and the treble cap after. As Mass recomended.
That way is fine, no problem.
What works in the real world isn't necessarily what a textbook calculation says.
My guess is the capacitor is being used to "pad back" that very high efficiency supertweeter. I'm assuming the FE208ESR is somewhat lower in efficiency, and also extends north of 8 kHz on its own, so you want to a) pad back the T900a quite a bit, and b) blend it in gently with what the fullrange driver is already doing.
But Kloss will know more about this than me.
Duke
Me being a dealer makes you leery?? It gets worse... I'm a manufacturer too.
Tom
You're on the right track: 159,000/(XO freq. in Hz) X Driver Imp. in Ohms= 2.28 uf, and 2.48 uF is a standard value which is probably close enough. A 0.47uF cap would indeed give you a XO freq. of 42K Hz, and it's difficult to make any sense of this in the context of the system you propose.
Try it and see what you think. I would add an inductor to make it real 2 way crossover: L (inductor value in mlliHenrys)=159.155 X Driver Imp. in Ohms/XO freq. in Hz. Note the decimal point in 159.155 here, and the comma in 159,155 above. You could try it both ways as a parallel 6 dB crossover or as a series 6 dB crossover, the parts values will be the same. I'd go series, but that's just my take on it.
BTW do you hear much from Paul Butterfield down Orlando way? Tell him we miss him here!
Paul E
Paul,
Thanks for your input. I'm sorry to say I don't hear much from Paul Butterfield very much at all these days. I know he's having health issues and like you, I too, miss Paul very much. I'll be sure to let him know you inquired about him Paul by providing a link to your post in an email I'll be sending him and if you'll email me privately at thetubeguy1954@yahoo.com I'll be happy to provide you with Paul B's email address, ok?
Listening to Points of View by Nando Lauria.
Thetubeguy1954 (Tom Scata )
Full-range/Wide-range Drivers --- Front & Back-Loaded Horns
Central Florida Audio Society -- SETriodes Group -- Space Coast Audio Society
=================================================================================
"The man that hath no music in himself, nor is not moved with the concord of sweet sounds, is fit for treasons, deceptions, and spoils;
The motions of his spirit are dull as night and his affections dark as Hell. Let no such man be trusted."
- William Shakespeare from The Merchant of Venice; Act V,i
fe208esr has a bit more response and a simple caps not really 6db slope so if you crossover lower you get a bit to much energy. One can use a 2nd or 4th order. But to me the t900a 1.uf r100t combo is the best match to fe208esr.
Kloss
That's a good point that the crossover freq. has to be chosen carefully with a 6 dB slope, with the response of the drivers dictating the freq. even more so than with higher order slopes. With a 1.0 uF cap (assuming that the L-pad holds the impedance close to the drivers 8 Ohms) you will be crossing the T900A at around 20k Hz where there is the beginning of about a 10 dB per octave downward trend from the data sheet. A 6 db slope would fit with the roughly 10 dB downward response trend at that point of the FE208ES-R, but there is a lot of response going on with the T900A below 10K Hz for a 6 dB slope to lop off. A 12 dB crossover might be more appropriate, but that adds more parts into the picture. Tom's plan to cross the T900A 6 dB at 6K Hz is feasible, as there is a downward trend in the response shown on the data sheet for the FE208ES-R beginning at 5K, but again I would add an inductor to make it a 2 way crossover, to avoid the rough area above 9K Hz with the 208ES-R. This is just my take on it from looking at the spec sheets, however someone working with the actual drivers may wind up with different conclusions of course.
Paul
Works best with t900a- fe208esr combo. You will need to attenuate the t900a to match fe208esr.R100t is the best way.
Thanks John, but I just cannot afford to purchase a pair of the Fostex R100t transformers to go along with the T900a tweeters at this time. Am I understanding you correctly in that you're suggesting I use a Mundorf silver/gold 1.00uf 1200v ---{ as seen at the link below }--- to run the FE208ES-R fullrange and cross the T900a in at 20K, while also using a resistor to attenuate the T900a as needed? Or are you suggesting I should use just the cap and allow the T900a to also mechnically crossover as it will, just like the FE208ES-R does?Listening to Points of View by Nando Lauria.
Thetubeguy1954 (Tom Scata )Full-range/Wide-range Drivers --- Front & Back-Loaded Horns
Central Florida Audio Society -- SETriodes Group -- Space Coast Audio Society
=================================================================================
"The man that hath no music in himself, nor is not moved with the concord of sweet sounds, is fit for treasons, deceptions, and spoils;
The motions of his spirit are dull as night and his affections dark as Hell. Let no such man be trusted."- William Shakespeare from The Merchant of Venice; Act V,i
Edits: 02/05/12
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: