I have a system with the XT-25's as the HF driver hoping to get at least a response flat to 25kHz plus... however when I measure with a radioshack SPL meter, the sinewaves show a gradual roll-off after 10kHz.
20kHz is about 13dB lower than 10kHz. A-weighted
Granted my measuring equipment could be better but this level of roll off is a little unnerving.
I am placing the SPL meter about 6" away from the tweeter, I have tried on axis as well as 30 degrees off and get same results.
Both Tweeters have the same issue.
The crossover is a 3rd order with a 4.7uF cap in series and 0.2mH coil in parallel and another 4.7uF in series with the XT-25. So basically cap/coil/cap then tweeter on the live terminal and nothing on the gnd terminal (except the other end of the coil).
Schem attached... apologies for the lousy MS paintjob
Anyone have any ideas?
20kHz is about 13dB lower than 10kHz. A-weighted
Granted my measuring equipment could be better but this level of roll off is a little unnerving.
I am placing the SPL meter about 6" away from the tweeter, I have tried on axis as well as 30 degrees off and get same results.
Both Tweeters have the same issue.
The crossover is a 3rd order with a 4.7uF cap in series and 0.2mH coil in parallel and another 4.7uF in series with the XT-25. So basically cap/coil/cap then tweeter on the live terminal and nothing on the gnd terminal (except the other end of the coil).
Schem attached... apologies for the lousy MS paintjob
Anyone have any ideas?
Attachments
A quick look at the manual for your radioshack SPL meter tells me that with an A weighting the meter is more responsive in the 500Hz - 10KHz range than other frequencies. You need to set the meter to a flat frequency response curve (although technically that's not a curve, I know).
It seems that your meter is only really of use up to 10KHz anyway. I wouldn't worry about your tweeter's response, I'd be more worried about the response of the meter at that frequency.
It seems that your meter is only really of use up to 10KHz anyway. I wouldn't worry about your tweeter's response, I'd be more worried about the response of the meter at that frequency.
You should use the C-weighting with a compensation chart.
It's not perfect, but it's better than what you're currently doing.
Here's the compensation chart :
http://forum.audiogon.com/cgi-bin/fr.pl?htech&983682086&openfrom&3&4
It's not perfect, but it's better than what you're currently doing.
Here's the compensation chart :
http://forum.audiogon.com/cgi-bin/fr.pl?htech&983682086&openfrom&3&4
Thanks Guys!
The thing is I took a second reading with C-weighting and the 10k to 20k diff was just as bad. There is no "Flat curve" setting on this meter.
I did have some metal film resistors in series with the Tweet which I though may have been inductive enough to roll off the highs, however being only 1.3ohms, I didnt think they'd make such a difference and removed them, the response is just the same.
I am using MIT Terminator2 cables... perhaps they have some roll-off in that black box of theirs attached to the Cable...
I am thinking that the meter is no good beyond 10kHz... but then why do the ears want more 16kHz?
Anyway I will also try C-weighting with the compensation curves (wonder what 1db compensation will do lol).
The thing is I took a second reading with C-weighting and the 10k to 20k diff was just as bad. There is no "Flat curve" setting on this meter.
I did have some metal film resistors in series with the Tweet which I though may have been inductive enough to roll off the highs, however being only 1.3ohms, I didnt think they'd make such a difference and removed them, the response is just the same.
I am using MIT Terminator2 cables... perhaps they have some roll-off in that black box of theirs attached to the Cable...
I am thinking that the meter is no good beyond 10kHz... but then why do the ears want more 16kHz?
Anyway I will also try C-weighting with the compensation curves (wonder what 1db compensation will do lol).
Network
K-amps,
The meter is suspect, to some extent, without equalization (I have one), but with a test disc and limited bandwidth frequency increments, you can spot general trends.
I also wonder at the component values you have shown in your crossover. These are not something I would recognize as a common 3rd order xover, with the second cap value the same as the first. Ordinarily the second cap is around 3X the value of the first in an 18 dB high pass circuit. In a 24 dB LR xover, this 2nd series capacitor is exactly double the first value (along with a second coil). If you have something near 15uf, I'd give that a try in this position, along with any padding resistors you will using, if needed to match levels with the woofer.
Just for testing purposes, you might want to try just a single 4.8 uf capacitor in series with the tweeter alone. Apply power carefully, however, as the tweeter has less protection in this simple 6dB circuit, and do not feed any low frequencies(below 1K Hz) to it.
Tim
K-amps,
The meter is suspect, to some extent, without equalization (I have one), but with a test disc and limited bandwidth frequency increments, you can spot general trends.
I also wonder at the component values you have shown in your crossover. These are not something I would recognize as a common 3rd order xover, with the second cap value the same as the first. Ordinarily the second cap is around 3X the value of the first in an 18 dB high pass circuit. In a 24 dB LR xover, this 2nd series capacitor is exactly double the first value (along with a second coil). If you have something near 15uf, I'd give that a try in this position, along with any padding resistors you will using, if needed to match levels with the woofer.
Just for testing purposes, you might want to try just a single 4.8 uf capacitor in series with the tweeter alone. Apply power carefully, however, as the tweeter has less protection in this simple 6dB circuit, and do not feed any low frequencies(below 1K Hz) to it.
Tim
Thanks Tim, some good info.
I actually also at one time had the arrangement at a simple 2nd order, i.e. cap/cap/coil wth the two 4.7uF caps being the same as a 2.35uF single, however Now I am not sure if I made readings while the caps were set in 2nd vs. 3rd order.
I will short out the second cap, making this a second order and re-test it.
I use a sinewave generator so I will not be using a full spectrum as such.
But that apart, the next issue I run into is, how do I determine polarity of individual drivers where the tweet is 3rd order, mids second order, midbass first order and Bass drivers that are actively crossed over...
I actually also at one time had the arrangement at a simple 2nd order, i.e. cap/cap/coil wth the two 4.7uF caps being the same as a 2.35uF single, however Now I am not sure if I made readings while the caps were set in 2nd vs. 3rd order.
I will short out the second cap, making this a second order and re-test it.
I use a sinewave generator so I will not be using a full spectrum as such.
But that apart, the next issue I run into is, how do I determine polarity of individual drivers where the tweet is 3rd order, mids second order, midbass first order and Bass drivers that are actively crossed over...
More on RS meter
K-amps,
I thought I'd kill two birds with one stone, here, by including both the fix for the meter, as well as a source for the Wallin jig used with the Speaker Workshop free program.
http://mysite.verizon.net/tammie_eric/audio/audio.html
Good luck!
Tim
PS Where's home?
K-amps,
I thought I'd kill two birds with one stone, here, by including both the fix for the meter, as well as a source for the Wallin jig used with the Speaker Workshop free program.
http://mysite.verizon.net/tammie_eric/audio/audio.html
Good luck!
Tim
PS Where's home?
Re: Crossover
LOL, yes it's been tough.
It started off as an Orion Variant with an MTM instead of the 8" + 1" on the Orion, I used dual Vifa P13's and an XT25. The bass was originally handled by dual PPI Pro 12 flat piston woofers.
However after realizing that the PPI's did'nt sound so good, I went with 4 stryke SA071's in the Midbass instead. These are much better suited to the Vifa's.
The sub-bass is handled by a pair of 15" JBL's actively crossed over.
Enter the wrench, as warm as the P13's sounded, I thought I was missing some low treble details.... so I added a TAD berllylium 4" driver.... ok a Pioneer DSS9 midrange that I had (but it looks like a TAD design) inserted between the XT25 and the P13's. It relieved the XT's from being crossed at 3kHz to now being crossed at more like 5-7kHz.
I have the system to now where it images much better, the soundstage is deep (and at times near, and at times even behind my ears ) and wide.
Hearing tests are now whre I prefer the system to my old trusty B&W 801 series 80's. However with the DSS9's mid treble driver padded down, I need to make sure that I got the polarities well set.
Tim Moorman said:
LOL, yes it's been tough.
It started off as an Orion Variant with an MTM instead of the 8" + 1" on the Orion, I used dual Vifa P13's and an XT25. The bass was originally handled by dual PPI Pro 12 flat piston woofers.
However after realizing that the PPI's did'nt sound so good, I went with 4 stryke SA071's in the Midbass instead. These are much better suited to the Vifa's.
The sub-bass is handled by a pair of 15" JBL's actively crossed over.
Enter the wrench, as warm as the P13's sounded, I thought I was missing some low treble details.... so I added a TAD berllylium 4" driver.... ok a Pioneer DSS9 midrange that I had (but it looks like a TAD design) inserted between the XT25 and the P13's. It relieved the XT's from being crossed at 3kHz to now being crossed at more like 5-7kHz.
I have the system to now where it images much better, the soundstage is deep (and at times near, and at times even behind my ears
) and wide. Hearing tests are now whre I prefer the system to my old trusty B&W 801 series 80's. However with the DSS9's mid treble driver padded down, I need to make sure that I got the polarities well set.
Difficulty Rating: 9.5 out of 10
K-amps,
Fort Wayne. Not tooo far.
Jeez, that is quite a full plate. Kinda scary in its complexity, and trying to get it all integrated, but I admire the effort.
Frankly, I prefer high efficiency designs, but still have all manner of speaker gear around here. Some I've passed on to my sons (old fart).
I've got a few amps around that need rebuilding. I'll send you an
email.
Tim
K-amps,
Fort Wayne. Not tooo far.
Jeez, that is quite a full plate. Kinda scary in its complexity, and trying to get it all integrated, but I admire the effort.
Frankly, I prefer high efficiency designs, but still have all manner of speaker gear around here. Some I've passed on to my sons (old fart).
I've got a few amps around that need rebuilding. I'll send you an
email.
Tim
I like high efficiency designs too but feel I am not that experienced in designing those. Last year I tried to build a system around 2 Altec 515G's mated to a 288-8k compression driver.
As impressive as those drive units were, I could not get them to image like 2 cents worth. I still feel the 515 has few peers in terms of impact... but I guess it is not as all round as the system I have now.
Awaiting your email,
Cheers,
As impressive as those drive units were, I could not get them to image like 2 cents worth. I still feel the 515 has few peers in terms of impact... but I guess it is not as all round as the system I have now.
Awaiting your email,
Cheers,
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.