Peerless 830869 + SB26ADC-04 crossover help?

[Mayuri]

Paul, it is fine to change the cap values. From your simulation I see we can get better FR with 15uF and 33uF than with 18uF and 33uF. I'm interested in getting the best response out of these, this is why the stepped baffle to physically time-align is also an option.

 

[Mayuri]

Bit more of crossover simulation. I'll try this one today as I got the parts for it at hand and measure drivers separately. Note that this is not optimal, lacks level matching etc.

 

[Mayuri]

Results from aforementioned filter. 1m measurements are gated 3,5ms, 30cm measurements are only smoothed 1/24 oct. Looks pretty decent.

Dropbox - filter1.mdat

 

[Paul Vancluysen]

Mayuri,

For your last simulation I have different results in Leap than yours.
Also your filter electrical transfer response is different for the woofer.

In Leap it looks not so good. The woofer has a strong peak at 1200Hz.


I will check also. It looks that your woofer impedance is different from mine.
I have attached my woofer impedance in the cabinet.

 

[Dissi]

Mayuri, your zma files do not contain any valid phase data, that's why your simulations are not correct. Attached are files that should work better.

 

[Paul Vancluysen]

Results from aforementioned filter. 1m measurements are gated 3,5ms, 30cm measurements are only smoothed 1/24 oct. Looks pretty decent.

Dropbox - filter1.mdat

The measurement looks conform the Leap simulation, but you measure a little higher level for the tweeter. Also the tweeter shape is a little different around 2kHz. It should fit better.
Maybe one time also measure again the driver responses without filter.
Then I can compare a second time with the models.

 

[Paul Vancluysen]

Mayuri, your zma files do not contain any valid phase data, that's why your simulations are not correct. Attached are files that should work better.

Hi Dissi,

yes, that explains the difference between the simulations

 

[Mayuri]

Here are measurements with and without filters at 30cm and 1m distances. Filter is the one in the attachment.
Dropbox - filter2.mdat

Mayuri, your zma files do not contain any valid phase data, that's why your simulations are not correct. Attached are files that should work better.

Thanks for the tip! However, when I try to load the files you provided into VituixCAD, it gives an error and says they don't contain frequency data.

 

[Paul Vancluysen]

Mayuri,

I have checked the Leap driver models versus the REW measurement filter 2. There is a good fit.
You can see in attach, models in color, measurements in grey.

 

[Mayuri]

Mayuri,

I have checked the Leap driver models versus the REW measurement filter 2. There is a good fit.
You can see in attach, models in color, measurements in grey.

Alright, so the simulations are correct. I'm going to make some more tests tomorrow when I get my soldering iron back (borrowed it to a friend for the a weekend, a mistake in hindsight). I'll check my own filter first and if it is unsatisfactory I'll order the parts for this filter:

Also I got the speakers with terminal cups atm, will detach them and use the hole for a test port as well (4cm diameter with 10cm lenght for 33Hz tuning). It should push the F3 a little lower.

 

[Paul Vancluysen]

Mayuri

Here the simulation with 15 uF and 33 uF. Coils are 3.9 mH / 0.30 Ohm and 0.22 mH / 0.82 Ohm. Z is minimum 3.25 Ohm. Serial resistor in tweeter filter 2.7 Ohm now.

This is the best I can make with this filter configuration.

Paul

 

[Mayuri]

Thank you kindly for your help Paul, Erik and Dissi!

I will post measurements of effects of porting and filters on the following week. I will probably put a small resistor in series before the tweeter filter to push level a bit down and to bring impedance up around 2kHz.

 

[Paul Vancluysen]

Mayuri,

I have made Leap models conform your REW measurement filter 2. They are a little different. See the SPL with filter V7.

Paul

 

[Paul Vancluysen]

Thank you kindly for your help Paul, Erik and Dissi!

I will post measurements of effects of porting and filters on the following week. I will probably put a small resistor in series before the tweeter filter to push level a bit down and to bring impedance up around 2kHz.

You are very welcome Mayuri. I find it always interesting to have some feedback of practical realizations of something on paper.

 

[Mayuri]

Mayuri,

I have made Leap models conform your REW measurement filter 2. They are a little different. See the SPL with filter V7.

Paul

Just to make sure: are you simulating this filter I posted earlier in here?

http://www.diyaudio.com/forums/mult...-sb26adc-04-crossover-help-5.html#post5315730

 

[eriksquires]

The minimum impedance is too low, and with many amplifiers may lead to a depressed output in the 2-5kHz range.

I would strongly suggest rethinking it, but it is impossible to do if you insist on keeping the caps.

Best,

E

 

[Paul Vancluysen]

Just to make sure: are you simulating this filter I posted earlier in here?

http://www.diyaudio.com/forums/mult...-sb26adc-04-crossover-help-5.html#post5315730

No, it is my last filter version 7, schematic in attach.

 

[Paul Vancluysen]

Mayuri,

Which filter are you going to built now? If it is not the last version 7, tell me. Then I do a simulation of the filter you will use with the last Leap models.

 

[Paul Vancluysen]

The minimum impedance is too low, and with many amplifiers may lead to a depressed output in the 2-5kHz range.

I would strongly suggest rethinking it, but it is impossible to do if you insist on keeping the caps.

Best,

E

Erik,

I agree, this is on the limit. With the last filter version 7, the minimum impedance is 3.25 Ohm at 2.5 kHz, impedance plot in attach.
The serial resistance of the 0.27mH coil has to be 0.82 Ohm.
I am looking to improve it. The problem is that this tweeter needs some filter boost between 2 and 5 kHz to become flat.


Paul

 

[Mayuri]

Surprise measurements! Got my soldering iron back, which means that I immidiately soldered together a filter (attachment) with the parts I got at hand currently. Tried out a port in place of terminal cup in this version, 4cm diameter 11cm long (should be 33Hz tuning). Ordered the parts for the filter v7 Paul made.

Dropbox - filter3.mdat