Heresy I, reducing squawker with autoformer?

[cfelliot]

In the original crossover let's say the driver is 16 ohms.

On tap 2 the impedance multiplier is 8.

This makes the impedance seen by the 2uF cap = 128 ohms = 16(driver) * 8(mult)

1 / (2 * pi * impedance * 2uf) = Fc = 621 Hz

============================================================================

Now move the tap down and put the 16 ohm across the driver.

The impedance of the driver || resistor is now 8 ohms.

On tap 1 the impedance multiplier is 16

This makes the impedance seen by the 2uF cap = 128 ohms = 8(driver||resistor) * 16(mult)

Fc is still 621 Hz

=============================================================================

[cfelliot]

One last item. A third solution is to just replace the 2uF cap with a 1uF cap (SoniCap $10) and move the tap down. No resistors to waste power.

[djk]

The resistor doesn't really waste power, not like the woofer does. The impedance will still rise to about 70Ω in the midrange with the added resistor on the lower tap.

The 1µF will work OK, but would not be as smooth on the lower midrange as the 2µF with 16Ω resistor. The resistor 'swamps' the impedance peaks caused by the horn.

[Deang]

I was just thinking how hanging off that last tap with no help might be a bit too much.

[cfelliot]

Waste may not be the right term. But there is no doubt that the 16 ohm resistor draws power. When not at resonance 1/2 the current is going through the resistor. When at resonance an even higher percentage of the total current goes through the resistor.What effect this has on the system is unclear to me.

I also know that this "swamp" is a shortcut method for performing this function. A full series notch filter LCR is what is generally used to supress an impedance peak.

It's times like this that I wish ALK was still here to explain this more fully. I have 3 Heresy Is that I intend to mod later this year with one of his universals which solves these and a number of other problems in the Heresy crossover.

[Deang]

Yes, you can use a notch filter, and the newer Klipsch networks employ them. They are modeled using LEAP in conjunction with an anechoic chamber. It's all very nice -- until you push the loudspeaker into the room, where most of the adjustments the LCRs are designed to handle are swamped by the effects of the room. They've dropped the autoformer, a superior method (and expensive) of handling attenuation, in favor of penny junk resistors.

Bob and myself have taken Al to task for years on the swamping resistor issue. Al claims it's only soaking up power that's not being used anyway, while the rest of us claim it plain just don't work that way. I guess it's no secret by now that I'm going to be helping Al out by taking over the Universal builds. We were talking on the phone and he mentioned seeing some recent posts of mine where I talk about the resistor along with his practice of removing the VC inductance from the traditional calculated low pass coil value. I told him I had issues with building things that I didn't think worked right. Two hours later, my mail box started filling up with plots and very well written explanations. He said it was important to him that people here had an understanding of how and why he did these things -- he wants people to understand his designs. However, I believe it might have been more for me than anyone -- I don't learn well verbally, having things in writing is more effective with me, and Al knows that. I'm at work right now, and can't download the papers -- I will get them out here tonight after I get off of work.

Some thoughts: I've spoken with Dennis (DJK) several times, along with the email -- and he insists that regardless of where the resistor is placed, performance improves. I would say that regardless of what one ends up thinking about it, consider that it is pretty much the only way for a person to adjust attenuation on the fly (for the bigger Heritage), which also allows them to try different drivers and horns. Very cool, right?

The inductance. This makes me groan, but he sent me plots, where he measured the difference with a microphone. It's miniscule. The transfer function improves doing it the conventional way, but there is more control over the impedance doing it his way. Al builds diplexers -- the top and bottom need to be mirror images of each other. So, who wants to present the argument that Al's designs sound bad? Good luck with that.

[cfelliot]

I started off my crossover exploration with the design for my MTM Heresy which you can find in this area. You all have way more experience than I, but I'm learning. Al contacted me indirectly and I have exchanged a large number of e-mails with him and had a number of phone calls. He's a great guy with lots of knowledge. I too have gotten the graphs etc which I prefer with my background as a software engineer. The old 1 picture is worth a thousand words thing.

I own both the Heresy I and the Heresy III. I have opened the crossovers to take a look at them. Yes, the autoformer is gone, but the IIIs absolutely blow away the Is in a number of areas as I have A/Bed them numerious times. I am sure the would be even better with higher quality components. I haven't been able to find a schematic, but will pop one at some point and trace it out.

I don't find the Heresy I's too bright, so attenuating the horns has never been an issue to me.

I think there is something to be learned from all the previous post here. I know I learned something. The bottom line is there are various ways to get to a similar results, but there's no free lunch. There are always tradeoffs.

[Daddy Dee]

cfelliot,

would you mind saying more about Heresy vs Heresy III? What do you hear as the differences?

[cfelliot]

More middle bass. This may be because the K28 is a 4Ohm driver, not sure.

Midrage is much more open.

Much better soundstage. On a good recording center is very well focused, spread is excellent.

It's not that the H1 sucks, it's just that the H3 is better.

[Deang]

They just pulled the midrange back a little and went to a higher order filter. John Albright's mod is killer (21uF/11R and dropping the taps a notch). Still owe the papers from Al, I just ran out of night - sorry.

[cfelliot]

They just pulled the midrange back a little and went to a higher order filter. John Albright's mod is killer (21uF/11R and dropping the taps a notch). Still owe the papers from Al, I just ran out of night - sorry.

By back a little do you mean the 800Hz crossover point?

I think the mid-range is on a bandpass too - one of the reasons I want to try the Universals on my H1s.

[Daddy Dee]

thanks cfelliot.

[Deang]

It's late I want to go to bed.

They dialed back the output of the midrange driver a little, it brings up the bass some.

The Universal Type A isn't compatible with the Heresy. You can however modify the Cornwall ALK to work with the Heresy. Then there's that new CornScala network Al has, it might work too, not sure.

[Deang]

First file.

xformers.pdf

[Deang]

The second.

XOshift.pdf

[Deang]

A nice picture.

[cfelliot]

Sitting in the airport this morning.

Dean,

These are Al's write-ups? Very good!

I should have been clearer in previous posts, meaning fractions of TOTAL current. But, total current does go up.

Al said the new Cornscala universal would be fine on the Heresy

[Daddy Dee]

Thanks Dean!

[John_Warren]

Here's how I believe the autotransformers behaves.

Let's simulate the measured impedance of the PD5VH loaded into a K401. Let's call this the "tap load".

[John_Warren]

With the tap load, we can apply it across any to the taps on the T2A and see what happens. This means we need to simulate the T2A which is easy to do give that we have the winding information.

Here's the Voltage response at tap 4 (-3dB) connected to the PD5VH+K401 tap load above and a 13uF cap on the source side of TAP 5. This is the AA configuration for the midrange.