I know how L-pad works.
But no one sees that L-pad in series crossover (to reduce level to the tweeter, I am ok there) is in parallel with inductor, thus making shelving filter for woofer...
I feel like repeating myself. Peace!
But no one sees that L-pad in series crossover (to reduce level to the tweeter, I am ok there) is in parallel with inductor, thus making shelving filter for woofer...
I feel like repeating myself. Peace!
And without the l-pad, the tweeter is in the exact same position with the exact same impedance, how is it any different?
Sent from my Nexus 5 using Tapatalk
Hi adason,
if you're looking to try another driver, the Audax PR170 wide range (mid) driver is back in production - there's a lot of rubbish written on the madisound site about them but a really good driver that works extremely well with current drive amps- not sure what of your tweeter arrays will marry up with them but if you want to go a bit crazy, there's a couple of the old rather good horn tweeters still available on Parts connection that'll fit quite well
... getting quite a bit of feedback about the crossovers!
if you're looking to try another driver, the Audax PR170 wide range (mid) driver is back in production - there's a lot of rubbish written on the madisound site about them but a really good driver that works extremely well with current drive amps- not sure what of your tweeter arrays will marry up with them but if you want to go a bit crazy, there's a couple of the old rather good horn tweeters still available on Parts connection that'll fit quite well
... getting quite a bit of feedback about the crossovers!
Hi,
Yes your repeating whatever it is you don't understand.
What you think you can see, and no-one else does, is wrong.
The L-pad makes no difference at all in basic simple theory.
In simple theory a tweeter + L-pad is the same as the tweeter.
The impedance across the inductor simply remains constant.
rgds, sreten.
Get the free TinaTi and sim it, you'll see no such effect.
Last edited:
Series versus Parallel? In a series crossover every single problem with the woofer impedance curve and drivers resonances shows up as a signal to the tweeter. This stinks like last weeks diapers. Series crossovers make mud out of the region around the crossover because the drivers are far from a perfect resistor. If you are going to use series crossover (YUK!) I strongly suggest placing a power resistor directly across each active driver. An 8 ohm big resistor across a nominal 8 ohm driver. This will greatly stabilize the impedance of the driver allowing the series crossover to half way do its job. Coils and other networks to stabilize impedance of the driver sort of fix the driver but are now in series with the tweeter and create a whole bunch more interactions so, this is not nearly as good an idea as it first sounds. Experience has shown in every case the parallel crossover outperforms the series crossover in every way both in listening test and laboratory measurements. The crazy notion the crossover frequency for the woofer should be identical to the crossover frequency for the tweeter is so dead the grave marker is missing but, this does not stop people from digging up this horrible idea and the horrible series crossover along with it. Both should die and go away forever!
Hi,
Strikes me you don't know when a series x/o is useful and when it isn't.
It has its rather limited uses, but can be useful for first order networks,
where unless you need offset x/o points, it is usually better than parallel.
Point is it deals with driver impedance variations much better than parallel.
An ideal series application is a 1st order FAST implementation.
rgds, sreten.
Thanks for the info jameshillj, but I am fond of faital pro midrange drivers...
* At high frequencies the tweeter//l-pad//inductor is 8 ohms and the capacitor shorts across the woofer reducing the voltage driving the woofer. At low frequencies the inductor shorts the l-pad//tweeter reducing tweeter drive while the capacitor//woofer stays around 8 ohms nominally.
2-way 1st order into XSim, zma-files are from real drivers SPH-250KE and TC9FD but cheat a bit link both drivers a frd-file with same BW2 40Hz - BW2 16kHz textbook bandwidth, it seems the two serial filters performs best in this simpler setup. Parallel gets better with resonance and inductance correction networks seen in second plot, XSim dxo file as zip folder attached.
Attachments
Last edited:
Wow BYRTT, that's nice work. So i may be wrong about l-pad, or the difference is miniscule. On the other hand your simulation does show better square wave using series crossover. That's what i plan to measure over the weekend.
Thanks.
Thanks.
Yes it seems when sticking to as simple a network as possible in only one capacitor and one inductor then serial wins, look forward hear what real world experience it brings you over the weekend.
And the fact that a tweeter is inductive is somehow not a problem when using an l-pad in a parallel crossover? Because it is...
I don't find one better or worse than the other. You can tend to use less copper in an SXO, whereas a parallel will require more copper in the inductors.
I agree that larger wire is better in both cases. You don't have to use a 1st order SXO to get benefits.
Also of note- you really can't do much SXO circuit work within a parallel xover, but you can do quite a bit of parallel style adaptations within a SXO. The whole 'you can't use difficult drivers' or 'you cannot notch a SXO' is rubbish.
As seen from my Xenoliths network above, you can do quite a bit of adapting to get your FR flat and phase aligned if that is your endgame.
Later,
Wolf
I agree that larger wire is better in both cases. You don't have to use a 1st order SXO to get benefits.
Also of note- you really can't do much SXO circuit work within a parallel xover, but you can do quite a bit of parallel style adaptations within a SXO. The whole 'you can't use difficult drivers' or 'you cannot notch a SXO' is rubbish.
As seen from my Xenoliths network above, you can do quite a bit of adapting to get your FR flat and phase aligned if that is your endgame.
Later,
Wolf
Sreten, yes series crossovers are useful when building electronic circuits using resistors and are completely useless for transducers. Parallel always is able to work better with the same component count. I understand crossovers extremely well having been doing speakers for 55 years and do not have dementia. All the wishful thoughts for series crossovers do not work in the real world with real drivers. I have proved this in the lab with testing and in listening test including hundreds of people. By far the proper parallel crossover always outperformed the series in bench testing and in listening. Messing around with series crossovers is a complete waste of time. I got better things to do than mess with notions which never work as well as something else.
Hi,
Like I said, you don't know when series x/o's are useful and when they are not.
You don't know how to use their interaction with drivers to your advantage.
I'll repeat unless you you want offset electrical x/o points 1st order series is
nearly always better than 1st order parallel, and the best choice for FAST's.
I admit higher order series doesn't make much sense, it gets too messy.
That is : 2nd order electrical parallel is always better than series, whatever
the acoustic target, and 3rd order electrical series just gets complicated.
I'm a big fan of series for first order electrical, but not much else.
rgds, sreten.
Last edited:
Hi all, i did simple experiment, simple 2way, exactly the same cap and coil used, once series, once parallel. I measured fr response and then fed it square wave.
Fr response close between both crossovers, but not identical.
Square wave so bad in both cases, no need to mention it. Small difference too, but nothing like theory says.
Fr response close between both crossovers, but not identical.
Square wave so bad in both cases, no need to mention it. Small difference too, but nothing like theory says.
I'm a bit surprised that you're still playing around with passive crossovers with your very wide driver options plus the different types of amplifiers (current drive & voltage drive g/clones)
Suggest you look towards electronic crossovers - they can avoid a number of problems you face with the passive filters and 'driver peculiarities', plus amplifier inequalities, or a combination of both - the filters still need to be designed for the specific driver characteristics but they provide a more direct connection between amp and driver and more flexibility in final result.
Suggest you look towards electronic crossovers - they can avoid a number of problems you face with the passive filters and 'driver peculiarities', plus amplifier inequalities, or a combination of both - the filters still need to be designed for the specific driver characteristics but they provide a more direct connection between amp and driver and more flexibility in final result.
- Status
- Not open for further replies.