I want a 70hz first order crossover for a 4 ohm speaker. The chart i have says that I will need a 568uf capacitor. I want to use a 500uf and 68uf capacitor in parallel to create the crossover. Should there be any problems with this?
yes C parallel adds.
capacitors used on speakers must be non-polar types of sufficient voltage withstand and higher AC currents, most buy them from audio / speaker suppliers. what kind of caps are you using?
most importantly your 4 ohm speaker most likely isn't 4 ohms at those frequencies, so 'the chart' will give errors ( wrong values ) in the numbers it spits out. it can be done but not very accurately, speakers have terrible impedance variances. look in on active crossovers, besides accuracy improvements , the capacitors themselves are smaller and not extremely expensive either.
capacitors used on speakers must be non-polar types of sufficient voltage withstand and higher AC currents, most buy them from audio / speaker suppliers. what kind of caps are you using?
most importantly your 4 ohm speaker most likely isn't 4 ohms at those frequencies, so 'the chart' will give errors ( wrong values ) in the numbers it spits out. it can be done but not very accurately, speakers have terrible impedance variances. look in on active crossovers, besides accuracy improvements , the capacitors themselves are smaller and not extremely expensive either.
Last edited:
AH that was my next question. I know impedance goes up near the resonant frequency thus lowering the effective frequency of the high pass. Im trying to filter out the low frequencies around the fs of my driver and let the sub do the work there so this would be bad. What should I do? I need a cheap crossover solution. (under $50 for 2 or hopefully 4 channels) and I want it at 70 hz.
I was looking at the 100v non polar electrolytic capacitors from parts express. Its for a 22 watt RMS 100 watt peak current.
The speakers are the stock 6by9 woofers in my car so I cant get much info on them. Unfortunately I have no tools for measuring impedance spikes and t/s parameters.
I was looking at the 100v non polar electrolytic capacitors from parts express. Its for a 22 watt RMS 100 watt peak current.
The speakers are the stock 6by9 woofers in my car so I cant get much info on them. Unfortunately I have no tools for measuring impedance spikes and t/s parameters.
Last edited:
you can use simple RC filters on the inputs to amplifiers, but you need to have access to the schematics or measurement tools to find driving source and higher input impedances. aka "poor mans active filter solution"
the same charts might work here as well. > use amps input impedance instead of the speaker ( typical values range from 10K to 47K ) sometimes you can tack them onto the amplifier PCB if you want it more or less permanent.
the same charts might work here as well. > use amps input impedance instead of the speaker ( typical values range from 10K to 47K ) sometimes you can tack them onto the amplifier PCB if you want it more or less permanent.
Last edited:
yes this shows some good understanding, unfortunately 70Hz is very close to the resonance Eg rising impedance of your speaker in a box so a series cap doesn't attenuate LF too well.
its not hard to measure source and input impedances.
use a 400 Hz test tone (recorded and played on your source) and use a DMM on AC volts ~ 1V to 200mV
use a couple of known value 1% test resistors 1K and 10K > apply ohms law on Thevenins theory
use a 400 Hz test tone (recorded and played on your source) and use a DMM on AC volts ~ 1V to 200mV
use a couple of known value 1% test resistors 1K and 10K > apply ohms law on Thevenins theory
Last edited:
The capacitor tolerance may be 10% or 20% so one marked 500uF could be somewhere between 450uF and 550uF. This is fine as a '4 ohm' speaker won't be 4 ohms either.
Bipolar electrolytic of 50Vac 500uF (for upto 250W into 8ohms) might be expensive or difficult to source.
voltage rails for a really big amplifier will give close to 200 V p-p or 71 VRMS choose 100 VAC ratings
- Status
- Not open for further replies.