Heresy I, reducing squawker with autoformer?

[rjsilva]

I read a post about reducing the output of the Heresy I squawker by 3dB by changing the position of the wire on the autoformer. Is someone able to give details about that?

[Fjd]

I read a post about reducing the output of the Heresy I squawker by 3dB by changing the position of the wire on the autoformer. Is someone able to give details about that?

 

It involves more than just changing the tap on the autoformer. John Albright did a modification of a Heresy Type E cross-over network in this thread; however, you have to, at least, read part way through the thread at the link below to see his final schematics on the topic. 

 

Look at John Albright's schematic in post #10 rather than the earlier posts.

 

https://community.klipsch.com/index.php?/topic/21969-heresy-crossover-mod-less-bright/page-1

Edited April 8, 2015 by Fjd

[rjsilva]

Thanks for the link, it was helpful! It basically says to move the squawker negative from tap #2 to #1 and the tweeter negative from #3 to #2, and add in a 16 ohm 10 watt resistor across the squawker terminals. The issue for me is that I'm not 100% certain which are taps #1 and #2. Can you clarify for my peace of mind?

[Fjd]

Thanks for the link, it was helpful! It basically says to move the squawker negative from tap #2 to #1 and the tweeter negative from #3 to #2, and add in a 16 ohm 10 watt resistor across the squawker terminals. The issue for me is that I'm not 100% certain which are taps #1 and #2. Can you clarify for my peace of mind?

If you have the T2A autoformer, I found this schematic. I'm traveling and do not have one that I can currently look at; however, I suspect that someone will wander by this thread and add to what I've found.

[cfelliot]

The 2uF connected to the tweeter is on the tweeter side of the autoformer. You can move this tap where ever.

The 2uF before the autoformer is the high pass for the mid-range. If you change the tap of the mid-range it will make the mid-range a dirrerent impedance and screw up the crossover point.

The crossover mod you a talking about puts a clamp on the input to the autoformer with the 11 ohm resistor. The 2uF cap is changed to a 21uF. Now the tap position of the mid-range can change with no effect on the high pass filter.

[rjsilva]

Thanks Fjd, I appreciate it. It looks right. I'll report back if it turns out to be different.

Did you read the thread that Fjd linked to cfelliot? I'm certainly no expert, but there was no capacitor change in that mod. There was much discussion in the thread about impedance issues. Did I miss something?

[cfelliot]

Look on Page 3 5th post!

[Fjd]

The 2uF connected to the tweeter is on the tweeter side of the autoformer. You can move this tap where ever.

The 2uF before the autoformer is the high pass for the mid-range. If you change the tap of the mid-range it will make the mid-range a dirrerent impedance and screw up the crossover point.

The crossover mod you a talking about puts a clamp on the input to the autoformer with the 11 ohm resistor. The 2uF cap is changed to a 21uF. Now the tap position of the mid-range can change with no effect on the high pass filter.

I was hoping that others would help clarify and did reference John's "final" schematic somewhat generically; however, I believe the 'final on page 1' will also work as it seems that the technique in John’s thread is essentially changing the driver’s

impedance by adding the resistor in parallel across the mid-driver to off-set the change in impedance in the autoformer when you drop down one tap on the autoformer and the

autoformer impedance changes. The 'net' impact of the two changes would appear to substantially off-set each other in the context of the network.

Why couldn't this technique, if the resistor is added in parallel to the driver, for a specific 'ohm value' that when inserted in conjunction with an autoformer tap change, would create a 'net match' of the old impedance previously seen in the network (see my example calculation below....Note that 'turns ratio' used in my calc was taken directly from the Klipsch Type E schematic attached below for the specific attenuation level), be

used to change taps, thus reducing the output of the mid-driver without replacing capacitors as per John's final page one example? In this example, why would the cross-over point change, thus necessitating a capacitor value change when we seem to just get back to the same starting point?

Is this a similar method

of adding a resistor in parallel directly to the autoformer as used by ALK and others (I may be missing something specific to the individual designs since I haven't studied the designs of either in detail) since the ALK design does not require a cap change everytime a tap on the autoformer is changed?

Attenuate 12 dB
Mid-driver impedance		11	ohms
autoformer impedance (Turn ratio)		16	x
New impedance		176	=
Desired impedance		8	ohms
Parallel Resistance formula
=1/((1/8)-(1/(11*16)))		8.38	resistor ohm value
Attenuate 6 dB
Mid-driver impedance		11	ohms
autoformer impedance (Turn ratio)		4	x
New impedance		44	=
Desired impedance		8	ohms
Parallel Resistance formula
=1/((1/8)-(1/(11*4)))		9.78	resistor ohm value

Since total parallel resistance is calculated as 1/Rtotal = 1/R1 + 1/R2 + 1/R3 +...... , in my example I just try to solve the equation for the unknown.

[rjsilva]

Look on Page 3 5th post!

Oh I see. I don't want to order more parts (I've already put more than I wanted into these Heresy's) so I'd prefer to stick with John's corrected version on page 1. Is there some problem with that version? He does say "The second schematic is for experimental purposes, but could be used indefinately"...

[cfelliot]

But, there is an another issue. The standard E/E2 crossover has a huge impedence swing due to the multiplied impedence of both horns.

The resistor across the driver fakes the starting impedence to a lower value just to be multiplied up by the lower tap.

You still have the huge impedence swing.

Adding the 11 ohm resistor across the autoformer smooths that out so overall impedence becomes pretty flat.

In the days of the Heresy design a 21uF cap would have been an expensive item.

This is a sweep of one of my original Heresy Is:

[djk]

Adding the 15Ω~20Ω resistor when going down a tap allows you to use the original capacitor values and retain the original crossover points.

The 11Ω resistor mod requires the cap be much larger and more expensive for only a minor improvement IMO, it also makes it difficult to go back to the original taps should you change the speaker placement and need to change back.

YMMV (I just changed a pair for a friend with the 16Ω resistors and he loves the change).

[rjsilva]

Ok, to clarify, John's final mod replaces the squawker cap with a 21uf cap (a film cap in that size would be expensive, no?) and solders an 11-ohm 21-watt resistor across taps 0 and 5 on the autoformer?

[Fjd]

But, there is an another issue. The standard E/E2 crossover has a huge impedence swing due to the multiplied impedence of both horns.

The resistor across the driver fakes the starting impedence to a lower value just to be multiplied up by the lower tap.

You still have the huge impedence swing.

Adding the 11 ohm resistor across the autoformer smooths that out so overall impedence becomes pretty flat.

In the days of the Heresy design a 21uF cap would have been an expensive item.

This is a sweep of one of my original Heresy Is:

Is it fair to say that the resistor only off-sets the change in autoformer impedence and the cap change fixes an inherent design "flaw" in the original schematic?

[cfelliot]

...it also makes it difficult to go back to the original taps should you change the speaker placement and need to change back....

Hey I don't pretend to be an expert on this, but my understanding is the the 11 ohm and 21uF locks the crossover point and you are free to put the squaker on whatever taps you wish.

[cfelliot]

Is it fair to say that the resistor only off-sets the change in autoformer impedence and the cap change fixes an inherent design "flaw" in the original schematic?

The resistor on the autoformer "clamps" the crossover point to itself vs. the reflected impedence of the driver which is tap dependent.

Not a flaw, just the original design.

[cfelliot]

DJK,

I just did the math and my understanding was incorrect.

Yes, if the tap moves the 11ohm resistor across the autoformer will need to change to maintain the same xover point with the 21uF cap!

[Fjd]

I don't want to order more parts (I've already put more than I wanted into these Heresy's) so I'd prefer to stick with John's corrected version on page 1.

For a direct answer to your question, given djk's recommendation for using a 15Ω~20Ω resistor (I've read hundreds of his technical audio-related posts on various forums and have never found where he has been incorrect) and by my working through the math in my other post on why it should work, I think you should be fine using the 16Ω resistor in the same way that John used it in his final schematic on page 1 of the link.

Adding the 15Ω~20Ω resistor when going down a tap allows you to use the original capacitor values and retain the original crossover points.

YMMV (I just changed a pair for a friend with the 16Ω resistors and he loves the change).

After working through the math (instead of just punching it into a calculator), I see why the resistor is often referred to as a "swamping" resistor.

[cfelliot]

After a bunch of though last night on this issue, I realized
the following. Please correct me if you think I’m in error.

Either case of adding a resistor 1) across the auto former; 2) across the driver; does not give you the freedom to move the
output tap wherever you want.

The original schematic of putting the 16 ohm resistor across
the driver was an attempt to half the impedance of the driver because moving
the tap down will multiply the new impedance by 2. Doing this will restore the
original impedance seen by the 2uF cap and keep the same xover point.

The final schematic parallels the reflected impedance with
the 11 ohm resistor, but requires a new value 21uF cap. The total impedance of
the system is flattened to around 10 ohms. There seems to be a difference of
opinion as to whether this benefits the load on the amp.

Because the tweeters filter cap is on the output side of the
autoformer, the crossover point is not altered by the tap placement.

[Deang]

Everything above is correct. Nicely done.

[Fjd]

Either case of adding a resistor 1) across the auto former; 2) across the driver; does not give you the freedom to move the
output tap wherever you want.

The original schematic of putting the 16 ohm resistor across
the driver was an attempt to half the impedance of the driver because moving
the tap down will multiply the new impedance by 2. Doing this will restore the
original impedance seen by the 2uF cap and keep the same xover point.

I also came to this same conclusion when I did the math problem on the prior page in that it seemed to show me that depending on the tap choice, a different resistor value would need to be chosen to restore the original network impedance. Since I did not know the original network impedance, I just 'plugged' 8Ω as the desired impedance for illustration purposes. I haven't tried the math yet, but I suspected that by selecting a resistor in the 15Ω~20Ω range, it may 'get you in the ballpark' on a 'net' basis impedance-wise and not 'materially' change the crossover point?