For sale. AA crossovers. SOLD

[CECAA850]

From your picture you have the input cap and tweeter cap circuit connected to tap 4, squaker connected to tap 3.

I never touched the tweeter circuit.

[Axz Hout]

x

[Deang]

My eyes aren't that good. Carl, if Chris is right then that has to be fixed first. It would also go a long way to explaining the original problem as well.

Is the Dayton a 6.8uF?

[CECAA850]

I'll have to go back and find out what I was told to move. I remember asking several times and I'm sure I moved what I was directed to move. I think it was in the "Roadshow" thread. I may have been instructed to move the tweeter wire but I thought it was change the squawker cap and move that wire.

[Axz Hout]

z

[CECAA850]

You are correct, here's what Dean told me.

Okay, so they are just Type AAs, and can't be used in a "CornScala" without a little bit of work, and I mean very little.

Drop the squawker wire to Tap 3.

Drop the tweeter section to Tap 4.

Remove the black capacitors (they look like Bennics or Daytons), and replace with 6.8uF

[Axz Hout]

x

[CECAA850]

If wired correctly you need 4 connections to the T2A, you still have the original 3.

I'll point the buyer to this thread.

[CECAA850]

My eyes aren't that good. Carl, if Chris is right then that has to be fixed first. It would also go a long way to explaining the original problem as well.

Is the Dayton a 6.8uF?

Yes it is.

I told the buyer to check this thread, the crossovers are already boxed up.

[CECAA850]

The squaker wire is correctly wired to tap 3, however the input and tweeter are common and were set to tap 4. If wired correctly you need 4 connections to the T2A, you still have the original 3. You want a Type B wiring on the autotransformer, with a 6.8 uf input cap

This is exactly why I have had Dean build 5 crossovers for me. I know just enough to be dangerous. I can follow directions perfectly but have no idea why I'm doing what I'm doing. I've pointed the issue out to the purchaser. Thank you for your keen eyes.

Carl

[Deang]

lol, you've just never sat down and forced yourself to think it through.

With the T2A autoformer, Tap 5 is the input tap. That means the primary capacitor, which is usually the one furthest to the right on the board, or the largest, runs from input positive to Tap 5. This capacitor determines the first transition (400, 500, 600, or 700). Okay, simple enough. So, that leaves the output taps: 4, 3, 2, & 1. With each one of these, you get a 3dB drop in attenuation. You also get something else, a doubling of impedance back to the amplifier with each tap you drop. We'll get back to that.

For now, because of this particular problem, we want to point out that with the Type A and AA, the tweeter section is also connected to Tap 5, but remember, this is the INPUT TAP, and so the autoformer isn't providing any attenuation. All that is happening here is that PWK is running, or attaching the tweeter filter directly to the backside of the input cap. Why would he do this, I mean, why not just connect it to input positive like any other rational human being would? We'll get back to that.

Okay, so on the Type A and Type AA we normally see a 13uF capacitor. So now it's time to talk about that doubling of impedance thing. The K-55 is roughly 15 ohms at the crossover point. Attached to Tap 4, the amplifier sees twice that, or 30 ohms. You can use any crossover calculator, and it will tell you that for a 400hz crossover point, with a driver impedance of 30 ohms, you're going to need a 13uF capacitor. If we want to take another 3dB off, we drop the wire down to Tap 3, but now the impedance doubles again - to 60 ohms. When the impedance doubles, the capacitor value gets cut in half, so, 6.5uF, but 6.8uF works very nicely, don't you think? Remember, this cap has to connect from input positive to Tap 5, the Input Tap.

So Carl had these new CornScalas, and two sets of networks. Now, a CornScala is basically a Cornwall on steroids, and may be built with bigger horns, bigger and more efficient midrange drivers, and a super tweeter from hell if you want, but it all has to be brought down to match the least efficient part of the system, which is the bass. The K-33 in that box is roughly 98dB. The K-77-M is 104/105dB, and for the Cornwall, PWK knocked 3dB off, which is 101/102dB. The K-55 is 110dB, but you lose 3dB when you put it on the K-400, and another 6dB when you put it on Tap 3, which is 101dB. Keep in mind that PWK counted on this thing being against a wall or in a corner, and so he was counting on at least 3 - 4dB of room gain. This is very important to remember if you want a good, balanced sound out of these. When I owned Cornwalls, they sounded best placed within a foot or so of the corners.

Bass, 101dB

Midrange, 101dB

Tweeter, 101/102dB

With two sets of networks to choose from, Carl asked which he should use. Since the CornScala uses a 400hz crossover point, the Type AA made the most sense. The fix is relatively simple, we tap it out like a Cornwall, and then halve the primary capacitor value to keep the crossover point the same (400hz). When I learned more about what was actually in these loudspeakers, I talked Carl into letting me build him some battery biased Type B2's.

But now we have a new situation: a LaScala is not a CornScala. Btw, it really bugs me that people have taken "CornScala" to mean anything they decide to put in a Cornwall sized cabinet. The only "CornScalas" are the ones my friend Mr. Crites designates as such. Bob would never say it publicly because he has too much class, but I, who am somewhat depraved, ahem, I mean deprived in that area, have no trouble telling you all that I'm pretty sure he would much prefer that everything else simply be referred to as "my modified Cornwalls" - which is what Carl has. : )

Oh yeah, the LaScalas. Well, it has a 104dB bottom, just like a Klipschorn. Okay, it's not like a Klipschorn, but it's treated like a Klipschhorn. To be real here, we need to call it what it really is - a loudspeaker with the sensitivity of a Klipschorn with the bass response of a Quintet. Okay, it's not quite that bad, but you better have them very deep into the corners, and/or a good sub or two playing along, or they are going to sound pretty forward. If you don't have those deep corners, or the subs - you start looking at mods.

An easy and very effective modification is the one we've already discussed - you simply knock another 3dB off of the squawker. The slightly reduced output allows you to hear more of the bass, and it also lowers distortion, since the diaphragm is moving less than it was while on Tap 4 (it also reduces throat overload modulation distortion). So, the wire for the midrange gets dropped to Tap 3, and the 13uF capacitor gets changed to 6.8uF. The tweeter section stays attached to Tap 5, the Input Tap. Nothing else changes.

To fix this network and to make it LaScala ready:

1) move the other end of the 6.8uF capacitor to Tap 5

2) move the wire attached to the first 2.2uF capacitor of the tweeter filter to Tap 5

3) make sure the wire attached to squawker positive on the barrier strip is connected to Tap 3.

The LaScala starts to roll off at 100Hz. That means it gets its 104dB sensitivity rating somewhere above that. It's a rare person that doesn't believe the LaScala midrange is too hot. Since most of us are experiencing either aged induced hearing loss, or hearing loss from too many concerts, loud music in general, firearms, and engines -- a tweeter that runs 3dB higher in relationship to the midrange isn't necessarily a bad thing.

Bass, 104dB

Midrange, 101dB

Tweeter, 104dB

Edited April 26, 2014 by DeanG

[Axz Hout]

x