Low tweeter output from passive crossover vs active

[TransformerNoob]

I am designing a 2-way speaker and when using the minidsp settled on a crossover of 3500hz. I purchased assembled crossover but they measure completely different than when active. Both are LR 12DB.

Initially when using active, I decided I would also attenuate the tweeter by 2.5db, as it was a bit too bright. With the passive crossover and no attenuation there is significant dropoff.

I bought 2 of the crossovers and tried both with the same result.

This is my first speaker build, is this normal behavior?

Here are the measurements.

Purple is active, green is passive.

Thanks!

 

[Mark Tillotson]

50dB per division makes those graphs really hard to compare. 10 or 5dB per division is much more informative.


And what are these things exactly? (And what impedance drivers)

 

[TransformerNoob]

50dB per division makes those graphs really hard to compare. 10 or 5dB per division is much more informative.


And what are these things exactly? (And what impedance drivers)

Here are with the tighter scales, active in purple, passive in green.

Tweeter: Dayton Audio DC28FT-8
Dayton Audio DC28FT-8 1-1/8" Silk Dome Truncated Tweeter

Woofer: Aurum Cantus AC-130MKII
Aurum Cantus AC-130MKII 5-1/4" Carbon Fiber/Kevlar Woofer


Crossover: Dayton Audio XO2W-3.5K 2-Way Speaker Crossover 3,500 Hz
Dayton Audio XO2W-3.5K 2-Way Speaker Crossover 3,500 Hz


Thanks

 

[madisonears]

Couple of issues.

Ask a mod to move this to "Multi-Way" speaker forum. You will get many better responses from experienced speaker builders.

No off-the-shelf passive crossover will give satisfactory results. If you want to stay passive, you need to do a lot more research and measuring and modifying of an off-the-shelf unit.

It appears that your crossover point with the active is too high. There is a large dip from 2k to 3.5k that your tweeter could probably fill quite easily. How did you determine 3.5k as the knee of the filter? Try moving it to 2.5k and see if it sounds and measures better.

Peace,
Tom E

 

[TransformerNoob]

Couple of issues.

Ask a mod to move this to "Multi-Way" speaker forum. You will get many better responses from experienced speaker builders.

No off-the-shelf passive crossover will give satisfactory results. If you want to stay passive, you need to do a lot more research and measuring and modifying of an off-the-shelf unit.

It appears that your crossover point with the active is too high. There is a large dip from 2k to 3.5k that your tweeter could probably fill quite easily. How did you determine 3.5k as the knee of the filter? Try moving it to 2.5k and see if it sounds and measures better.

Peace,
Tom E

Yeah I tried to cut a corner with an off the shelf unit. I'm sure a custom crossover would be better but this is for a portable setup so it didn't need to be perfect. I'm just not sure why the complete falloff after 3.5k, if it's affecting the impedance somehow. I figured it'd be fairly standard.

Typically the tweeter would need attenuation which I anticipated, but these results were unexpected.

I'll measure the components and see what they used, guess I'll be returning it anyway.

Thanks

 

[jdarg]

We're you driving with identical amps when active?

 

[TransformerNoob]

We're you driving with identical amps when active?

I was using a single 2 channel amp. Testing a single speaker.

One channel was for the tweeter, one channel for the woofer.

When passive, I was using a single channel of the same amp, with the xovers disabled in the minidsp.

 

[Mark Tillotson]

Its odd, the passive crossover seems to be without any resistors, so shouldn't attentuate the high pass (unless the tweeter impedance was completely wrong - that tweeter is 5.4 ohms resistive which might account for some small attenuation).


You have the passive crossover configured for 8 ohm woofer correctly? No obvious damage to it?
Perhaps capacitance meter across the high-pass capacitor to check its OK?

 

[TransformerNoob]

Its odd, the passive crossover seems to be without any resistors, so shouldn't attentuate the high pass (unless the tweeter impedance was completely wrong - that tweeter is 5.4 ohms resistive which might account for some small attenuation).


You have the passive crossover configured for 8 ohm woofer correctly? No obvious damage to it?
Perhaps capacitance meter across the high-pass capacitor to check its OK?

Yes, 8 ohm setting for the woofer.

I have tried both boards I purchased, both have the same result.

Capacitance between the input (+) and tweeter (+) is 5.53uf, which is the same measured between the woofer (+) and woofer (-).

Inductance measured between tweeter (+) and (-) is .76mh.

Inductance measured between input (+) and woofer (+) is .73mh.

 

[sgrossklass]

I'd try tracing out the tweeter connection while inspecting the underside of the board carefully. By the looks of it there's a large series resistance, which points to maybe a bad joint in the signal path. Stuff like that happens.

 

[TransformerNoob]

I'd try tracing out the tweeter connection while inspecting the underside of the board carefully. By the looks of it there's a large series resistance, which points to maybe a bad joint in the signal path. Stuff like that happens.

Unfortunately the Dayton crossover has a cardboard bottom layer protecting the pcb, so I can't see the connections.

 

[TransformerNoob]

Ok.....false alarm.

I reconnected my dac and used it's output, and now the measurement looks as expected!

I have no idea why the minidsp output is affected using the passive crossover.

I bypassed the xover settings for the minidsp, but something must be "off" with it.

Here is the measurement using the dac output.

MUCH better. Still needs a bit of passive tweeter attenuation but that's easy enough.

Thanks everyone for the help!

 

[TransformerNoob]

Attenuated measurement

 

[Mark Tillotson]

Much closer match to original purple.

 

[tuantran]

What is the value of inductance of L3 (air core) on the Dayton crossover network?

 

[tuantran]

I guess that the tweeter lost very high treble if the inductance of L3 is too small.

 

[tuantran]

Tweeter: DC resistance 5.4 Ohms, Voice coil inductance 0.09 mH