Crossover region for K-77 / K-77M tweeters

[Erik Mandaville]

I'm building a pair of first-order parallel networks with band-pass, and have calculated the filter in the tweeter circuit (8ohms at 6kHz). Is 6,000 Hz the factory suggested crossover point for the K77?

Thanks,

Erik

[boom3]

The T-35 series tweeters were used at crossover points between 4 KHz and 7 KHz. PWK originally used them at 6 KHz; the newer Khorn has dropped that to 4.5 KHz, to clean up some response problems in the midrange which used to have a nasty 9KHz spike, so I am told.

Unless you are going to have some other protective device(s), I am queasy about using this tweeter with a first order network. I'm not sure that 6 dB/octave will protect the tweeter well enough. I recommend you look at third or even fourth order, depending on home much insertion loss you can tolerate and what your crossover budget is.

[Erik Mandaville]

Thanks!

I have in fact been looking at a 12db/octave slope for the very reason you mention. The issue is that, except for a very few parts, I'm building this network from things I have on hand and have taken from other networks that have been spending time sleeping in a dark closet. I think crossing over at around 6kHz should be alright, but I appreciate your suggestion that there may be 'safer' ways of doing this. I have used the type 'A' network for a reasonable amount of time, which is a lower order network, without any problems.

I would actually be interested in an even higher order network, but those can get costly quickly, especially when using more expensive coils capacitors.

Thanks for this helpful information,

Erik

[mikebse2a3]

Hey Erik

EV suggest a minnimum for the T35 of 3500HZ and using a 12db crossover slope

I believe with a 6db slope I would stay with a 5000HZ to 6000HZ crossover point.

Also EV suggest long axis orientation od the T35 doesn't matter as much with a higher crossover point.

mike

[Deang]

Isn't the Type AA 12db/octave between the tweeter and midrange.

You're a turkey -- you're reinventing the wheel. John Albright's network has everything you're trying to build into this thing.

[Erik Mandaville]

Hey, Dean:

Actually, all I'm doing is building a band-pass, three-way, first order parallel network which is one of millions out there. John didn't do anything new, either, but I think that auto light protection on the tweeter is kind of cool.

I first started building my own speakers and crossover networks more than 10 years ago -- some were pretty good; some were really cruddy sounding. I happened to have just about all the parts I needed for this project on hand, and the majority of parts came from a pair of networks I built in about 1995.

I'm not trying to reinvent anything. I'm building a very common type of crossover to split the signal between three different drivers using parts I already have. That's it, nothing more. The one thing I'm not going to be able to do is account for the reactance between the series connection between the cap and coil, which will often result in slightly different values than calcuated going on the driver impedance and frequency alone. I'm going to get as close as I can, though.

But for Turkey, I'll take dark meat! It's always had more flavor to me.

Oh: John's network doesn't have the flexibility and fine tuning ability of a variable L-pad. Mine will.

I just thought it would be fun to do, since I have everything (almost) on hand. It's a way to make use of some decent parts that might sound half-way decent, as well.

Erik

[Deang]

You're a good sport.

Did you see my "L-pad" network in 2-channel?

[Erik Mandaville]

Hey, Dean:

Yah!

I just finished this pair of networks, but in building them in mirror images (talk about audio neurosis!?), I suddenly realized I did the same thing for pos. and neg. to drivers. So, I'm glad I discovered this and switched it, since that would have put the low pass branch into the tweeter!

I'll take a probably fuzzy picture tomorrow and show you. Actually I just have to connect the L-pads, which are supposed to be here tomorrow. But go easy on me, cuz I made the darn things with what I had! Maybe I should call it the 'Gumbo type 'A' (that's as close as I can get to a Mardi Gras mask!)

Erik

[Deang]

I wonder what's wrong with Marie's camera? Have you tried dropping it on the floor a few times?

[WMcD]

The T-35 by EV apparently thought by EV to go down to 3500 Hz, hence the name.

I have a scematic they used in the building block set up which is a second order, i.e. 12 dB per octave.

The A is a first order and the spec is that it is 6000.

The AA is a third order 18 dB with a Zobel and again we see the 6000 Hz spec. I don't have it just here. I find it strange that some people claim it is bad, causes sibalance, and surgery is necessary, etc.

For the record, EV made the T-35W as part of the Wolverine series. It seems to have a weaker magnet and less output. In the series the crossover is a first order. This may have been dumbing down for no good reason.

Best,

Gil

[Erik Mandaville]

Dean: I don't know what the problem is with her camera -- which probably isn't the case. I'm probably not doing something right...

I meant to bring my school camera home (very high quality Sony), but was so occupied with grading and getting exams finished up, that as soon as we were told we could go, I was gone! Forgot the camera.

Gil: Thanks for the extra information. It might be hard sometimes to improve on simplicity, and the type A might be a good example of that. I have tried two different stock AA networks, as well as one where I included more modern poly tubular capacitors in the HF branch, and in all cases it just sounded rolled off to me on the top end. I think part of that might just be because my hearing isn't as good as it used to be, the other networks I have tried, including the recent A version I made with a variable L-pad, sounded much more open.

I'm building this whole thing in my head, but using a table I have for desired values of inductance and capacitance at specific frequencies and impedances (voice coil values).

Erik