DHAxovers for my '75 Khorns are done

[Deang]

The sound is on the lean side because you are running your squawker too hot. With a constant impedance design, tap 4 brings you 3dB down, but in the A or AA -- tap 4 brings you 6dB down. This silly fact is something I missed myself until BEC and Al spelled it out for me using my Crayons.

The squawker is roughly 110db/w, but it's a 16 ohm driver. So while the 8 ohm K-77 is getting 1 watt with 2.83V applied, the K-55 is getting a 1/2 watt. So, 3dB is effectively "lost" because of the K-55's higher impedance, and the other 3dB of attenuation is provided by the autotransformer.

You should be using taps 5 and 2, but you won't be able to do that until you rewire the DHA like the ALK.

[Kudret]

You should be using taps 5 and 2, but you won't be able to do that until you rewire the DHA like the ALK.

Thanks Dean. Does this mean DHA is no more?

So, do I need .3mH inductor 2.2uF and 6.2+1uF caps to convert them to ALK universal A s?

How about the networks you built for Craig? (EDIT: I also came across the schematic for KBL.) Where do they all fit in?

Kudret

[BEC]

Kudret,

Have you tried a Type A with all new parts yet?

Bob

[Kudret]

Bob,

I converted my AAs to As and tried those, but not with new components

(only 2uF caps were new). I'll get some 13uF caps and build

a pair of new type As with 2.4mH Solen and your 3636s. Thanks for

the suggestion.

Kudret

[Deang]

Kudret -- I didn't say build an ALK, I said wire the DHA like an ALK. I hope the attached helps.

[Kudret]

Thanks again, Dean.

Just to be on the safe side, is that 5 to + and 2 to - (ALK schematic shows the opposite)?

Tried it, vocals improved and I am hearing a full sound, but bass got loose now.

Kudret

[Deang]

The ALK has a phase reversal for the squawker that doesn't apply to the DHA. Also, and I forgot about this -- taps 5 and 2 apply to the T2A and UT 3619. Since you are using the UT 3636, you will be using taps 5 and 3 (squawker positive to tap 5, squawker negative to tap 3). Note the chart I supplied showing the tap settings for the UT 3636 in my previous post.

It makes sense that you would hear more bass, but I can't think of any reason why the bass should be "loose" now.

BTW, the filter I built for Craig is similiar except his has a 3rd order tweeter filter to accomodate more power (high SPL's).

[JohnA]

Dean,

Don't you have a math error? The K-55-V/PD-5VH is 107 dB on a

horn like the K-400 and 110.7 dB (at 1mW) on a plane wave tube.

If the T2A drops its output by 3dB on taps 0-4, then adding the

swamping resistor will bleed off additional voltage lowering it's

output further. I don't think I'd want it to be dropped too much

more than 3 or 4 dB, since the 107 figure could be called 108 easily

enough looking at the peaks.

I might want to change taps if I had a live room or one even deader than mine or my brother's.

[Kudret]

Dean, Here is what I've done (I hope it was ok to mod your schematic).

I have one more newbie question: If I want to try different attenuation levels, the 10ohms reistor stays connected across 0-5 and 0.2mH inductor goes to 5 on 3636, but squawker connections change to X-1 (for 5.3dB), Y-1 (for 3.9dB), and 4-0 (for 3dB), right?

Sorry for all these questions, but the chart (labeled autotransformer AT-1) I got from Bob for 3636 shows different tap settings - and I was using that chart to try different attenuation levels all along. []

Kudret

EDIT. I just looked at your diagram above and corrected the tap settings.

[Deang]

Math error? Me?

"The K-55-V/PD-5VH is 107 dB on a horn like the K-400 and 110.7 dB (at 1mW) on a plane wave tube."

Why Golly Mr. Albright, that's exactly what I used to think, but then that Crites feller started spout'in off about some Arkansas' Law called Ohm or sumpin like that, and he told me dat 16 ohm driver be only get'in half the power dat 8 ohm tweeter be get'in. So me be'in so good at math I got on one of them there ohmish calcerlators and figured out dat Arkansas Law wuz right! So now I tellz people:

The squawker is roughly 110db/w, but it's a 16 ohm driver. So while the 8 ohm K-77 is getting 1 watt with 2.83V applied, the K-55 is getting a 1/2 watt. So, 3dB is effectively "lost" because of the K-55's higher impedance, and the other 3dB of attenuation is provided by the autotransformer.

But yer know, that Crites feller is always try'in to confuse me, so I calls that Al feller -- and he starts talk'in bout batteries and light bulbs and all kinds of stuff, and the next thing yer know he starts tell'in me lot's of stuff about impertence -- but I thought he wuz gett'in a little personal if ya know what I mean.

So I finally get the whole thing figured out, and now you come in here mudding up the river.

"If the T2A drops its output by 3dB on taps 0-4, then adding the swamping resistor will bleed off additional voltage lowering it's output further."

Sounds like you need to talk to that Al feller and get yer lesson about impertance. I think a battery fixes it, but then you walk kinda funny for a while.

Seriously though, when I RTA the K-401 using a constant impedance filter -- 6dB of attenuation is "correct". I've spent enough time with the math and the RTA that I can say there isn't always a perfect correlation between the math and the acoustic response. All I know is that if I terminate the different parts of the network with resistors and measure the drivers independantly -- a 6dB attenuation level lines up the squawker with the woofer and tweeter response. When I measure the whole deal together -- it looks and sounds good. If I do the same thing with the Type A, the squawker level is slightly elevated but only a dB or two -- and only a few spots.



Kudret -- Look at your schematic -- you still show common from the autoformer attached to squawker negative.

[Kudret]

Kudret -- Look at your schematic -- you still show common from the autoformer attached to squakwer negative.

Dean,

I don't know what you are talking about.[]

Kudret

[BEC]

Dean,

The circuit Kudret posted above looks more like around 12 db of squawker attenuation to me.

Bob

[Deang]

"Sorry for all these questions, but the chart (labeled autotransformer AT-1) I got from Bob for 3636 shows different tap settings - and I was using that chart to try different attenuation levels all along."

That will never work. Bob's chart shows the way it works if you are using the first number as the input tap for the 13uF capacitor value. The second number is the output tap for the squawker. That's the chart you use for using the 3636 to rebuild stock Klipsch networks. Also, that chart shows the attenuation from an electrical perspective. To get what is happening acoustically you have to add 3dB of attenuation. In other words, if you are using input tap 5, and output tap 3 -- the autoformer attenuates by 6dB, but acoustically it's 9dB down -- 16 ohm driver = half the power in relationship to the other drivers = 3dB of loss.

Bob, -12dB? I hadn't really thought about it -- I just noticed the schematic was wrong for what he is trying to do. To be honest, with the resistor in place and the fact that he still has the squawker attached to common -- I have no idea what taps 5 and 3 do.

http://www.electroacoustics.co.uk/article/anon/fulldb.htm

[Kudret]

Bob, -12dB? I hadn't really thought about it -- I just noticed the

schematic was wrong for what he is trying to do. To be honest,

with the resistor in place and the fact that he still has the squawker

attached to common -- I have no idea what taps 5 and 3 do.

Squawker is not connected to common.

The schematic was wrong at frst, but it was corrected right away (see my [] above).

The resistor is connected across 5 and 0. Squawker negative is

connected to 3 and positive to 5. Everything should be ok now (-6dB attenuation and all), no?

Kudret

[BEC]

Dean,

I was figuring a few more db for the resistor and taking a guess at around 12 db total hooked up like that.

Bob

[Kudret]

What would be a direct approach to directly measure how many dB tweeter is down from squawker, etc.

How about using different test tones while keeping only one driver

connected to the crossover at a time and measuring the SPL? What kind

of test tones (frequency?) would be required for each driver?

Kudret

[Al Klappenberger]

Guys,

There seems to still be a lot of conufsion about the transformer settings and the swamping resistor on these netwroks. The 10 Ohm swamp dose not drop the voltage going into the transformer (5-0). It only draws addition current from the amp. As the amp is a voltage source, it doesn't drop the level any more than screwing in a light bulb makes another light bulb on the other side of the room get dimmer. The additional bulb just draws more current from the source. That's what lowering the impedance is all about. The truns ratio of the transformer fixes the attenuation to its ratio. The actual attenuation of THE SYSTEM is also a fuction of the impedance of the driver. Becasue the Zo of the K55 is about 13 Ohms you need to figure that number into the power equation to get the actual attenuation (E*E)/R. I think you really need to set the taps using instruments initially to get the squawker output level the same as the woofer output level. That is, a flat frequency response across the crossover frequency. Once you have done that you adjust it by ear to you liking. I think the setting will be the same on the DHA netwrok or my Universal network. Most people like X-4 or 5-2 with the K55 driver.

BTW: The phasing will probably be different between my Universal network and DHA becasue the DHA has fewer parts to shift the phase.

Al K.

[Deang]

Bob, -12dB, what are you thinking here? -6dB of attenuation from using tap 3, -3dB because it's a 16 ohm driver (1/2 the power), and -3dB because of the resistor? Hmmm. Well, the amplifier isn't delivering for a 16 ohm driver - it's delivering for an 8 ohm driver. Hey look, we got back 3dB!

I'm sick of impertince.

[BEC]

Dean,

Nothing has happened to the driver. It is still merrily swinging through its impedance curve and there is nothing you can do about that from outside the driver.

Bob

[Deang]

Baloney. You can't possibly believe that.