CT 125 Tweeters

[Erik Mandaville]

I recieved our third CT 125 for a center channel upgrade, and once again was impressed with the quality of the horn Bob developed. The other two in the Klipschorns, with a lower crossover point of 4kHz, have provided much needed HF extension and presence, and this last one will allow for better matching between the three front speakers of our HT/music system.

Thanks,

Erik

[Daddy Dee]

Erik,

Yeh. I bet your system sounds good. Wish I could hear it. The 4000 hz crossover allows that little guy to show it's stuff. Pretty amazing for the price.

I'd also say that Bob's K-77 inspired horn sounds good. I've used the same tweeter behind a baby cheeks type horn and found no improvement over Bob's horn.

[Erik Mandaville]

Dee:

Absolutely. I've been aware of the availability of the driver, but the integration with the new horn is what is so nice. I have wanted to maintain the overall appearance of the Klipschorns, and what Bob worked out fits perfectly in that respect.

As we've mentioned in the past, too, the lower crossover plus band-pass in the midrange is an important part of the modification -- IMO.

Erik

[fini]

OK, so what do I have to alter on my AAs to change the crossover to 4K?

[Erik Mandaville]

"OK, so what do I have to alter on my AAs to change the crossover to 4K?"

Bob has worked out a 4.5K crossover for the new tweeter, which is only 5 kilocycles 'faster' than the one I made. Dee sent me those networks to try along with the new tweeters for a first audition, and I liked what I heard very much. Actually, maybe they were type 'As' (?) -- I can't recall.

[Daddy Dee]

Gregg,

Shoot an e-mail to Bob on that. For the 4500hz mod it's just adding a cap to change the value on the tweeter and an inductor on the squawker. I'm no help with the values.

Done with the right connectors it's easily convertible.

[djk]

The tweeter inductor on a AA is 245uH (nominal).

A 4Khz Butterworth is 3.3uF+250uH+10uF, the ALK is a 6Khz Butterworth.

[Erik Mandaville]

My own is a 4KHz 3.5uf - .24mH - 9.5uf Butterworth on the tweeter.

Narrow bandpass - (neg) 2dB 16 ohm L-pad (fixed) on the squawker (no autoformer)

The first experiment was an even-order design using 5uf in series with the tweeter and .57uf and 28uf (approx) in the bandpass.

I also made a second order reverse polarity design, and prefer the mixed order network using the 3rd order filter on the tweeter.

Erik

[Erik Mandaville]

Since the tweeter is run wide-open, you can treat it as a separate entity, connecting as mentioned above directly to the amp input.

However, if you wire a true bandpass, the values chosen need to reflect not only the low-pass crossover point, but for the high-pass, too. So some adjustment is needed, but it's not difficult. The autoformer adds a little oddness, though, because of the reflected load impedance concerns.

Erik

[fini]

I just want to do it the way it sounds best.[]

[Erik Mandaville]

"I just want to do it the way it sounds best."

Maybe the way it is will do just fine, but I would defer to Bob in terms of what he's done with this while maintaining the autoformer as an essential part of the design.

BTW: You could also convert to an 'A' and just use a single first order series cap on the tweeter that gives you the desired reactance at 4kHz or 4.5Hz. The suggested crossover for this tweeter is, if I remember right, 3rd order at 3.5kHz, so if crossing at 4000, it still might be better to use a sharper cut-off than what you get with 6dB/octave.

[Daddy Dee]

Gregg,

I'll be interested to hear your take on the sound after you make this mod.

After I did this, I was talking with Trey about how much better the Khorns sounded crossed at 4500hz he mentioned that Klipsch already knew that and had made that the crossover point in the AK4 crossover. (however, their crossover is sixth order or some such thing, which allows them to safely employ the K77 which otherwise would be endangered by such a crossover point in a Type A or AA crossover) What they had discovered was that the K55 just pooping out at 6000hz was not on it's best behavior. I don't think I ever noticed the K55 distortion at it's limit until I didn't hear it. The CT125 picks up there nicely and as Bob Crites says, "it's essentially bulletproof" in an A or AA crossover.

Simply using the CT125 as a drop in replacement for the K77 is a solid improvement. Using it in a crossover modded to cross at 4500hz, the improvement is sweet.

[Erik Mandaville]

4kHz Butterworth for the type AA:

1) Keeping everything else connected as shown on the schematic for this first part, you can remove the tweeter section from its connection point after the 13uf input capacitor.

2) Adding lead length if necessary, connect the wire you just removed from the downstream side of 13uf cap (where it connects to the autoformer) and reconnect it to the positive input to the crossover (or just to the opposite end of the cap -- either way it's the same).

3) Add 1.5uf in parallel with the first 2uf capacitor, and add 8uf in parallel with the second 2uf capacitor. (This may actually lower ESR, and possibly make a difference you can hear)

This is for the tweeter section.

Any takers on the squawker branch? A change needs to be made there, as well.

[BEC]

This works well as a first order crossover crossing at about 4500 to the tweeter and rolling off the midrange to match. Lots of people are using this design.

Bob Crites

[fini]

Bob, do you sell the caps and inductors, or is there a recommended supplier (and brand of parts)?

[BEC]

Gregg,

Yes, send me an email.

Bob

bobcrites@mac.com

[psg]

Bob posted a type A mod. Is a type-AA mod fully documented? I'd like to try this.

[Erik Mandaville]

I have the type AA modified for 4000Hz on paper, and will wire the midrange branch with the 13 uf cap and autoformer the same way he has done with the type 'A.' That's one part of this design I continue to ponder.

In any event, I have the needed capacitors, autoformers, and two chokes. The chokes are only slightly lower than what's needed, and I can add a few turns of magnet wire to reach the needed inductance.

The tweeter part as outlined above is pretty easy, though.

Who knows, maybe this will sound good enough to convert me back to Autoformerism.

Erik

[Erik Mandaville]

Actually, I've forgetten that I have already made this very network. It's the one I'm using now, but with a fixed L-pad rather than the autoformer. No matter, I'll make it anyway. You'll note that the tweeter branch will use the values posted by DJK (as well as me according to my own schematic), and the midrange branch will be done as Bob had done the type 'A' with the choke between the autoformer and squawker. I just have to use a slightly different value choke because of the lower crossover point.

This is a straightforward modification.

[Erik Mandaville]

Bob:

" This works well as a first order crossover crossing at about 4500 to the tweeter and rolling off the midrange to match."

Did you ever try:

1) remove tweeter from top of autoformer and reconnect to x-over input. Change capacitor from 2uf to approx 4.5uf (for an 8ohm driver)

2) calculate BOTH squawker bandpass elements (L & C) for 32ohms, and use about 1mH in series with 13uf but connect those IN FRONT of the autoformer rather than with the autoformer between them. The values of the capacitor and inductor in a first order bandpass are normally calculated for the same impedance, no? In other words, if the amplifier is looking at a 32 ohm load (reflected), what do you see happening if you use the inductance needed for that impedance? What you have done seems to work, and I liked the way it sounded, but I'm wondering if you considered this other approach.

Some small adjustment may need to be made to the 1mH as it relates to the autoformer, however I'm wondering if this was something you ever tried when developing the lower crossover point for the type 'A.'

Erik