I agree with sreten, everything about the tweeter looks like a SEAS unit, from the hexagrid cover to the face plate to the way the magnet and chamber look.
From what I recall, SEAS appear to have made drivers for MA for a long time anyway, so this isn't a surprise.
If SEAS did manufacture this unit then I think it goes without saying that they are of decent quality and as it also happens, SEAS prestige range of chambered tweeters (what this looks like) tend to be excellent performers. One of the things these tweeters are good at doing is crossing over low ~2kHz etc.
Now metal coned mid and mid/bass units tend to suffer from large breakup modes that are usually between around 5-10kHz and are usually closer to 5kHz for 6.5" units. This breakup tends to put peaks into the harmonic distortion at multiples below the original breakup mode. In other words if you've got a peak at 6kHz, you'll have a corresponding peak in the third harmonic @ 6/3 = 2kHz. How bad this peak looks depends on how linear the motor is and how severe the breakup is.
As far as I am aware the MA has the tweeter crossover at 2.7kHz, this would not be (on the whole) low enough to attenuate this peak and it could be contributing to the more lively sounding treble. The same could be true if MA haven't attenuated the large peak in the midranges response by enough. They have however used a notch filter on the midrange and with one of the small air core inductors, so I would expect this to be used for combating this peak. It might not, but it's a good enough guess.
On the whole the crossover components used are of decently high quality and MA have elected to use the right type of component for the right job. Nothing really needs changing.
As sreten says you can add a resistor to the upper part of the design, to pad down the tweeter and upper mid/bass. This will increase the apparent level of baffle step compensation and make the sound 'warmer' it could also make it too bass heavy. You could try padding down the upper mid/bass on its own, or just the tweeter on its own and see what happens. In any of these cases you only want to use small 0.5-2R resistors and you want the resistors to be placed on the side of the amplifier, rather then the driver.
To really get to the bottom of this however would require that you measure the distortion and frequency response of the loudspeakers and see if anything sticks out as being responsible for your dislike in the way they sound.
Without know anything in detail about how the drivers measure I would say that the xover of 2.7kHz is too high as to be optimal, but may have been chosen on purpose to better protect the tweeter given some peoples tendencies towards turning it up too loud.
From what I recall, SEAS appear to have made drivers for MA for a long time anyway, so this isn't a surprise.
If SEAS did manufacture this unit then I think it goes without saying that they are of decent quality and as it also happens, SEAS prestige range of chambered tweeters (what this looks like) tend to be excellent performers. One of the things these tweeters are good at doing is crossing over low ~2kHz etc.
Now metal coned mid and mid/bass units tend to suffer from large breakup modes that are usually between around 5-10kHz and are usually closer to 5kHz for 6.5" units. This breakup tends to put peaks into the harmonic distortion at multiples below the original breakup mode. In other words if you've got a peak at 6kHz, you'll have a corresponding peak in the third harmonic @ 6/3 = 2kHz. How bad this peak looks depends on how linear the motor is and how severe the breakup is.
As far as I am aware the MA has the tweeter crossover at 2.7kHz, this would not be (on the whole) low enough to attenuate this peak and it could be contributing to the more lively sounding treble. The same could be true if MA haven't attenuated the large peak in the midranges response by enough. They have however used a notch filter on the midrange and with one of the small air core inductors, so I would expect this to be used for combating this peak. It might not, but it's a good enough guess.
On the whole the crossover components used are of decently high quality and MA have elected to use the right type of component for the right job. Nothing really needs changing.
As sreten says you can add a resistor to the upper part of the design, to pad down the tweeter and upper mid/bass. This will increase the apparent level of baffle step compensation and make the sound 'warmer' it could also make it too bass heavy. You could try padding down the upper mid/bass on its own, or just the tweeter on its own and see what happens. In any of these cases you only want to use small 0.5-2R resistors and you want the resistors to be placed on the side of the amplifier, rather then the driver.
To really get to the bottom of this however would require that you measure the distortion and frequency response of the loudspeakers and see if anything sticks out as being responsible for your dislike in the way they sound.
Without know anything in detail about how the drivers measure I would say that the xover of 2.7kHz is too high as to be optimal, but may have been chosen on purpose to better protect the tweeter given some peoples tendencies towards turning it up too loud.
Hi,
Warmer is more bass heavy and too bass heavy is too warm.
Simple, reduce the series resistor added, if that is the case.
rgds, sreten.
Yes, but a lack of warmth might not be the problem here. The bass could very well end up being overpowering whilst the upper midrange/lower treble/treble issues still remain.
Another way of getting work done, without "checking" to much, is testing with two variable L-pads, one for mid and one for tweeter, replacing later with fixed resistors.
As was said if the problem is on the quality of the drivers it will not work that much, quality wise this is, behind having a little attenuation.
http://www.toutlehautparleur.com/attenuateurs-c-77_154.html
As was said if the problem is on the quality of the drivers it will not work that much, quality wise this is, behind having a little attenuation.
http://www.toutlehautparleur.com/attenuateurs-c-77_154.html
@roender
I own a pair of MA RS8 and i also found them to be too bright and thought about doing something about this. What i did was to move them to the bedroom that has no exposed walls. There is real wood furniture and thick drapes. The sound difference is astonishing. They sound very warm and the bass response is very much improved. I now believe that the brightness was caused by the reflections in the old room, that had three naked walls.
Also, i doubt you can improve anything in the crossovers with supposedly better components. I am sure that MA tweaked the crossovers considering the components they used. 'Better' components could make it worse.
I own a pair of MA RS8 and i also found them to be too bright and thought about doing something about this. What i did was to move them to the bedroom that has no exposed walls. There is real wood furniture and thick drapes. The sound difference is astonishing. They sound very warm and the bass response is very much improved. I now believe that the brightness was caused by the reflections in the old room, that had three naked walls.
Also, i doubt you can improve anything in the crossovers with supposedly better components. I am sure that MA tweaked the crossovers considering the components they used. 'Better' components could make it worse.
Last edited:
The GS20 speakers are good but a little, just a little, to bright... and this brightness somehow covers the mid bass.
What about getting felt rings for the tweeters?
Dampers: Felt Defraction Ring for Tweeters: Madisound Speaker Store
http://www.audiotweaks.com/tweaks/tweak_184.htm
Hi,
Fair enough, but then its simply the wrong approach and the "brightness"
issues don't exptend as far dow as described. Adding an external 1R is
very easy to try, and abandon if it doesn't work.
rgds, sreten.
This are the specs:
Frequency Response: 32 Hz - 43 KHz
Sensitivity (1W@1M): 89 dB
Nominal Impedance: 6 Ohms
Maximum S.P.L (Per pair in room): 111 dBA
Power Handling (RMS): 150 W
Recommended Amplifier Requirements (RMS): 40 - 150 W
Bass Alignment: Dual Chamber Bass Reflex with Differential Tuning (HiVe®II Technology)
Tweeter Crossover Frequency: 2.7 KHz
Bass Crossover Frequency: 250 Hz
Drive Unit Complement: 1 x 6” RST®II bass driver
1 x 6” RST®II bass / mid driver
1 x 1” (25 mm) gold dome C-CAM® tweeter
What I don't understand is how the 2.7Khz tweeter crossover freq can be blamed for brightness and why correct it at 2kHz will supposedly solve the problem.
L-Pad trick (1ohm//20ohm) didn't solve the problem ... I just try it and the mids are as well affected not only the highs :-(
Frequency Response: 32 Hz - 43 KHz
Sensitivity (1W@1M): 89 dB
Nominal Impedance: 6 Ohms
Maximum S.P.L (Per pair in room): 111 dBA
Power Handling (RMS): 150 W
Recommended Amplifier Requirements (RMS): 40 - 150 W
Bass Alignment: Dual Chamber Bass Reflex with Differential Tuning (HiVe®II Technology)
Tweeter Crossover Frequency: 2.7 KHz
Bass Crossover Frequency: 250 Hz
Drive Unit Complement: 1 x 6” RST®II bass driver
1 x 6” RST®II bass / mid driver
1 x 1” (25 mm) gold dome C-CAM® tweeter
What I don't understand is how the 2.7Khz tweeter crossover freq can be blamed for brightness and why correct it at 2kHz will supposedly solve the problem.
L-Pad trick (1ohm//20ohm) didn't solve the problem ... I just try it and the mids are as well affected not only the highs :-(
(Do) You mean that, your problem being the mids and high frequencies, you need a solution that addresses both...
[=check post #24]
You have to go back to the design of loudspeakers, choosing the (preferably good) drivers and designing the crossovers.
[Go back to #21: "Now metal coned mid and mid/bass units tend to suffer from large breakup modes "]
Last edited:
With metal coned drivers you tend to get what's called distortion amplification due to the resonances of the cone.
In other words the motor + soft parts at say 2kHz would normally produce a third harmonic @ 6kHz that's at a level of -50dB, this is a respectable figure. However this cone actually rings at 6kHz, with a peak of 15dB. So now the third harmonic @ -50dB is amplified by the ringing cone to a level of - 35dB, thus grossly increasing the distortion. Now the distortion is created by the presence of a 2kHz tone, to get around this, you need to limit the amount of signal getting to the driver @ 2kHz, simply cutting the signal at 6kHz will do nothing. 6kHz is way into treble territory so could potentially be adding some artificial brightness to the sound and it's produced by the upper mid/bass driver.
This driver will also have a peak in the fundamental at 6kHz and if listened to full range will sound very artificial. The filter on the driver will need to be steep enough or used with a notch to make sure that the 6kHz peak is attenuated by enough, otherwise this could be injecting false brightness.
In a nutshell, with metal coned drivers you need to crossover low enough so that the peak in the third harmonic is attenuated to a reasonable level, but you also need to make sure that the peak in the fundamental is also attenuated by an acceptable amount. Usually a steep low filter, to satisfy the first part will be enough to satisfy the second, but if a designer has neglected to satisfy the first, then the second could be causing a problem. As it stands with the component values given, the notch filter used on the mid/bass will more then likely be used to attenuate this peak.
This is not a good idea, the crossover will have been designed appropriately for the drivers. The stand mount version of this speaker, the GS10 was measured by stereophile and although not perfect, the crossover does a decent enough job for being as simple as it is.
Monitor Audio Gold Signature GS10 loudspeaker Measurements | Stereophile.com
First of all we can see that the response is reasonably flat, save for a peak in the top octave. This would be my first target for an overly birght balance.
Stereophile is also good enough to show us the individual driver responses with the filter in place.
Here we can see that the mid/bass clearly has a strong resonance with MA targeting the notch filter at just above 6kHz.
The design isn't bad but there is definitely room for improvement. One would need to be capable however to actually do this correctly.
In other words the motor + soft parts at say 2kHz would normally produce a third harmonic @ 6kHz that's at a level of -50dB, this is a respectable figure. However this cone actually rings at 6kHz, with a peak of 15dB. So now the third harmonic @ -50dB is amplified by the ringing cone to a level of - 35dB, thus grossly increasing the distortion. Now the distortion is created by the presence of a 2kHz tone, to get around this, you need to limit the amount of signal getting to the driver @ 2kHz, simply cutting the signal at 6kHz will do nothing. 6kHz is way into treble territory so could potentially be adding some artificial brightness to the sound and it's produced by the upper mid/bass driver.
This driver will also have a peak in the fundamental at 6kHz and if listened to full range will sound very artificial. The filter on the driver will need to be steep enough or used with a notch to make sure that the 6kHz peak is attenuated by enough, otherwise this could be injecting false brightness.
In a nutshell, with metal coned drivers you need to crossover low enough so that the peak in the third harmonic is attenuated to a reasonable level, but you also need to make sure that the peak in the fundamental is also attenuated by an acceptable amount. Usually a steep low filter, to satisfy the first part will be enough to satisfy the second, but if a designer has neglected to satisfy the first, then the second could be causing a problem. As it stands with the component values given, the notch filter used on the mid/bass will more then likely be used to attenuate this peak.
This is not a good idea, the crossover will have been designed appropriately for the drivers. The stand mount version of this speaker, the GS10 was measured by stereophile and although not perfect, the crossover does a decent enough job for being as simple as it is.
Monitor Audio Gold Signature GS10 loudspeaker Measurements | Stereophile.com
First of all we can see that the response is reasonably flat, save for a peak in the top octave. This would be my first target for an overly birght balance.
Stereophile is also good enough to show us the individual driver responses with the filter in place.
Here we can see that the mid/bass clearly has a strong resonance with MA targeting the notch filter at just above 6kHz.
The design isn't bad but there is definitely room for improvement. One would need to be capable however to actually do this correctly.
You can try replacing the current crossover components with better ones of the same values.
In stead of iron core inductors use air core inductors with the same inductance / Rdc as the originals. nice polypropolyne capacitors (don't go crazy on these), MOX resistors. Can give an improvement, but probably no huge differences....
In stead of iron core inductors use air core inductors with the same inductance / Rdc as the originals. nice polypropolyne capacitors (don't go crazy on these), MOX resistors. Can give an improvement, but probably no huge differences....
Indeed, the problems that roender is experiencing are unlikely to be solved by changing any of the components.
As a suggestion however I would recommend using your PC as a source and playing around a little with some EQ. Using the stereophile review as a baseline, you should be able to correct for that peak, or at least tone down the upper octave and see if it helps. If it does you know where the problem lies.
Then as sreten suggested you could buy some small resistors and try padding down each section, either individually or together and see if that helps.
If you do some of this and find out roughly where the problem is coming from, then we can offer other suggestions about how to permanently fix it, say for the upper tweeter peak for example.
Do you think that SEAS Excel T25CF002 with crossover borrowed from Thor will solve the problem?
E0011-06 T25CF002
E0011-06 T25CF002
No!
The SEAS tweeter you've already got is most likely, more then capable.
If you want to spend some money on this then my advice to you is to buy a microphone and mic preamp, then use some freeware, such as ARTA to take measurements of the loudspeakers. This doesn't have to be too expensive, the Behringer ECM8000 should be fine as a mic, as should any cheap mic pre with phantom power and without a rumble filter, be fine too. I know you're experienced in electronics design so you could perhaps even build your own mic pre
If you can take meaningful measurements of the drive units, without a crossover, in your loudspeakers, then there will be plenty of people here (myself included) who would be more then happy to assist you in designing a new crossover. I have written a short guide on making useful loudspeaker measurements when using ARTA too if you're interested.
"But ... I would really appreciate any suggestion which may lead to a better crossover
Yeah, that's why I gave you one.
Hi,
Nevermind your suggestions are just clueless conjecture and wrong.
They won't lead to a better crossover and are not useful in any sense.
rgds, sreten.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.