Hi All
My first speaker project in almost a decade is up and running. Stumbled onto some Pro-Ac response 2 Cabinets that were about to be trashed, and seeing how I couldn't let that happened, I decided to build something into them. After some researching, decided to put some Seas CA 18RNX woofers and 27 TFFC tweeters into them.
Put them together first with just a 4.7 cap on the tweeter, and found a lot of detail, but INCREDIBLY bright sound. Measured the response in a local anechoic chamber, and found a big hump from about 3k on up.
With some local help, put a crossover together. 2nd order butterworth on the woofer (.72mh coil and 5.6 cap) and a 3rd order on the tweeter (7.1 cap followed by a 15 cap with a .29mh coil and a 2.2 ohm and 10 ohm resitor). Sounds much smoother now, with still some brightness/thinness. Letting them break in now. Still have to tune the port (impedance measurements showed that Pro-Ac tuned it to 50hz whereas the driver looks like it would be happier at 45hz). Oh, and the straws in the port are gonna go too!
Anyway, anyone have any thoughts or suggestions on this combo? Anything I have obviously done wrong?
Thanks!
Erik
My first speaker project in almost a decade is up and running. Stumbled onto some Pro-Ac response 2 Cabinets that were about to be trashed, and seeing how I couldn't let that happened, I decided to build something into them. After some researching, decided to put some Seas CA 18RNX woofers and 27 TFFC tweeters into them.
Put them together first with just a 4.7 cap on the tweeter, and found a lot of detail, but INCREDIBLY bright sound. Measured the response in a local anechoic chamber, and found a big hump from about 3k on up.
With some local help, put a crossover together. 2nd order butterworth on the woofer (.72mh coil and 5.6 cap) and a 3rd order on the tweeter (7.1 cap followed by a 15 cap with a .29mh coil and a 2.2 ohm and 10 ohm resitor). Sounds much smoother now, with still some brightness/thinness. Letting them break in now. Still have to tune the port (impedance measurements showed that Pro-Ac tuned it to 50hz whereas the driver looks like it would be happier at 45hz). Oh, and the straws in the port are gonna go too!
Anyway, anyone have any thoughts or suggestions on this combo? Anything I have obviously done wrong?
Thanks!
Erik
Has the reponse when the tweeter polarity is reversed been measures? You can most of the times always get a nice smooth response with the correct polarity, but when the polarity is reversed you can quite easily not get a notch.
What you want to see is a nice deep null around the xover frequency when you reverse the polarity, this shows that the drivers are in phase thru the xover region.
What you want to see is a nice deep null around the xover frequency when you reverse the polarity, this shows that the drivers are in phase thru the xover region.
These are very nice units.
.7mH on the woofer sounds a bit on the small side. No wonder that the sound seems bright. I would shoot for something like 2.5mH combined with 12uF, which probably would give a xover freq of about 2000Hz with a nice 4th order rolloff.
And to match that: a second order topology on the tweeter with around 10uF, 0.2Mh in parallell and some resistors for setting of level should be a good place to start.
Happy testing!
EE
.7mH on the woofer sounds a bit on the small side. No wonder that the sound seems bright. I would shoot for something like 2.5mH combined with 12uF, which probably would give a xover freq of about 2000Hz with a nice 4th order rolloff.
And to match that: a second order topology on the tweeter with around 10uF, 0.2Mh in parallell and some resistors for setting of level should be a good place to start.
Happy testing!
EE
Dear bigerik,
I happened to have these drivers measured and ready for simulation, so I can give you a possible crossover solution for your drivers.
In series with the woofer a 2.4mH inductor, and in paralell a 10uF capacitor. Simple 2. order network.
For the tweeter also a 2. order with a twist....
Series cap 7.5 uF and in paralell a resistor and inductor (in series with each other) 3.6 ohm and 0.33 mH. The dc resistence of the inductor should be subtracted from the resistor value.
Say you have an inductor with a dc resistence of 0.6ohm , then the resistor should be 3 ohm. Not extremely critical, but it's a way to shape the highpass Q-factor to get a good phase transition between HP and LP.
Then in series with tweeter a 5.1 ohm resistor , and in paralell, forming an L-pad, a 10ohm resistor.
Both drivers should be connected with the same polarity.
This should give you a good starting point for some fine tuning. Maybe you have to adjust the series resistor for the tweeter to find the right balance that you like.
Remeber that this is a simulation and I have not listened to this, so don't be afraid to adjust the filter to your liking. This simulation works with a inter-driver spacing of 158.5mm center to center, and a baffle layout I also can send you, if you are interested.
Crossover frequency is 2.2 kHz, and system sensitivity with bafflestep compensation is 85 dB SPL @ 1m with 2.83V input.
Regards,
Bjorn Magne Idland
Lab technician
Seas Fabrikker AS
I happened to have these drivers measured and ready for simulation, so I can give you a possible crossover solution for your drivers.
In series with the woofer a 2.4mH inductor, and in paralell a 10uF capacitor. Simple 2. order network.
For the tweeter also a 2. order with a twist....
Series cap 7.5 uF and in paralell a resistor and inductor (in series with each other) 3.6 ohm and 0.33 mH. The dc resistence of the inductor should be subtracted from the resistor value.
Say you have an inductor with a dc resistence of 0.6ohm , then the resistor should be 3 ohm. Not extremely critical, but it's a way to shape the highpass Q-factor to get a good phase transition between HP and LP.
Then in series with tweeter a 5.1 ohm resistor , and in paralell, forming an L-pad, a 10ohm resistor.
Both drivers should be connected with the same polarity.
This should give you a good starting point for some fine tuning. Maybe you have to adjust the series resistor for the tweeter to find the right balance that you like.
Remeber that this is a simulation and I have not listened to this, so don't be afraid to adjust the filter to your liking. This simulation works with a inter-driver spacing of 158.5mm center to center, and a baffle layout I also can send you, if you are interested.
Crossover frequency is 2.2 kHz, and system sensitivity with bafflestep compensation is 85 dB SPL @ 1m with 2.83V input.
Regards,
Bjorn Magne Idland
Lab technician
Seas Fabrikker AS
Attachments
Hi All
Looked all the great info you guys have given me and I must say I am confused. When we ran the speaker specs thru some software up here, it kicked out a .72mh coil and 5.6 cap for a 2.5k crossover, yet everyone here is recommending 2.0+ for the coil. Now my understanding is that this would be due to baffle step compensation. But it is a pretty huge difference! Now, using the numbers that originally got me the .72 coil, 2.4 would give me a 800 mhz crossover point. Just a little low for a 2 way, me thinks!
Anything else I am missing here? Just wanna make sure I understand!
Thanks
Erik
Looked all the great info you guys have given me and I must say I am confused. When we ran the speaker specs thru some software up here, it kicked out a .72mh coil and 5.6 cap for a 2.5k crossover, yet everyone here is recommending 2.0+ for the coil. Now my understanding is that this would be due to baffle step compensation. But it is a pretty huge difference! Now, using the numbers that originally got me the .72 coil, 2.4 would give me a 800 mhz crossover point. Just a little low for a 2 way, me thinks!
Anything else I am missing here? Just wanna make sure I understand!
Thanks
Erik
The coil is so large because the bafflestep occurs in the 500 - 1200 Hz range. The large coil starts the rolloff gradually at low frequencies, and the capacitor increases the steepness around the lowpass crossover frequency.
To look at the bafflestep in your speaker, use this tool (Excel spreadsheet): http://www.pvconsultants.com/audio/diffraction/downloadbds.htm
BM
To look at the bafflestep in your speaker, use this tool (Excel spreadsheet): http://www.pvconsultants.com/audio/diffraction/downloadbds.htm
BM
This is one of the reasons why you never can use textbook calculations for crossover values that work in real life.
Because of the baffle step the speaker has a rising response all the way from about 200 Hz upwards. To compensate this is why you want to let the big 2.5mH (or so) coil kick in from about 300 Hz.
Speakers (in small baffles) without bafflestep compensation sounds thin and overly bright. The downside is that bafflestep compensation reduces efficiency.
EspenE
Because of the baffle step the speaker has a rising response all the way from about 200 Hz upwards. To compensate this is why you want to let the big 2.5mH (or so) coil kick in from about 300 Hz.
Speakers (in small baffles) without bafflestep compensation sounds thin and overly bright. The downside is that bafflestep compensation reduces efficiency.
EspenE
Re: Thanks!
Hi Erik,
You can get caps and coils from www.solen.ca . I dont know of any shops in the GTA though.
bigerik said:Wow!
So much great advice. Thanks everyone!
Bjorn, I would love to see the baffle layout you suggest.
Now I just have to fin caps and coils locally of those values!
Thanks again
Erik
Hi Erik,
You can get caps and coils from www.solen.ca . I dont know of any shops in the GTA though.
Update
Well, I did the woofer part of Bjorns recommended crossover tonight. Made a big difference allready. As to be expected, I have way to much tweeter. Still using my original part of the tweeter crossover (3rd order on the tweeter (7.1 cap followed by a 15 cap with a .29mh coil and a 2.2 ohm and 10 ohm resitor). Even put in a 12 ohm resistor instead of the 10 and that was still nowhere close to enough. Soooo, I will have to do the second half of the crossover too. Or put in a much heavier resistor in the L-Pad. But I can hear what sounds like some weird stuff happening in the crossover range anyway, so that will certainly not be the answer.
Gotta tune the port too at some point. Tuned to 50 right now. Should be about 45.
Thanks!
Erik
Well, I did the woofer part of Bjorns recommended crossover tonight. Made a big difference allready. As to be expected, I have way to much tweeter. Still using my original part of the tweeter crossover (3rd order on the tweeter (7.1 cap followed by a 15 cap with a .29mh coil and a 2.2 ohm and 10 ohm resitor). Even put in a 12 ohm resistor instead of the 10 and that was still nowhere close to enough. Soooo, I will have to do the second half of the crossover too. Or put in a much heavier resistor in the L-Pad. But I can hear what sounds like some weird stuff happening in the crossover range anyway, so that will certainly not be the answer.
Gotta tune the port too at some point. Tuned to 50 right now. Should be about 45.
Thanks!
Erik
Simulation Results (Bjorn)
Hi Bjorn,
I was able to duplicate your results in Lspcad, but the dZ on the woofer was 0. If the Proac's Erik inherited have a flat baffle the results aren't quite as good, the dip in the response between ~2 and 5k gets pretty ugly. Any idea of what dZ should be using these two drivers?
I uploaded a screen shot of my freq response with a dZ of 25mm, I figured a 1" offset would be pretty close.
It was darn nice of you to post your schematic for these two drivers. I am considering the CA 18RNX myself, for use with a T25CF001.
I'm kind of a rookie on this board, if my graph isn't visible I also uploaded it to my personal webspace...
http://home.comcast.net/~nikonguy/Bjorn_CA18_27TFFC.jpg
Best,
Steve
Hi Bjorn,
I was able to duplicate your results in Lspcad, but the dZ on the woofer was 0. If the Proac's Erik inherited have a flat baffle the results aren't quite as good, the dip in the response between ~2 and 5k gets pretty ugly. Any idea of what dZ should be using these two drivers?
I uploaded a screen shot of my freq response with a dZ of 25mm, I figured a 1" offset would be pretty close.
It was darn nice of you to post your schematic for these two drivers. I am considering the CA 18RNX myself, for use with a T25CF001.
I'm kind of a rookie on this board, if my graph isn't visible I also uploaded it to my personal webspace...
http://home.comcast.net/~nikonguy/Bjorn_CA18_27TFFC.jpg
Best,
Steve
Re: Simulation Results (Bjorn)
In the data I used for simulation there is already a driver spacing of 158.5mm center to center on a flat baffle. You could adjust the dY value on the woofer. Take bigerik's driver spacing and subtract 158mm. The dZ should not be changed because the different acoustic centers are already accounted for (it's in the phase part of the datafile).
Bjorn
nikonguy said:Hi Bjorn,
I was able to duplicate your results in Lspcad, but the dZ on the woofer was 0. If the Proac's Erik inherited have a flat baffle the results aren't quite as good, the dip in the response between ~2 and 5k gets pretty ugly. Any idea of what dZ should be using these two drivers?
I uploaded a screen shot of my freq response with a dZ of 25mm, I figured a 1" offset would be pretty close.
Steve
In the data I used for simulation there is already a driver spacing of 158.5mm center to center on a flat baffle. You could adjust the dY value on the woofer. Take bigerik's driver spacing and subtract 158mm. The dZ should not be changed because the different acoustic centers are already accounted for (it's in the phase part of the datafile).
Bjorn
Re: Re: Simulation Results (Bjorn)
Thanks Bjorn, I knew that I had to be missing something. I bet my CA18RNX-T25 XO will be *MUCH* simpler now.
Best,
Steve
bidland said:
In the data I used for simulation there is already a driver spacing of 158.5mm center to center on a flat baffle. You could adjust the dY value on the woofer. Take bigerik's driver spacing and subtract 158mm. The dZ should not be changed because the different acoustic centers are already accounted for (it's in the phase part of the datafile).
Bjorn
Thanks Bjorn, I knew that I had to be missing something. I bet my CA18RNX-T25 XO will be *MUCH* simpler now.
Best,
Steve
Re: Re: Re: Simulation Results (Bjorn)
I bet it will.
Gotta say, that is one nice sounding woofer. Even with my completely screwed up crossover as I listed above.
Would be interesting to measure these things now. I bet some weird stuff is happeneing at the crossover point.
Allrerady tho, they do sound great on good recordings. Totally unforgiving on bad ones tho. Once the tweeter is right, these guys are gonna be awesome.
Erik
I bet it will.
Gotta say, that is one nice sounding woofer. Even with my completely screwed up crossover as I listed above.
Would be interesting to measure these things now. I bet some weird stuff is happeneing at the crossover point.
Allrerady tho, they do sound great on good recordings. Totally unforgiving on bad ones tho. Once the tweeter is right, these guys are gonna be awesome.
Erik
nikonguy said:
Thanks Bjorn, I knew that I had to be missing something. I bet my CA18RNX-T25 XO will be *MUCH* simpler now.
Best,
Steve
Re: Re: Simulation Results (Bjorn)
Hi Bjorn
Measured my distances between woofer and tweeter and came up with about 147mm or so. Is the approx. 10mm difference throw things off a lot?
Thanks again
Erik
Hi Bjorn
Measured my distances between woofer and tweeter and came up with about 147mm or so. Is the approx. 10mm difference throw things off a lot?
Thanks again
Erik
bidland said:
In the data I used for simulation there is already a driver spacing of 158.5mm center to center on a flat baffle. You could adjust the dY value on the woofer. Take bigerik's driver spacing and subtract 158mm. The dZ should not be changed because the different acoustic centers are already accounted for (it's in the phase part of the datafile).
Bjorn
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Loudspeakers
- Multi-Way
- Would like advice on my new Seas project