Hey guys, I'm looking for some advice here. I'm new at in depth tinkering with passive crossovers but I have some large three way speakers with rising tweeter response and they're killing me from 9kHz and up. In my opinion it's too severe for moving the speakers to try and fix it. It helps a little bit, but not enough.
I didn't build or design the speakers but there is a 2ohm resistor used in series before the crossover to attenuate the tweeter. I've adjusted this quite a bit and when it feels right in the lower treble, the rising response is really taking my head off.
My question is what is the next logical step in fixing this? Do I get rid of the series resistor and try an L-Pad instead? Or do I leave the series resistor and build an RL circuit to compensate? The tweeter being used is 6ohms.
I didn't build or design the speakers but there is a 2ohm resistor used in series before the crossover to attenuate the tweeter. I've adjusted this quite a bit and when it feels right in the lower treble, the rising response is really taking my head off.
My question is what is the next logical step in fixing this? Do I get rid of the series resistor and try an L-Pad instead? Or do I leave the series resistor and build an RL circuit to compensate? The tweeter being used is 6ohms.
How did you come to conclusion that tweeter output
beyond 9kHz is an issue ? Was is just by listening ?
beyond 9kHz is an issue ? Was is just by listening ?
Yeah, I tried that. I think it's a more in depth adjustment than that. It seems fairly well known that a series resistor can cause rising response, so I'm curious if trying an L-Pad instead of the series resistor first would be a good first step, or if leaving the series resistor and going to a RL circuit would be best.
Unless you provide a great deal of information about the crossover and the drivers of the system, it's difficult to provide any worthwhile advice. Frequency response measurement makes it easier to correctly identify a problem, where your subjective response might not be entirely correct. Can you contact whoever it is who did the design and build about your problem?
In most audio reproduction, there is very little intensity at frequencies above 10 kHz, in this range it's mostly low amplitude noises, clinking of keys, bow scrapings, etc. So assuming that reproduction is too bright, I would suggest that it is occurring in a frequency range starting much lower than 9 kHz.
If you remove or change the resistance of the series resistor, you are shifting the high-pass cut-off frequency of the tweeter's filter and messing up crossing over from the mid-range driver. So if you are going to try a L-pad, you should leave the 2 Ohm resistor in circuit.
Adding an RL circuit would be greatly altering the original design- I would say do that only if you have good objective reasons to do so.
My two cents,
cT
In most audio reproduction, there is very little intensity at frequencies above 10 kHz, in this range it's mostly low amplitude noises, clinking of keys, bow scrapings, etc. So assuming that reproduction is too bright, I would suggest that it is occurring in a frequency range starting much lower than 9 kHz.
If you remove or change the resistance of the series resistor, you are shifting the high-pass cut-off frequency of the tweeter's filter and messing up crossing over from the mid-range driver. So if you are going to try a L-pad, you should leave the 2 Ohm resistor in circuit.
Adding an RL circuit would be greatly altering the original design- I would say do that only if you have good objective reasons to do so.
My two cents,
cT
Thanks for your reply. It's definitely above 9kHz. I can shoot it with fuzz measure but I really don't need to. I can hear it.
The speaker designer used a resistor in series pre crossover as an adjustable feature to adjust tweeter level. The problem is that I'm getting a rising response issue. I realize this is going to be trial and error but I thought someone here had maybe had a similar issue and a preference for an L Pad or filter network.
The speaker designer used a resistor in series pre crossover as an adjustable feature to adjust tweeter level. The problem is that I'm getting a rising response issue. I realize this is going to be trial and error but I thought someone here had maybe had a similar issue and a preference for an L Pad or filter network.
The only way it is possible to adjust the level of the tweeter without shifting the crossover frequency (of the tweeter) is with a L-pad. There is no such thing as a "pre- crossover" series resistor.
-cT
Yes, I understand. That's how these were designed. I'm not saying it's right or wrong, that's just how it is in this case. It seems to be a very common first step in adjusting tweeter level in a lot of designs.
Assuming the tweeter SPL is correct, for the rising response you can use an RC in parallel to the tweeter after the current crossover. Try some values, 1.5-2.2uF and 1-3R.
Ralf
Ralf
What he is referring to is usage of a resistor between the amplifier and dividing network portion of the xover. Yes the resistor is part of the xover network, but being to one side or the other of the dividing network portion has differing affects on the outcome.
If you place a resistor before, there will likely be no added tilt to the response. If you place it after (between driver and network), the response will tilt downward with increasing frequency compared to the other placement, and shift the Fc of the xover if it is a first order type. If the xover is of higher order the Q of the xover will change, but the Fc will not.
An L-pad is usually defined as having a series and parallel resistor to form the "L" in the circuit drawing. You can have the series resistor before or after, and the parallel can also be before (with the series also before, or issues happen) or across the driver, and attenuation will still happen.
Yes you can have a resistor before and use that single resistor as an attenuation resistor and be very effective vs using a conventional L-pad after the network.
Later,
Wolf
Good advice. Be also advised that the resistance level in the CR will be in direct parallel with the driver used above the Fc of the cap used. If there is no series resistance before the CR and driver, then you could be causing a very low impedance issue with a 1-3 ohm resistor value.
Just a thought,
Wolf
series crossover ?
... maybe you can, maybe you can'tbut mostly very very difficult to hear where it comes from, or what the cause may be
and that makes you believe it is what you hear ?
basicly the tweeter may just be playing too loud, and thats all
but many times the problems that might be heard as tweeter issues could very well be casued by other crossover problems
impossible to say
well, maybe it is, maybe it isn't ... I would doubt it 😉
what tweeter are we talking about ?
Thanks for the input. That's exactly what I was looking for. Although I thought an RL (resistance and induction) in series would tame some rising response starting at a specific frequency? You're recommending an RC in parallel instead. I take it they both work but differently?
Last edited:
series crossover ?
but mostly very very difficult to hear where it comes from, or what the cause may be
and that makes you believe it is what you hear ?
basicly the tweeter may just be playing too loud, and thats all
but many times the problems that might be heard as tweeter issues could very well be casued by other crossover problems
impossible to say
What's up with the snarky comments? I'm looking for experienced input, not questions about what I can or cannot hear. I was actually having problems with the high frequency balance (rising at about 9kHz) and I googled rising response and found that the series resistor can cause it. The tweeter level is about right (for me) with a 2ohm resistor. A 2.5ohm is a bit too quiet from 3-7kHz.
I've got 2 tweeters I've tried, a SEAS t25cf001 and t25cf002 (Millenium). I actually prefer the less expensive 001 as the Millenium seems to exacerbate the problem with the extended response. There's something colored (harmonic distortion) that I don't like about the Millenium as well but a lot of people really love them. I've been using Scanspeak ring radiators for the past 10 years in my old speakers so this is a bit of a change.
I have used RC on a tweeter, with some positive effect
often a driver has some small impedance ressonances
might cause a little tendency to make the sound less refined
btw, the only snarky one is yourself 😉
have you tried any tweeter that did not have this problem ?
and ... you cannot just change a tweeter without crossover changes as well
often a driver has some small impedance ressonances
might cause a little tendency to make the sound less refined
btw, the only snarky one is yourself 😉
have you tried any tweeter that did not have this problem ?
and ... you cannot just change a tweeter without crossover changes as well
Yes there is. It's also known as input side attenuation. It's a resistor before the first capacitor, and equally attenuates the magnitude response of the driver.
I have not tried a tweeter in this particular speaker that didn't have this problem. The manufacturers recommendation is that you can change from the SEAS 001 to 002 with the only adjustment being the series resistor pre crossover. Again, not saying it's right or wrong. I didn't design it...
I'm looking for possible solutions to try first. Not looking for a lesson in what I'm hearing or technical lessons about the 'right' way to design a speaker. I'll know when the frequency response is right for me and I'm trying to get the last 10% out of my speakers.
have to tell you that any component you throw into the mix is likely to affect phase
sometimes it will have the wanted positive effect
but it always holds some negative ones
the only thing certain with speaker problems, there is no miracle cure
only some rather poorly develloped ways to lessen them
good luck
sometimes it will have the wanted positive effect
but it always holds some negative ones
the only thing certain with speaker problems, there is no miracle cure
only some rather poorly develloped ways to lessen them
good luck
- Status
- Not open for further replies.