This is my first time doing anything with diy audio, so i'm confused on what I actually need.
So far I have picked these components
This is the amplifier TDA7498E 2.1
Full range drivers Eminence Alpha 4-8 4" 90-12,000hz
Tweeters Peerless by Tymphany DX25TG59-04
With the tweeters and drivers, do I need a cross over with them? Do I build the cross over or would a premade board work?
Any other tips would be great also.
So far I have picked these components
This is the amplifier TDA7498E 2.1
Full range drivers Eminence Alpha 4-8 4" 90-12,000hz
Tweeters Peerless by Tymphany DX25TG59-04
With the tweeters and drivers, do I need a cross over with them? Do I build the cross over or would a premade board work?
Any other tips would be great also.
That Eminence is sort of small to use with a tweeter, but you can. It may be as simple as a single cap in series with the tweeter, or cap a resistor.
My "minimalist method" is as follows.
o Android APP Frequency Sound Generator, finger-sweep and compare-by-ear frequency response
o a few typical small-value MKP caps 1 1.5 2.2 2.7 3.3 4.7 6.8 uF add in parallel
o a few typical small air-core inductors 0.15 0.2 0.3 mH add in series
o typical 5 or 10V resistors 1 3 10 ohm add in series
o small electrician's conehead twist-caps / \ to "build" test crossover, sometimes directly over the speaker or amp binding-posts
o banana-plug/alligator-clip wire, connectors, binding-posts, etc.
o simplified formula for 1st-order XO -6dB per octave i.e. halved (driver impedance in ohm)
high-pass-filter=-3dB@(160khz/imp)/uF
low-pass-filter=-3dB@(160hz*imp)/mH
(e.g. imp=8 HPF@20khz/uF LPF@1280/mH, typically set to equal before tweaking by ear
o easiest to choose sensitivity-matching drivers; attenuate with resistors only if absolutely necessary
o 2nd-order XO -12dB per octave merely adds a complementary (capacitor/inductor) bypass across speaker poles, but 1st-order should be enough
o if the full-range reaches high enough but has a resonance peak (tyically ~5-7khz) or otherwise needs to be "tamed", the simplest notch filter is to treat it as two virtual parallel drivers, and connect both HPF and LPF (in parallel) to speaker + (so-called in series), the XO frequencies being unequal so as to sum (precisely and predictably) to a valley at the resonance peak
o if the above-formed notch valley is detectable by ear (using the APP), add a series-chain of resistors in parallel with HPF and LPF
o the proper way to position speakers is to form an approximately equilateral triangle with the listening position, speakers toe-in aimed directly at an ear (but front-left/right to ear-canal); otherwise, high frequencies are lost
o that's basically it
o Android APP Frequency Sound Generator, finger-sweep and compare-by-ear frequency response
o a few typical small-value MKP caps 1 1.5 2.2 2.7 3.3 4.7 6.8 uF add in parallel
o a few typical small air-core inductors 0.15 0.2 0.3 mH add in series
o typical 5 or 10V resistors 1 3 10 ohm add in series
o small electrician's conehead twist-caps / \ to "build" test crossover, sometimes directly over the speaker or amp binding-posts
o banana-plug/alligator-clip wire, connectors, binding-posts, etc.
o simplified formula for 1st-order XO -6dB per octave i.e. halved (driver impedance in ohm)
high-pass-filter=-3dB@(160khz/imp)/uF
low-pass-filter=-3dB@(160hz*imp)/mH
(e.g. imp=8 HPF@20khz/uF LPF@1280/mH, typically set to equal before tweaking by ear
o easiest to choose sensitivity-matching drivers; attenuate with resistors only if absolutely necessary
o 2nd-order XO -12dB per octave merely adds a complementary (capacitor/inductor) bypass across speaker poles, but 1st-order should be enough
o if the full-range reaches high enough but has a resonance peak (tyically ~5-7khz) or otherwise needs to be "tamed", the simplest notch filter is to treat it as two virtual parallel drivers, and connect both HPF and LPF (in parallel) to speaker + (so-called in series), the XO frequencies being unequal so as to sum (precisely and predictably) to a valley at the resonance peak
o if the above-formed notch valley is detectable by ear (using the APP), add a series-chain of resistors in parallel with HPF and LPF
o the proper way to position speakers is to form an approximately equilateral triangle with the listening position, speakers toe-in aimed directly at an ear (but front-left/right to ear-canal); otherwise, high frequencies are lost
o that's basically it
(Edited)
My "minimalist method" is as follows.
o Android APP Frequency Sound Generator, finger-sweep and compare-by-ear frequency response at multiple points (and against manufacturer's FR curve if available)
o a few typical small-value MKP caps 1 1.5 2.2 2.7 3.3 4.7 6.8 uF add in parallel
o a few typical small air-core inductors 0.15 0.2 0.3 mH add in series
o typical 5 or 10V resistors 1 3 10 ohm add in series
o small electrician's conehead twist-caps / \ to "build" test crossover, sometimes directly over the speaker or amp binding-posts
o banana-plug/alligator-clip wire, connectors, binding-posts, etc.
o simplified formula for 1st-order XO -6dB per octave i.e. halved (driver impedance in ohm)
high-pass-filter=-3dB@(160khz/imp)/uF
low-pass-filter=-3dB@(160hz*imp)/mH
(e.g. imp=8 HPF@20khz/uF LPF@1280/mH, typically set to equal before tweaking by ear
o easiest to choose sensitivity-matched drivers; attenuate with resistors only if absolutely necessary
o 1st-order XO should be enough if drivers are well-behaved (better phase-matched for fullness of sound and precise imaging, besides costing less)
o 2nd-order XO -12dB per octave merely adds a complementary (capacitor/inductor) bypass across speaker poles; by keeping the complementary uF*mH product unchanged but varying uF and mH, the fall-off inflection curve can be tweaked (so-called Bessel, Butterworth, or Linkwitz-Riley crossovers)
o if the full-range driver reaches high enough but has a resonance peak (tyically ~5-7khz) or otherwise needs to be "tamed", the simplest notch filter is to treat it as two virtual parallel drivers, and connect both HPF and LPF (in parallel) to speaker + pole (so-called in series), the XO frequencies being unequal so as to sum (precisely and predictably) to a notch at the resonance peak
o if the above-formed notch valley is detectable by ear (using the APP), add a series-chain of resistors in parallel with HPF and LPF to in-fill the notch
o the proper way to position speakers is to form an approximately equilateral triangle with the listening position, speakers toe-in aimed directly at an ear, but front-left/right to ear-canal for max hearing sensitivity (otherwise, very high frequencies are lost); beware, especially during testing/tweaking!
o that's basically it
My "minimalist method" is as follows.
o Android APP Frequency Sound Generator, finger-sweep and compare-by-ear frequency response at multiple points (and against manufacturer's FR curve if available)
o a few typical small-value MKP caps 1 1.5 2.2 2.7 3.3 4.7 6.8 uF add in parallel
o a few typical small air-core inductors 0.15 0.2 0.3 mH add in series
o typical 5 or 10V resistors 1 3 10 ohm add in series
o small electrician's conehead twist-caps / \ to "build" test crossover, sometimes directly over the speaker or amp binding-posts
o banana-plug/alligator-clip wire, connectors, binding-posts, etc.
o simplified formula for 1st-order XO -6dB per octave i.e. halved (driver impedance in ohm)
high-pass-filter=-3dB@(160khz/imp)/uF
low-pass-filter=-3dB@(160hz*imp)/mH
(e.g. imp=8 HPF@20khz/uF LPF@1280/mH, typically set to equal before tweaking by ear
o easiest to choose sensitivity-matched drivers; attenuate with resistors only if absolutely necessary
o 1st-order XO should be enough if drivers are well-behaved (better phase-matched for fullness of sound and precise imaging, besides costing less)
o 2nd-order XO -12dB per octave merely adds a complementary (capacitor/inductor) bypass across speaker poles; by keeping the complementary uF*mH product unchanged but varying uF and mH, the fall-off inflection curve can be tweaked (so-called Bessel, Butterworth, or Linkwitz-Riley crossovers)
o if the full-range driver reaches high enough but has a resonance peak (tyically ~5-7khz) or otherwise needs to be "tamed", the simplest notch filter is to treat it as two virtual parallel drivers, and connect both HPF and LPF (in parallel) to speaker + pole (so-called in series), the XO frequencies being unequal so as to sum (precisely and predictably) to a notch at the resonance peak
o if the above-formed notch valley is detectable by ear (using the APP), add a series-chain of resistors in parallel with HPF and LPF to in-fill the notch
o the proper way to position speakers is to form an approximately equilateral triangle with the listening position, speakers toe-in aimed directly at an ear, but front-left/right to ear-canal for max hearing sensitivity (otherwise, very high frequencies are lost); beware, especially during testing/tweaking!
o that's basically it
With 105Hz - 12kHz of usable range its more like extended midrange than full-range.
You may try just a cap in series with tweeter, but I would be more comfortable with some basic x-over somewhere around 4kHz.
Eminence response is fairly smooth with some rise towards highs. Taking that rise into account I would go with 0,47mH in series with it.
For tweeter, around 4,7uF cap in series and maybe 1 - 1,5 ohm resistor in series to match efficiency, if necessary. Both values fine tuned by ear to your taste.
To do that, get few smaller caps instead single 4,7uF value, and combine their values in parallel to find the best final value.
You may try just a cap in series with tweeter, but I would be more comfortable with some basic x-over somewhere around 4kHz.
Eminence response is fairly smooth with some rise towards highs. Taking that rise into account I would go with 0,47mH in series with it.
For tweeter, around 4,7uF cap in series and maybe 1 - 1,5 ohm resistor in series to match efficiency, if necessary. Both values fine tuned by ear to your taste.
To do that, get few smaller caps instead single 4,7uF value, and combine their values in parallel to find the best final value.
o to check phase-alignment of the XO-ed drivers, simply play the APP at the XO frequency, and manually move the tweeter forward and back to locate the periodically recurring volume/fullness peaks.
o XO network staying the same, higher impedance driver will play more (passes more, filters less) than lower impedance driver; this applies to both tweeter/HPF and midwoofer/LPF.
o XO network staying the same, higher impedance driver will play more (passes more, filters less) than lower impedance driver; this applies to both tweeter/HPF and midwoofer/LPF.
Sorry, I missed that tweeter is 4 Ohm while full-range is 8 Ohm. Thats not very best situation. Unless you already bought the drivers I suggest you to get suitable 8 ohm tweeter.
If you still going to use 4 Ohm tweeter, it will work, but cap value in series with tweeter must be doubled to get the same x-over point.
If you still going to use 4 Ohm tweeter, it will work, but cap value in series with tweeter must be doubled to get the same x-over point.
It just seems like an odd combination. If you really need to top off a 4" driver, why this tweeter?
Although you won't get any bass from this speaker, if you want to blend them together you might try a first order crossover at about 6 kHz. Remember that your woofer is 8 ohm (for the inductor) and your tweeter is 4 ohm (for the cap).
Meaning somewhere around 6.8 uF for the cap and 0.2mH for the coil (inductor). You might need a resistor inline with the tweeter to bring its level down a decibel or three.
Although you won't get any bass from this speaker, if you want to blend them together you might try a first order crossover at about 6 kHz. Remember that your woofer is 8 ohm (for the inductor) and your tweeter is 4 ohm (for the cap).
Meaning somewhere around 6.8 uF for the cap and 0.2mH for the coil (inductor). You might need a resistor inline with the tweeter to bring its level down a decibel or three.