Crossing over Philips 1" dome

Status
Not open for further replies.
greenhorn

greenhorn

Enclave Plus
Juggernaut
I'm making a speaker as sort of an extra curricular activity :P

so far i have a pair of 'Dainty' 6" Poly cone woofers ( I'll be losing them soon, just waiting to get hold of some good paper cone drivers)

and a pair of philips 1" textile dome tweeters. The tweeters are completely dominating the sound, and something needs to be done about it. from what i heard, they go down till about 1K, and are about 94 dB sensitive. Should i move the crossover frequency high up, or use some resistors to bring down the output?

I know there are better places to ask this , but i think there are folks here who can answer this:)

check out my rig moviemachine to see the speaker and the drivers :)

for now , i've tweaked the eq in my pc to take care of it , but something needs to be done about it....
 
I think you need to bring down the tweeter's level. But I'm quite a noob in this, so don't go by my word :P
 
I presume that is a first order crossover, try using this calculator to set your capacitance and inductance values.

But for your case try the resistor method first with a variable resistance. Try adjusting the knob and if you hit a sweet spot then you could always shift to a static resistor combination :)

All the best!
 
I presume that is a first order crossover, try using this calculator to set your capacitance and inductance values.

But for your case try the resistor method first with a variable resistance. Try adjusting the knob and if you hit a sweet spot then you could always shift to a static resistor combination :)

All the best!
 
thanks anish, thats what i plan to do if all else fails . but its a commonly used tweeter (default recommended tweeter for all indian DIY efforts, so someone here might have had some experience with it , so i could straight to making the crossover without playing around with it ( not much time these days... how i wish i were in college again :( :( :( )
 
Oh man, those are some memories.

The faceplate on the tweeter is square?

I would use a 3rd order electrical, as there is a nasty distortion peak somewhere. I don't remember the values, I think it had a 10uF or 4.7 uF and 18 uF cap and a .4mH coil, plus an L-pad. This was about 10 years ago, so forgive the memory loss.

I have no idea what the response of that woofer is like.

That tweeter does NOT work well with a cap in series only, and it's probably got a nasty impedance peak at the bottom as it has no ferrofluid in the gap.

Philips had a book with the exact crossover values for their various woofers (which mate pretty well with it, BTW) IIRC, including the 8" and 6". Don't remember any of it.

Are you using something on the woofer or is it a direct connect across the signal? You will never get good sound like that. The correct way is to take some basic measurements on baffles and then design the crossover. Any other way is a sure path to failure.

See this guy, he did it in India, with some very basic stuff.

The Asawari | dhandanought.org/tcpip
 
I've always argued against using those calculators. They assume a flat resistive load, which a speaker is not. Actually, it's not a minor error, you will get major deviations and will have an unusable crossover. This calculation is how the 'crossover in a box' solutions are created. As you will agree those sound horrible.

The common simulators use mathematics, and unless you're 100% sure of your speaker parameters you can't use them. Remember speakers have very wide tolerances and can be up to 20% off spec in either direction, so measurement of individual units is the only waay to assure consistent results.

You're lucky it's Philips, those guys at least have some standard crossover design recommendations, which can be a starting point.

Unless you're doing it for giggles, in which case a calculator will work fine.

To answer your original question, the answer is really in raw driver response. This is going to be a ghetto method, but sometimes works OK. Connect the woofer to a pink noise source and a white noise source, and try and listen for the sound spectrum.

Sometimes woofers will exhibit a nasty upper breakup node that sounds like extra treble. This is correctable via notch filters in the crossover. Some of the best woofers in the world also exhibit this breakup node, but the world's best crossover designers manage to tame it. Us folks have to manage otherwise.

Poly cone woofers will not have this breakup, by and large. Paper cone woofers will, though, so if you change out the woofers the crossover will have to change.

The tweeter can be corrected through an L-pad. You can use a small 5 watt wirewound potentiometer (you get them in 50 ohm and 100 ohm values) to check what the L-pad ratio should be for your preferred listening habits, then get some higher wattage fixed resistors to seal it.

I still maintain you need to measure or at least have the stock specs with you when undertaking an exercise like this one.
 
right now , I'm thinking about a cap in series followed by a potential divider to tone it down somewhat. I'm not planning to go beyond first order. On the other hand , my bigger , made to order floorstanders use the same tweeters and sound quite nice. maybe I'll take them apart and see how they're crossed over

@ sangram, yup, the same one with the square face plate

EDIT: oh, an L pad is a potential divider... Nice :D
 
You don't even have a cap in series now? Your tweeters may already be damaged from excursion. Or not, if you're lucky and not feeding them too much power. Crossover are to properly divide the signal and to ensure the final product has a flat frequency response. For any changes to one driver, you will need to make a corresponding change in the crossover.

The tower crossover will definitely not work correctly with your driver combination, if the drivers are different. Then there is BSC, which I'm not sure is implemented in the towers, but is required to make the speakers work correctly in room conditions.

Read up a bit. It's a bit like trying to make a car with a chassis from one vehicle and an engine from the other. It may work, but you need to consider a lot of things pretty carefully, and have to know exactly what you're doing.

Oh yeah, it's a blast, too, in more ways than one.
 
Status
Not open for further replies.
Top Bottom