External passive crossover?

Scott75s1974

Scott75s1974

Well-Known Member
I have two sets of speakers, that's I'd like to use to complement each other.
Heath AS-2A's which I am in the process of re-capping, and a newly acquired set of PSB Imagine Mini's.
Those take care of tweeter and mids in a way the AS-2As never will.
So I lifted the jumper and am only using the 2As as woofers.

The PSB's take energy all the way down to 50hz, and that's wasted on them. I'd like to direct all of that to the 2A woofers.

Is it possible to rig an external crossover so the PSB's only get frequencies above 100 or 200hz, or whatever a good midrange cut off would be? If so, any advice on what that circuit would look like? The AS-2A just has a 6uf cap in series with the midrange. That would be easy enough to rig. Would that do any good?

I know bi-amping with an active crossover is the way to go, it's just not in my plans.

Thanks!
 
Last edited:
Register to hide this ad
So, using an online calculator for a high pass RC filter: 200hz w/ 6uf cap and 133ohm resistor. What would that do to the PSBs on top of their own well-designed internal crossovers?

Is it that simple? Or are there 4 orders of crossover complexity I don't know that I don't know? (I'm assuming "yes" on that last one.)
 
That puts 133 ohms in series with your speaker and you won't get any volume.

You need to figure that your speaker impedience is the "R" and calculate "C" accordingly. For 6 mfd, you would set the high pass at 3,315Hz (assuming an 8 ohm impedience), and that's not what you want.
For 200 Hz (again assuming 8 ohms) you need ~100 mfd.
 
Thanks, that's very helpful, since I don't really know what I'm doing. Would using 100uf do the trick? That's easy enough to try. Seems easy, too easy.
 
Last edited:
Ideally, you would use an in-line inductor on your woofer (AS-2A) and an in-line capacitor on your mid/hi speaker (PSB).

A 100 uF cap as noted above works for the mid/high speaker. A 6.4 mH inductor works for the woofer. Assuming each is rated at 8 ohms. You could mount each one to the back of the speaker and make your speaker termination to it directly, with a jumper from the other end to the built-in speaker terminal.

See https://www.diyaudioandvideo.com/Calculator/SpeakerCrossover/

Plug in the nominal speaker impedance and the desired crossover frequency. A first order filter should be adequate and not complex.
 
RTally's suggestion to use an external first-order crossover is a good one provided the internal crossovers are entirely disabled. Adding an additional stage will cause undesired and unexpected interactions with whatever is inside the speaker. So best to entirely disable the internal crossovers in favor of an entirely external unit OR disable drivers in the cabinets provided the crossover points are identical, which is unlikely to be the case.

If the Heath AS-2A has already been modified to eliminate the midrange and tweeter, and only use the internal woofer crossover, and the point is compatible with the PSB, then no external crossover work is needed for it.

If the PSB Imagine Mini is jumpered to remove the woofer and only use its internal tweeter and capacitor, and the crossover point is compatible with the AS-2A, then no external crossover network is needed.

Again, this is unlikely to be the case and your experiment will fail.

Picking the crossover point depends upon where the woofer and midrange cross. So you must have the value for the woofer inductor and run the formulas (see below) to make sure the both points match. Otherwise you will have overlap which will create humps in your response curve; that will sound awful. If you have the same frequency bands being produced by two separate cabinets you will have comb artifacts. Drivers must be co-located within 1/4 wavelength to avoid this problem. You will not have this with two separate cabinets.

In addition to what toddalin noted about volume reduction, adding impedance to the woofer or midrange circuit also interferes with damping factor, which is governs dissipation of the back EMF produced as the cone returns to its rest position. (Moving a coil in a magnetic field induces a current in the coil.) This is critical for a woofer and somewhat important for a midrange.

One difficulty with using larger capacitors is that the less-expensive non-polar electrolytic (aka NPE) results in distortion and limited lifespan. This has been well discussed on AK. If you do not have the cash to assemble a larger capacitor from smaller poly film capacitors, I suggest using a combination of film and NPE.

The problem with the crossover calculators is the formulas are generally not explained. Understanding how this works is helpful.

Given:
f is the crossover frequency (Hertz)
L is the inductance (Henrys)
C is the capacitance (Farads)
R is the driver's nominal impedance (Ω)​

The low-pass filter uses an inductor whose value is calculated by the formula:
f = R / (2π x L)​
The high-pass filter uses a capacitor whose value is calculated by the formula:
f = 1 / (2π x R x C)​

A band-pass filter combines a low-pass (for the ceiling) and a high-pass (for the floor), both calculated using the above formulas.

The formulas are given in radians which is why the 2π is used to convert to Hertz.
Divide L by 1,000 to obtain mH and C by 1,000,000 to obtain uF.

Both frequency points must be identical for the woofer and midrange to cleanly mesh.
 
The purple trace below is for the PSB Image Mini. The Mini is a two way speaker intended to be used with a subwoofer (the blue trace). It looks like the Mini has a high pass filter at about 90 Hz. The slope is steep. It could be a 2nd or 3rd order. It would be helpful to see a crossover circuit for that speaker.

1111psbsig.meas.jpg

https://www.soundandvision.com/content/psb-imagine-mini-speaker-system-ht-labs-measures

With this information, I revise my above suggestion. Don't do anything to the Mini's crossover. Run it as-is. But, as retrovert suggests, remove the crossover from the AS-2A. I suggest you try a 2nd or 3rd order high cutoff filter at about 90 Hz. If you want a higher crossover frequency between the two speakers, it would help to see what the Mini has for a crossover because you will need to modify or replace the woofer crossover.

To truly make this work, I suggest getting a good mic and software to measure the frequency response. You may need to adjust the values to avoid a hole or hump in the FR.
 
RTally said:
The purple trace below is for the PSB Image Mini. The Mini is a two way speaker intended to be used with a subwoofer (the blue trace). It looks like the Mini has a high pass filter at about 90 Hz. The slope is steep. It could be a 2nd or 3rd order. It would be helpful to see a crossover circuit for that speaker.
Click to expand...

I concur. My first thought from that plot (which I had not earlier seen) was third order. That will cause time misalignment when the woofer is first order, and it will also cause driver overlap because the woofer is rolling off at 6 dB / octave while the midrange/tweeter is rolling off at 18 dB / octave. That ends badly, with bleeding eardrums and headaches.

RTally said:
With this information, I revise my above suggestion. Don't do anything to the Mini's crossover. Run it as-is. But, as retrovert suggests, remove the crossover from the AS-2A. I suggest you try a 2nd or 3rd order high cutoff filter at about 90 Hz.
Click to expand...

Yeah, but you know, this isn't going to work. High-slope crossovers have all sorts of phase shift across the pass-band because the delay isn't uniform. Maybe better to replace the Mini with a 6 dB / octave.

The big reason is that the older woofers like those in the AS-2A (nee AR-2) were designed to gently roll off and merge with the midrange/tweeter. Crossover point is 1,000 Hz. Strange, as it was 2,000 Hz in the AR.

RTally said:
If you want a higher crossover frequency between the two speakers, it would help to see what the Mini has for a crossover because you will need to modify or replace the woofer crossover.
Click to expand...

Right. Replacement is in the cards.

RTally said:
To truly make this work, I suggest getting a good mic and software to measure the frequency response. You may need to adjust the values to avoid a hole or hump in the FR.
Click to expand...

Oh, that's a lot of work. I have a better suggestion. Upgrade the AS-2 with a soft-dome tweeter to become an AR-2x. First order crossover for woofer and tweeter. Or, better yet, add a midrange and make a three-way.

I see this hybrid, with unknown drivers in the Mini, as doomed to fail. Beyond the driver mismatch, the distance between the drivers will be high which means the sound takes longer to converge, and any comb artifacts or driver misalignment will be more prominent.
 
djnagle said:
For the price of good passive xover parts you can buy a MiniDSP and go active.....providing you have two amps.
Click to expand...

I like the way you think.

Retrovert said:
Oh, that's a lot of work. I have a better suggestion. Upgrade the AS-2 with a soft-dome tweeter to become an AR-2x. First order crossover for woofer and tweeter. Or, better yet, add a midrange and make a three-way.

I see this hybrid, with unknown drivers in the Mini, as doomed to fail. Beyond the driver mismatch, the distance between the drivers will be high which means the sound takes longer to converge, and any comb artifacts or driver misalignment will be more prominent.
Click to expand...

Yep. Or get a real-subwoofer and dial it in with the Mini and be done with it.
 
Thanks everyone for this detailed advice. I'm always astounded by the kindness of AK experts.
I don't know much about crossovers at all. This gives me some good things to learn (especially multiple orders and phase shifting).

I still plan to restore the AS-2A's. Ironically, I had placed the order for new tweeter and caps the night before I found the PSB's at an estate sale, having languished until I showed up on 50% off day. They were far too good a deal to pass up. So if the restored 2A's do it as designed, I'll find some other use for them.

They seem to take less energy than the mids and tweeters on the 2A's (or they are just working and have a higher db rating) so the mixed combo I'm running now has more energy to fill in some demanding spots of complex pieces.

I had initially miswired the 2A's to run only the mids and tweets with the PSB's. Oh my that was indeed bleeding ear bad! I don't hear any phase shifting problems now with just the woofers. But I'll give it a listen more attentive to that.

I like the 2A's for their real bass...not so sure about subwoofer range. There are a number of pieces I love that feature double basses and before listening on these, I've been missing so much of what that music is doing. (E.g. Part's Cantus in Memoriam Benjamin Britten)
 
RTally said:
The slope is steep. It could be a 2nd or 3rd order. It would be helpful to see a crossover circuit for that speaker.
Click to expand...
I wonder if the steep slope is more about the driver diameter - 4.5". When I run low freq through them, they certainly do some serious extensions as designed. So it doesn't seem like there's a huge drop in signal coming. They just don't make much sound below 50 hz.
 
You must log in or register to reply here.
Back
Top Bottom