Not sure what level you are coming from (maybe you are an expert
. This is more advanced than the "textbook", cookbook, calculator, filter theory. Hard to explain without advanced circuit theory or SPICE models.
On the bass low pass filter (LPF)
There is a pole or 1st order filter formed with Re and L. This frequency can be set lower so that it corresponds with the baffle width. You know about the BSC right?
The C is then adjusted to form another pole further up towards the tweeter crossover. So electrically this is still a 2nd order LPF. Now depending on combining with drivers Le and acoustic response it can have up to a net 3rd or 4th order response.
. This is more advanced than the "textbook", cookbook, calculator, filter theory. Hard to explain without advanced circuit theory or SPICE models.On the bass low pass filter (LPF)
There is a pole or 1st order filter formed with Re and L. This frequency can be set lower so that it corresponds with the baffle width. You know about the BSC right?
The C is then adjusted to form another pole further up towards the tweeter crossover. So electrically this is still a 2nd order LPF. Now depending on combining with drivers Le and acoustic response it can have up to a net 3rd or 4th order response.
infinia said:Not sure what level you are coming from (maybe you are an expert
On the bass low pass filter (LPF)
There is a pole or 1st order filter formed with Re and L. This frequency can be set lower so that it corresponds with the baffle width. You know about the BSC right?
The C is then adjusted to form another pole further up towards the tweeter crossover. So electrically this is still a 2nd order LPF. Now depending on combining with drivers Le and acoustic response it can have up to a net 3rd or 4th order response.
Me ? An expert ? I past the 28 last years to do... anything but speakers
lol C'mon, I'm a beginner. I don't want to be ... don't know english term but i don't want to do like if i am better than you. I'm just frustrated about how complex and onerous it's supposed to be from some people when you open ordinary loudspeaker and see what it used on it. So small parts, with tole woofer and mini rikiki tweeter and very cheap cable. No reinforcement, low density wood material. And funny to see beautiful gold plated big plug to the back but small wire cable inside and no gold plated connector at all inside too No filtering. Just first or second order with little L-Pad. I make a living room table with old brass garbaged bed head and wood from box spring. Just cost some paint. I do my own router table with about 15$ of wood and 4$ of fixation screw. I do many cool project with recycling material and very low budget. It's a kind of motivation.
so understand kind of guy like me when he hear "hmm. buy 500$ measurement system with an active crossover, do a measurement box and be prepared to scrap some set of crossover parts hihihi
But I'm a hi-fi passionate fan too. I have Shanling STP-80 amp, Cambridge Audio DacMagic (the new one) DAC, good cables and i buy Monitor Audio RS8. I expense over 4 500$ for my sound system without count the stuff already own in the past... My first DIY project cost me over 1000$.
But now, i have a budget project in peace with my recycling passion one. Hey ! Some people listen music from 200$ Home Theater in a box system or 100$ mini stereo. Anyway, you know what i mean.
I think i'm searching of piercing the manufacturer crossover secret. Only one parts like inductor cost to us probably more than two complete crossover cost to many loudspeaker company !
infinia said:
BSC... Not sure.
Last week, i read a text about that.
If i'm right :
The "wall" where transducer is mounted create obstacle to wave and result to an reorientation of kinetic energy trough the front, in the listener direction. But if i undestand properly, when the wavelength is longer than panel (baffle) width, this phenomenon decrease until you get spheric shape wave dispersion in bass domain.
Drivers is tested on infinite baffle (wall ?) but when installed on real enclosure, the efficiency decrease in lower frequency. The degrees of attenuation and frequency where it's produced depend on baffle width.
The things i don't understand is what to do next ? ...
Do math and attenuate all the non affected frequencies ?
Someone talk about diffraction in box corner too. Is that possible some manufacturer create design with multiple little radius, angle etc. instead of just sharp 90 deg corner ? More quieter diffraction site instead of huge hearing one ?
Hmm... I see 16 000$ ProAc with absolutely sharp corner. --> Is there another way to reduce this phenomenon ?
Hoo.. my reply is long like romance book
Sorry about that
Study the expensive drivers, just because they provide more data (measured SPL response on and esp. off axis) and the T/S is more reliable. Look at the cone profile and materials for smooth HF shape. You can make some assumptions from thier data to similar cheap ones. Some cheap makers like you say have minimum or even no LF crossovers because they can rely on smooth natural rolloff of the driver. If you want to buy cheap midbass driver fine but measure the T/S parameter yourself then build the box.
IMO the best crossover matches the offaxis responses between the drivers. Since non horn tweeters have wide directivity at the low end the midbass should cross at or before the point where its offaxis deviates from its on axis responses. It's quite possible to do this with just a series inductor and a cap. But not the best for really cheap drivers or even not very good for expensive metal cone or fancy bullet pole pieces.
Keeps searching and reading on BSC. There is a shareware program called the edge. DL that program and understand what it tells you.
Some drivers response is measured on a smaller IEC type baffle and you can see the bass loss effects below a few hundred Hz.
Sorry but even the simple things can even be more complicated to do right or nothing is as easy as it seems, unless you know how.
IMO the best crossover matches the offaxis responses between the drivers. Since non horn tweeters have wide directivity at the low end the midbass should cross at or before the point where its offaxis deviates from its on axis responses. It's quite possible to do this with just a series inductor and a cap. But not the best for really cheap drivers or even not very good for expensive metal cone or fancy bullet pole pieces.
Keeps searching and reading on BSC. There is a shareware program called the edge. DL that program and understand what it tells you.
Some drivers response is measured on a smaller IEC type baffle and you can see the bass loss effects below a few hundred Hz.
Sorry but even the simple things can even be more complicated to do right or nothing is as easy as it seems, unless you know how.
Here is what you can do
1) 1st thing get an 8 ohm vifa tweeter the series cap will be half the size than for a 4 ohm saves$ then get data and response curves
2) model bass reflex encosure in Unibox or WinISD alpha
Get Vb and Fb find port size
3) pick box dimensions from 2) use golden mean
4) go to peerless / vifa web site get midbass response data
5) model BSC using edge & pick Inductor using 1st order calc
& calc loss due to Inductor
6) pick x over frq and order using 1,2,4&5
7) calc L pad using 1,4,&5
crossover s/b around 3ish KHz
1) 1st thing get an 8 ohm vifa tweeter the series cap will be half the size than for a 4 ohm saves$ then get data and response curves
2) model bass reflex encosure in Unibox or WinISD alpha
Get Vb and Fb find port size
3) pick box dimensions from 2) use golden mean
4) go to peerless / vifa web site get midbass response data
5) model BSC using edge & pick Inductor using 1st order calc
& calc loss due to Inductor
6) pick x over frq and order using 1,2,4&5
7) calc L pad using 1,4,&5
crossover s/b around 3ish KHz
I think that the other thing to remember about commercial speakers is that while they may use some very inexpensive parts and drivers, there may also be a lot of engineering going on there to get the acceptable sound out of them.
Designing a good sounding speaker with a very tight budget is a lot harder than using the best drivers and unlimited funds for cabinets and crossover parts.
Designing a good sounding speaker with a very tight budget is a lot harder than using the best drivers and unlimited funds for cabinets and crossover parts.
FWIW
I looked at those peerless 5,25 they s/b perfect for an easy 1st or 2nd order LPF. Since the crossover will be around 3 KHZ or a little higher for 1st order, the tweeter will be crossed closer to 4-5Khz or 4 times Fs means you probably can get away with a single series cap as well.
A little research and work without measuring can get you in the right direction. And then some tweaking with the cap values and a Lpad you can adjust to your ear.
This a DIY forum so we mean to encourage the Y part.
I looked at those peerless 5,25 they s/b perfect for an easy 1st or 2nd order LPF. Since the crossover will be around 3 KHZ or a little higher for 1st order, the tweeter will be crossed closer to 4-5Khz or 4 times Fs means you probably can get away with a single series cap as well.
A little research and work without measuring can get you in the right direction. And then some tweaking with the cap values and a Lpad you can adjust to your ear.
This a DIY forum so we mean to encourage the Y part.
From what i've read in this topic it seems you would be better served if you build an already working project from other people here in the forums.
The advantage would be a nice learning effect and a bigger probability to build a good sounding loudspeaker.
As previously mentioned the correct design of the crossover filter is complex and not to forget the driver/box/damping interaction......
Best regards,
Mike
The advantage would be a nice learning effect and a bigger probability to build a good sounding loudspeaker.
As previously mentioned the correct design of the crossover filter is complex and not to forget the driver/box/damping interaction......
Best regards,
Mike
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.