Hi,
I've built my first pair of speakers (other than subwoofers). I've used components from a well-known American speaker component retailer.
The drivers are all Dayton Audio 8 ohm drivers - each speaker has a DC300-8 12-inch woofer, a PC105-8 4-inch "full range" driver as the midrange (one of its suggested applications), and a DC25T-8 1-inch dome tweeter.
Each speaker has a ~1.5 cu ft sealed main/bass enclosure (built from 18mm / 0.7" thick MDF). Inside, the top, bottom, left and right walls, and part of the back wall (around the PRV Audio 3DF500/3500 8 ohm 3-way crossover - 500Hz and 3.5KHz at 12dB per octave) are lined with two thick sheets of polyester wadding.
Built into each speaker is a ~0.07 cu ft sealed midrange enclosure (made from a 10mm / 0.4" thick rigid cardboard tube, with an 18mm / 0.7" thick chipboard baffle and rear piece). Inside, polyester wadding is fitted to one side of the length of the tube.
The woofer impedance jumpers on the crossovers are correctly set to 8 ohm, and the attenuation jumpers on the midranges and tweeters are set to -6dB (it's not enough attenuation, so I'd like to add L-pads, but I'll worry about that another time).
Now, obviously, until I add the L-pads, I'm using equalisation (system-wide software on my source PC, in my case), cutting the frequencies in the passband of the midranges by one amount, and cutting the frequencies in the passband of the tweeters by another amount.
However, unlike the woofers and tweeters, the frequency response of the midranges within their passband is far from flat. There's an enormous peak around 400-600Hz, into the double digits of dB.
This is despite me having been advised that the very reason for making the midrange enclosures as "big" as 0.09 cu ft (granted, I had to make them ~0.07 cu ft to fit), is to not cause such a peak around the midrange HPF of 500Hz.
Other than the midrange enclosures being slightly smaller than I was advised to make them, what could be causing such a big peak around 400-600Hz? And what can I do to get rid of it?
I understand that in 3-way speakers with passive crossovers, the woofers and tweeters are wired normally, but the midranges are wired in reverse polarity.
To my knowledge, I've done this - I've wired the woofers and tweeters normally at the driver terminals. Then, I've wired the midranges "backwards" at the driver terminals, then normally (both internally and externally) at the midrange enclosure terminals.
That of course is assuming that the colour coding of the midrange wires from the crossover is not reversed (black wire for +, and coloured wire for -). Could it be the case that it is?
Tomorrow, I'll swap the + and - wires "externally" at the midrange enclosure terminals and see (or hear) whether or not there's an improvement.
Other than that though, what can I do? Fit more polyester wadding in the midrange enclosures? Apply bitumen deadening to the midrange enclosures? Wrap polyester wadding round the midrange driver frame openings? Use smaller (3") and/or different make/model of midrange drivers in the same enclosures? Modify the midrange circuits of the crossovers to raise the midrange HPFs to say 1KHz? Anything else?
Any help would be much appreciated 😊
I've built my first pair of speakers (other than subwoofers). I've used components from a well-known American speaker component retailer.
The drivers are all Dayton Audio 8 ohm drivers - each speaker has a DC300-8 12-inch woofer, a PC105-8 4-inch "full range" driver as the midrange (one of its suggested applications), and a DC25T-8 1-inch dome tweeter.
Each speaker has a ~1.5 cu ft sealed main/bass enclosure (built from 18mm / 0.7" thick MDF). Inside, the top, bottom, left and right walls, and part of the back wall (around the PRV Audio 3DF500/3500 8 ohm 3-way crossover - 500Hz and 3.5KHz at 12dB per octave) are lined with two thick sheets of polyester wadding.
Built into each speaker is a ~0.07 cu ft sealed midrange enclosure (made from a 10mm / 0.4" thick rigid cardboard tube, with an 18mm / 0.7" thick chipboard baffle and rear piece). Inside, polyester wadding is fitted to one side of the length of the tube.
The woofer impedance jumpers on the crossovers are correctly set to 8 ohm, and the attenuation jumpers on the midranges and tweeters are set to -6dB (it's not enough attenuation, so I'd like to add L-pads, but I'll worry about that another time).
Now, obviously, until I add the L-pads, I'm using equalisation (system-wide software on my source PC, in my case), cutting the frequencies in the passband of the midranges by one amount, and cutting the frequencies in the passband of the tweeters by another amount.
However, unlike the woofers and tweeters, the frequency response of the midranges within their passband is far from flat. There's an enormous peak around 400-600Hz, into the double digits of dB.
This is despite me having been advised that the very reason for making the midrange enclosures as "big" as 0.09 cu ft (granted, I had to make them ~0.07 cu ft to fit), is to not cause such a peak around the midrange HPF of 500Hz.
Other than the midrange enclosures being slightly smaller than I was advised to make them, what could be causing such a big peak around 400-600Hz? And what can I do to get rid of it?
I understand that in 3-way speakers with passive crossovers, the woofers and tweeters are wired normally, but the midranges are wired in reverse polarity.
To my knowledge, I've done this - I've wired the woofers and tweeters normally at the driver terminals. Then, I've wired the midranges "backwards" at the driver terminals, then normally (both internally and externally) at the midrange enclosure terminals.
That of course is assuming that the colour coding of the midrange wires from the crossover is not reversed (black wire for +, and coloured wire for -). Could it be the case that it is?
Tomorrow, I'll swap the + and - wires "externally" at the midrange enclosure terminals and see (or hear) whether or not there's an improvement.
Other than that though, what can I do? Fit more polyester wadding in the midrange enclosures? Apply bitumen deadening to the midrange enclosures? Wrap polyester wadding round the midrange driver frame openings? Use smaller (3") and/or different make/model of midrange drivers in the same enclosures? Modify the midrange circuits of the crossovers to raise the midrange HPFs to say 1KHz? Anything else?
Any help would be much appreciated 😊
Last edited:
it could be cabinet resonance, baffle step, the mid enclosure or resonance from the crossover or a bad luck combination of all.
Could you post some pics of the speaker and the build such as the bracing and baffle layout with dimensions?
I think that would help.
Have you measured this peak or is it by ear only?
Are you sure it’s from the midrange and not the bass?
Perhaps try to make some more customized crossovers after you have figured out it’s not related to the cabinet.
Could you post some pics of the speaker and the build such as the bracing and baffle layout with dimensions?
I think that would help.
Have you measured this peak or is it by ear only?
Are you sure it’s from the midrange and not the bass?
Perhaps try to make some more customized crossovers after you have figured out it’s not related to the cabinet.
Its hard to tell, but if big peak is around 1st xo point it could be some sort of electrical resonance/interaction between crossover and driver(s).
If impedance at input terminals drops very low at same frequency as response peak thats most likely the case.
Are you are using generic (of the shelf) xo?
Such xo is made with theoretical values of components, its not optimized. So problems like this are possible.
You say -6dB of attenuation is not enough.
Thats strange, since woofer sensitivity is 4db above the midrange, so effectively you should have 10dB of bass boost.
That also ponts to xo issues to me.
Try to experiment with polarity of drivers, it may help sometimes.
If impedance at input terminals drops very low at same frequency as response peak thats most likely the case.
Are you are using generic (of the shelf) xo?
Such xo is made with theoretical values of components, its not optimized. So problems like this are possible.
You say -6dB of attenuation is not enough.
Thats strange, since woofer sensitivity is 4db above the midrange, so effectively you should have 10dB of bass boost.
That also ponts to xo issues to me.
Try to experiment with polarity of drivers, it may help sometimes.
Given your apparent experience level, it's also worth looking at how you are measuring to make sure it's not a measurement error. Taking quasi-anechoic measurements in a home is not as straightforward as it may seem at first glance. How far away from the speaker are you measuring? How far away from the mic and speaker are the nearest reflective surfaces? What gating/windowing are you using? Have you moved the microphone and speaker to see if the peak changes based on those conditions? Etc.
Have you measured the drivers individually nearfield to try to narrow down the source of the problem? You can also do that with and without crossover for the woofer and midrange. Don't try to measure the tweeter without crossover since it's not where your problem is and it's easier to damage them.
Have you measured the drivers individually nearfield to try to narrow down the source of the problem? You can also do that with and without crossover for the woofer and midrange. Don't try to measure the tweeter without crossover since it's not where your problem is and it's easier to damage them.
Greets!
Hmm, (400*600)^2 = ~490 Hz mean
(open cylinder) 34400/2/490 = ~35.11 cm, so if yours is somewhere reasonably close, then it's likely the pipe, especially since it's only lightly lined Vs the well stuffed to 'taste' that's required.
Today, I put more polyester wadding in the midrange enclosures (a lot, but not crammed in, so some air can still get through I guess?), and that made a slight improvement.
Then, I changed the polarity at the midrange enclosure's "external" terminals, and that made a much more significant improvement, so I guess the crossovers could indeed have their midrange outputs wired "backwards".
Now, the speakers are at the point where the frequency response within each driver's passband seems reasonably flat-ish (I've only been "measuring" with my ears).
Therefore, I now actually have a fairly decent starting point for when I add L-pads for the midranges and tweeters to pad them down.
Also, I reckon that if the main/bass enclosures and the midrange enclosures were lined with bitumen deadening before the polyester wadding was added, there would be a further improvement in the "resonance issues" department.
Thanks very much again everyone, and sorry for not replying to everything/everyone!
Then, I changed the polarity at the midrange enclosure's "external" terminals, and that made a much more significant improvement, so I guess the crossovers could indeed have their midrange outputs wired "backwards".
Now, the speakers are at the point where the frequency response within each driver's passband seems reasonably flat-ish (I've only been "measuring" with my ears).
Therefore, I now actually have a fairly decent starting point for when I add L-pads for the midranges and tweeters to pad them down.
Also, I reckon that if the main/bass enclosures and the midrange enclosures were lined with bitumen deadening before the polyester wadding was added, there would be a further improvement in the "resonance issues" department.
Thanks very much again everyone, and sorry for not replying to everything/everyone!
It may be got lost in al the replies, but you seem to use an "universal" passive crossover.
Please accept that these do not work. Never, ever.
Any set of drivers needs a crossover exactly designed to match them.
You will ask:"but they sell them even at PartsExpress, so they should work?" Sorry, we live in a demand fullfilling economy. If people want an "universal" part, there will be a vendor selling them. If you are lucky, these crossovers will prevent the tweeter and mid to be blown on the first loud passage of music. Sound and frequency response will be unusable in any case. Sorry again, no exceptions.
There is another thing, your ears adapt to a weird frequency response after a while. So, after some listening, you may even think "hey, that's not bad". As long as you do not listen to a good speaker, you may stay happy.
If you understand this fact about crossovers, you can ask for help. Your problems are not solvable by some more stuffing or reversing driver polarity. This may change what you hear, but nothing will get better. The crossover is the most critical part of any loudspeaker.
Please accept that these do not work. Never, ever.
Any set of drivers needs a crossover exactly designed to match them.
You will ask:"but they sell them even at PartsExpress, so they should work?" Sorry, we live in a demand fullfilling economy. If people want an "universal" part, there will be a vendor selling them. If you are lucky, these crossovers will prevent the tweeter and mid to be blown on the first loud passage of music. Sound and frequency response will be unusable in any case. Sorry again, no exceptions.
There is another thing, your ears adapt to a weird frequency response after a while. So, after some listening, you may even think "hey, that's not bad". As long as you do not listen to a good speaker, you may stay happy.
If you understand this fact about crossovers, you can ask for help. Your problems are not solvable by some more stuffing or reversing driver polarity. This may change what you hear, but nothing will get better. The crossover is the most critical part of any loudspeaker.