True Profiguy, some things are unavoidable like reflections and other issues which has to be dealt with on an experimental basis. You start with a good sound source else you have no idea where to look for environmental oddities.
No, I did not say "cancel the ground reflection", I wrote a ground reflection can cause a dip in the speaker's response. The frequency of the dip is dependent on the ground reflection vs direct path length distance to the drivers.
The height of the drivers and mic from the ground in the OP measurement could be the cause of the 3kHz dip, and may not be present in the usual listening position.
At any rate, until ciccio64 makes further measurements, we won't know the cause of the dip.
The crossover values just need to be changed slightly
to get it to sum better.
Big phase jump can happen around 3k with one of
the drivers. So it is nothing to surprising.
Center to center is close.
But their is still a good amount of distance
that cant be changed.
Depends also on order of the filters.
likely the woofer isnt as steep.
So its top end response can cancel some.
It is usual challenge of wider center to center.
you'd usually end up with a tweeter crossed to low
to compensate and make it look great.
And then the phase flips up top with both drivers
cause this problem at higher more ideal crossover points,
for of course less tweeter distortion.
Solution usually make both woofer and tweet
3rd order or higher.
When instead is common to have cheaper 2nd order
on woofer and third order on tweet.
The phase flip flop up top just welcomes dips like this.
Realistic model, or correct measurements will help solve
a few component value changes
to get it to sum better.
Big phase jump can happen around 3k with one of
the drivers. So it is nothing to surprising.
Center to center is close.
But their is still a good amount of distance
that cant be changed.
Depends also on order of the filters.
likely the woofer isnt as steep.
So its top end response can cancel some.
It is usual challenge of wider center to center.
you'd usually end up with a tweeter crossed to low
to compensate and make it look great.
And then the phase flips up top with both drivers
cause this problem at higher more ideal crossover points,
for of course less tweeter distortion.
Solution usually make both woofer and tweet
3rd order or higher.
When instead is common to have cheaper 2nd order
on woofer and third order on tweet.
The phase flip flop up top just welcomes dips like this.
Realistic model, or correct measurements will help solve
a few component value changes
Hi welter Speaker are on a stand some 70 cm high.the mic is in line with tw I would say 1 meter from floor
Well Kevin at least I do not have such problem my speakers are relatively light but still quite heavy for what they are
Indeed
And infact this is the point of discussion here
Sound wise they are quite ok already for me
Hi White
Tks for your useful input. The crossover is indeed third order on the TW and second order on the woofer this was done not to be on the chip but just because it was looking good on the XSim simulation that way (i actually invested more than i should on crossover component)
As said i am by no means an expert but i try to make some plan when i do something. So the plan was to build a speaker decent quality, small size but with decent bass last but not least keeping cost on human level. Hence the choice of component. The SPH135c monacor has a low Fs so is capable on going quite low for the size on the other hand he cannot be crossed over to high in my opinion also i believe a good TW should perform better in that range than
a Wf,. Anyway i chose the monacor Dt254 which has an fs of 900 Hz so supposedly at 1800 Hz with a third order filter should be ok
I do not have all data with me to share as i am out for work but i appreciate all your input .I belive i shall try a third order filter on the woofer as well as i already tried to play with value of component in the crossover in present configuration with no success. The attachment is the xsim simulation but some final touch were made after critical listening.
Thank you again
Hi all one last post just to thank all of you for your input I have now so much information that shall take me sometime to digest.i shall try to put those at good use and I shall update you with results.
Thank you again to this great community 🙏 😁
Thank you again to this great community 🙏 😁
That would eliminate ground bounce as the cause of the 3kHz dip...
Considering the frequency and impedance response of both drivers are pretty smooth in the 3kHz range, there is chance matching the filtered response roll off of the woofer and tweeter (both to third order) could correct the problem, if it is crossover phase related rather than the baffle edge diffraction effect.
Best of luck!
Art
Art, someone mentioned before. Maybe it is a simple phase inversion and requires simply to flip the terminals on the tweeter. The originator should just try it and see.
Looks like the crossover should be at 2.5 khz .
Bringing the tweeter a little lower should pull it up a fair bit.
Bringing the tweeter a little lower should pull it up a fair bit.
Fifteen years ago I discovered "audio nirvana" by stacking two very different speakers (se300B low-end Monitor Audio Monitor2.5 for tweeter/bass, BelCanto ICEpower Fostex F120A for mids). I posted to audiogon and got both anecdotal support and principled ridicule, of course. I should have left well enough alone but broke up the set instead. I think I finally matched or exceeded that "stack" but only with even more unconventional DIY.
Last edited:
Some off topic noise has been removed. If there's a next time it would be easier to exclude the noisemakers.
Hi Shadowplay62
Tks for your advice if I understand properly the idea is to move the crossover point little higher and it looks promising on the simulation but so was my iteration and also this wold require a complete crossover re build. In time I shall give a try thank you again, 😁
Tks for your advice if I understand properly the idea is to move the crossover point little higher and it looks promising on the simulation but so was my iteration and also this wold require a complete crossover re build. In time I shall give a try thank you again, 😁
Hi Ciccio,
the FR of the Monacor woofer it's not so easy to manage without LCR compensations as a 1.8KHz peak, followed by a dip with a sharp slope, and then two additional peaks at 5 and 10KHz.
You can avoid the LCR notches but the response will be rugged.
Ciao
S.
the FR of the Monacor woofer it's not so easy to manage without LCR compensations as a 1.8KHz peak, followed by a dip with a sharp slope, and then two additional peaks at 5 and 10KHz.
You can avoid the LCR notches but the response will be rugged.
Ciao
S.
Driver inversions and improper phase look very different. A driver inversion is a much deeper null. Here's a picture from a speaker I built recently. It's a 3-way center channel. The dip at ~ 2.5 kHz in black is measurement 45 degrees above the tweeter axis:
Above the tweeter axis the tweeter is now closer and the midrange further away from the measurement microphone, causing moderate cancellation. Note that the green line is 45 degrees horizontally and shows no such problem. The black line is very similar looking to your original FR which is why I suspect your simulation did not properly take into account the tweeter/woofer offset. Reducing the tweeter to second order, and other adjustments such as inverting the tweeter would need to be made but I suggest properly measuring the tweeter/woofer distances before proceeding. Simply reversing the tweeter will probably not get you to fixed either. You now have run into why designers of passive crossovers often must change their filter orders to match the phases properly.
Of course, another option is to add a baffle for the woofer, bringing it further forward, but that's crazy talk. 🙂
As I mentioned before, the proof of this is to measure below the tweeter axis to see if the null fixes itself. You can do this partially by tilting the speaker up. This would also eliminate the idea of it being floor bounce.
Above the tweeter axis the tweeter is now closer and the midrange further away from the measurement microphone, causing moderate cancellation. Note that the green line is 45 degrees horizontally and shows no such problem. The black line is very similar looking to your original FR which is why I suspect your simulation did not properly take into account the tweeter/woofer offset. Reducing the tweeter to second order, and other adjustments such as inverting the tweeter would need to be made but I suggest properly measuring the tweeter/woofer distances before proceeding. Simply reversing the tweeter will probably not get you to fixed either. You now have run into why designers of passive crossovers often must change their filter orders to match the phases properly.
Of course, another option is to add a baffle for the woofer, bringing it further forward, but that's crazy talk. 🙂
As I mentioned before, the proof of this is to measure below the tweeter axis to see if the null fixes itself. You can do this partially by tilting the speaker up. This would also eliminate the idea of it being floor bounce.
Hi Erik
Thank you for your input
It is true due to my inexperience I did not give importance to the speakers off set
All you say is very intriguing..as I get home o shall make some measure and test to see Better what's going on
It not such a crazy idea to add an additional baffle for the woofer,,,,,😁😁😁😁
Thank you for your input
It is true due to my inexperience I did not give importance to the speakers off set
All you say is very intriguing..as I get home o shall make some measure and test to see Better what's going on
It not such a crazy idea to add an additional baffle for the woofer,,,,,😁😁😁😁