Hi all. Recently, I built speakers using the Thor/Odin crossover, but in a bass reflex. I’d like to try bringing up the tweeter level just a little bit to see how it sounds. In the tweeter crossover, instead of just a resistor in parallel for the attenuation (L-pad), there is a resistor and a capacitor in series, and that circuit is in parallel.
My question: how do I calculate the appropriate resistor values to bring the tweeter level up properly? Since the capacitor has ESR, I would guess that I need to factor that in to the resistance of the parallel resistor, but I’m not too sure. Perhaps the cap gets changed as well? Can anyone provide some input for the crossover challenged? Thanks in advance.
My question: how do I calculate the appropriate resistor values to bring the tweeter level up properly? Since the capacitor has ESR, I would guess that I need to factor that in to the resistance of the parallel resistor, but I’m not too sure. Perhaps the cap gets changed as well? Can anyone provide some input for the crossover challenged? Thanks in advance.
Hi,
I'm not exactly sure what that capacitors purpose is, seems superfluous.
What is does mean is that the c/o will see ~ 8R at lower frequencies
and ~ 6R at higher frequencies for the tweeter, strange arrangement.
You'd have to model it in a simulator but it seems to me C4 and R3
do not form an L-pad with R2, its some form of high frequency zobel,
which will also prevent the response rising a little towards the top.
(An normal L-pad does this so why the capacitor ?)
Which means its not exactly easy to increase the tweeter level,
without changing the impedance loading on the crossover. You
could try increasing (x2?) the value of R3, this should bring up
extreme treble a little without affecting upper mid levels.
🙂/sreten.
I'm not exactly sure what that capacitors purpose is, seems superfluous.
What is does mean is that the c/o will see ~ 8R at lower frequencies
and ~ 6R at higher frequencies for the tweeter, strange arrangement.
You'd have to model it in a simulator but it seems to me C4 and R3
do not form an L-pad with R2, its some form of high frequency zobel,
which will also prevent the response rising a little towards the top.
(An normal L-pad does this so why the capacitor ?)
Which means its not exactly easy to increase the tweeter level,
without changing the impedance loading on the crossover. You
could try increasing (x2?) the value of R3, this should bring up
extreme treble a little without affecting upper mid levels.
🙂/sreten.
sreten said:
Thanks sreten.
Maybe I’ll try eliminating it and using a standard L-pad to play with the level (?). I have a spreadsheet crossover tool that I’ll play around with to see what it thinks will happen if I change various component values.
You mention that a normal L-pad keeps the response from rising up top… could you please elaborate on why that happens? My guess is that the tweeter’s impedance increases with frequency, while the impedance of the pad is flat, so the pad eats up more power as frequency increases. Is that even close? I’ve got so much to learn. 🙂
Hi,
the problem with a standard Lpad is your going to reduce the
impedance seen by the crossover somwhat, as it 3rd order it
could upset things somewhat.
Download TinaTI (free easy to use circuit simulator) and model
the crossover for the tweeter to see whats will be going on.
The parallel part of an L-pad reduces the effect of any rise in
the drivers impedance, either due to fundamental resonance
or inductance - C4 seems to allow it to rise towards 500Hz.
🙂/sreten.
the problem with a standard Lpad is your going to reduce the
impedance seen by the crossover somwhat, as it 3rd order it
could upset things somewhat.
Download TinaTI (free easy to use circuit simulator) and model
the crossover for the tweeter to see whats will be going on.
The parallel part of an L-pad reduces the effect of any rise in
the drivers impedance, either due to fundamental resonance
or inductance - C4 seems to allow it to rise towards 500Hz.
🙂/sreten.
sreten said:
I modeled this, and it simulated quite flat, once I adjusted for the fact that the Seas curves assume 1w and not 2.83v... I was surprised the first time I ran the numbers! It looks like with some changes to the last parallel leg of the circuit (the cap and resistor), it will add some top end, just as you suggested, which is what I want. At least, what I think I want.
For the record, these speakers sound great. I'm really nit-picking about the tweeter thing. Don't take it to mean that I am not happy with the setup. The midband is fantastic, and they really pump out strong bass for 7" woofers. The box is 2.0 ft^3 tuned to about 30-35Hz. I have some Scan-Speak monitors with the 18w-8546 (7" kevlar) and 2010 tweet (3/4" soft dome), and Seas compares very well. Better bass for sure, no surprise there. The Seas midband and a bit 'warmer' but with all the detail, and the treble is silky smooth, although as I mentioned, I think they could use just a little more up top. It could be my room or my equipment (all NAD), so take it for what it's worth.
Thanks again for all the help, sreten.
Hi mathman,
I'm thinking of building this design (well a variant of it - probably the Small Thor MLTL enclosure), but have some crossover / enclosure questions for you.
- I'm interested in whether you used the original (current Seas website) published crossover or whether you've tweaked other parts as suggested elsewhere on this forum?
- Did you also maintain the baffle dimensions?
- Did you consider the TL version? (or was space an issue?)
Cheers,
David.
I'm thinking of building this design (well a variant of it - probably the Small Thor MLTL enclosure), but have some crossover / enclosure questions for you.
- I'm interested in whether you used the original (current Seas website) published crossover or whether you've tweaked other parts as suggested elsewhere on this forum?
- Did you also maintain the baffle dimensions?
- Did you consider the TL version? (or was space an issue?)
Cheers,
David.
- Status
- Not open for further replies.