Heresy Crossover Mod - Less Bright

[Erik Mandaville]

And lastly:

Bob Crites offers very good quality autoformers with a greater number of steps of attenuation. We have an early pair of Heresies that I used his replacement autoformers for, and I'm keeping these entirely stock in terms of crossover design. I'm mentioning this in case you can't quite find what you need with the stock autoformer, which has 'courser' steps between taps.

The center channel I'm building will have variable L-pads for both the 4kHz tweeter and bandpass squawker.

Erik

[WMcD]

Thanks, Gil.

What you summarize is very interesting, it connects acoustic/electrical systems for me very well and is much appreciated.

Please note that the Olson and JBL links you refer to do not appear on your post.

Let me test my assumptions now to see if I understand...

If I attenuate more than stock at the autoformer, I am cutting voltage, increasing current; equivalent to reducing impedance. Couple that with changing the inductance, which requires a change in cap value to prevent the crossover point from shifting.

Almost correct. Let's assume you are driving with a constant level. As you move the output to lower taps, the output voltage (delivered to the driver) is lower. Every step, it is reduced by a factor of 0.707. The actualy output current does change but not because of the autotransformer. The current through the driver is set by Ohm's law I = V/R. We've reduced the voltage and thus the current must change. (Note that if the current went up by the same factor the voltage went down, the power delivered would be the same. And we don't have that.)

You are think about current capability in a transformer or auto-transformer. That goes up. But here we're not particularly concerned.

What impedance (let me say R) does change?

Answer: It is the R* looking into the autotransformer. It changes as we connect the load to lower and lower taps on the output.

This is a new concept. The transformer has an input and output (a two-port device). We're used to thinking that what happens at the input affects the output. But also, what happens at the output affects the input. Also, as we change taps, we're affecting the quanitative nature of the transformer. We're making it more of a voltage step down transformer.

As we make the tranformer more of a voltage step down tranformer, the input impedance ® goes UP. This means that there is less current at the input. Again I = V/R. Now that R is bigger and bigger (at the input) the current (and power) at the input goes down.

This makes sense because we have reduced the power at the output by lowering the output voltage. So the input power is lower due to less current with the constant voltage input.

The resistor across autoformer 0 and 5 gives attenuation of both mid and hi.

No. We're assuming a constant voltage at the input. Remember our R* at the input to the autotransformer? The resistor across 0 and 5 is just in parallel with that changing R*. Now we've got a new effective R** in accordance to the rules of parallel resistors. We're now reducing the effective input impedance to the auto transformer and that "swamping" resistor.

That resistance in parallel plus the autoformer's impedance in series act like a fixed, though perhaps frequency dependent, L pad for both mid and hi.

Almost correct. They're not in series, you were right with parallel. The driver's shift in impedance does, as you say, affect things because it changes the load on the output of the autotransformer.

- - - - -

We have to now step back a minute and consider the design of the crossover. Here we have a simple cap at the input. The effective crossover frequency depends on the load on the cap, and the cap value. And that load is the input R* to the autotransformer.

A) Suppose we want to keep the same crossover freq and the cap value. We can use the parallel swamping resistor to reduce the effective load.

Suppose we don't want to use a swamping resistor but maintain the crossover freq. Now we have to alter the value of the cap. Generally making it lower.

I am still collating reflected impedance....

We have to go back to the basic theory of a two-port device. Both the horn and the autotransformer are two port devices and there is reflected impedance in each. But for different reasons. (In an effort to be accurate, I may be confusing people.)

The horn is a two-port. I'm pointing out that the mouth condition (end of the transmission line) can cause a reflection which goes back to the throat, The fact that energy goes both aways makes sense because an ear horn is a speaker horn working backwards. The reflection causes a variation in impedance at the input.

The transformer has a reflected impedance too. We're not concerned with it here. But consider that if there was no load on the output, there would be no load on the input. If there was a heavy load on the output, there would be a heavy load on the input. In our situation with changing taps, we're actually changing the autotransformer, not the load on the output. That load on the output is just the driver(s).

For sticklers, the impedance transfer ratio of the autotransformer is being varied by the winding ratio by the taps. Which sort of what I was explaining.

I have little knowledge, that makes me dangerous.

Me too. Smile. Wm McD

[WMcD]

The links to Olsen and JBL were in an earlier post by another in the thread.

John A., long ago, asked whether I had changed a cam profile. Nope.

OTOH, I did early on manage to reset the timing of the spark to 5 degrees adc instead of 5 degrees bdc. This drove me nuts. Smile.

Wm McD.

[DrWho]

In other words, a 3 dB drop in output will be a 3dB reduction regardless of whether an autoformer or resistor (or pair of resistors or a variable L-pad) is used as the tool.

Not true.

[WMcD]

http://forums.klipsch.com/forums/p/59538/582694.aspx#582694

You can read why PWK went with the autotransformer rather than L-Pads.

I can't speak for Mike B, of course.

The basic issue for PWK is that resistors get in the way of the driver seeing the low output impedance of the amp.

Wm McD.

[Deang]

It also robs efficiency by increasing insertion loss -- almost 3dB. I have a Dope from Hope on that at work. Maybe Bob will save me the trouble and post it.

[Invidiosulus]

Al K sells an autoformer with a circuit boad and a jumper block built in so it's very easy to change between 0,3,5,7,and 9 dB of attenuation.

[Erik Mandaville]

"The basic issue for PWK is that resistors get in the way of the driver seeing the low output impedance of the amp."

PWK's ideas behind the use of the autoformer have been published here on many occasions, but the link provided will be helpful for those who might be interested.

I've experimented with all approaches to attenuation mentioned here -- autoformers and fixed and variable L-pads -- and find that resistive attenuation in no way degrades performance. I'm sure you're aware that many very well known and highly regarded speaker manufacturers (Altec was one mentioned above) used resistive L-pads. PWK didn't care for them, and I both understand and respect why. There are those on this forum who have made far more drastic changes to his original designs -- new horn lenses, new drivers -- that doubtless would have much greater impact on the sonic presentation of the original design than simply using a different means of reducing gain in one or more drivers.

I suspect that if an autoformer had such a clear advantage over a resistor, and that what they contributed to the sound of the speaker was a significant element in the design criteria, they would continue to be used today. As far as I know (please correct me if I'm wrong), the autoformer is no longer used. If variable L-pads were so inferior, why do so many other well-known brands, both in the professional and domestic worlds, make frequent use of them?

Again, each of us is free to use what works best, and that is why I proposed another means of even finer gain adjustment ability. BTW: the main driver in question above is the tweeter, which as I said can be separated from the autoformer and connected directly to the input of the network through a slightly larger capacitor, and an L-pad installed on the driver. I have done it many, many times over the past 20 years, and it provides not only entirely acceptable performance, but a much finer and directly comparable type of control, IMO, than an autoformer. Again, we have the freedom to choose, and that choice, IMO, does not have to be influenced by anyone other than ourselves.

Erik

[Erik Mandaville]

"It also robs efficiency by increasing insertion loss -- almost 3dB. I have a Dope from Hope on that at work. Maybe Bob will save me the trouble and post it."

I would rather you post the schematic of your passive Jubliee crossover.

Erik

[Deang]

I suspect that if an autoformer had such a clear advantage over a resistor, and that what they contributed to the sound of the speaker was a significant element in the design criteria, they would continue to be used today.

A resistor cost 50 cents, a good autoformer around 30 dollars -- it doesn't surprise me in least that they're not used anymore since maximizing profit is the priority now.

If variable L-pads were so inferior, why do so many other well-known brands, both in the professional and domestic worlds, make frequent use of them?

You'll find plenty of L-pads, but I'm pretty sure you won't find "frequent use" of variable L-pads. Feel free to post links to "other well-known brands" that use them.

I would rather you post the schematic of your passive Jubliee crossover.

Why, so you can blather about how it doesn't have an autoformer? I guess it doesn't matter that it doesn't have a fixed or variable L-pad either. The series resistor is optional, and I run mine unattenuated.

[John Warren]

A resistor cost 50 cents

NO IT DOESN'T!!!

It can cost from 5 cents to as much as $1.50.

When you come back with the precise, accurate answer, I will counter with yet, another ignorant statement based in subjectivity!

[Invidiosulus]

A resistor cost 50 cents

NO IT DOESN'T!!!

It can cost from 5 cents to as much as $1.50.

When you come back with the precise, accurate answer, I will counter with yet, another ignorant statement based in subjectivity!

Wait,

He said a resistor "cost" 50 cents.

Past tense.

So maybe they did cost that much when our speakers were being designed and built.

I paid at least $2.50 each for the swamping resistors I put in my crossovers!!!

You are both too cheap!!!!

[Deang]

A resistor cost 50 cents

NO IT DOESN'T!!!

It can cost from 5 cents to as much as $1.50.

When you come back with the precise, accurate answer, I will counter with yet, another ignorant statement based in subjectivity!

LOL! Well, I took a mean average. I pay almost $4.00 for the resistors I use, but I didn't think anyone here could handle hearing that without going into anaphylactic shock.

[DrWho]

One of the main reasons autoformers aren't used more frequently - especially in pro sound applications (like the markets of Altec and JBL) is that you get into issues with core saturation. PWK even mentions that it introduces distortion, but at normal listening levels in the home he measured it to be insignificant. When really cranking it, the autoformer starts to act like a compressor (and will sound like pants because of the hystersis).

An interesting behavior with resistors is that their resistance will change with temperature. I noticed it just yesterday at work while I was rebuilding a network for a pair of 15" 2-way McCauley mains. It is quite annoying trying to get the tweeter attenuation to be the same at all listening levels. Of course that also means the crossover point slides around too. Here I am striving for proper summation between drivers and everything is sliding around on me. That's hardly my idea of "fine-tuned". I didn't realize it was the resistors until I pulled out an ohm-meter.

[Invidiosulus]

Isn't that heat/resistance thing the way that the lightbulb tweeter protection circuit works?

[DrWho]

Isn't that heat/resistance thing the way that the lightbulb tweeter protection circuit works?

Probably.

http://www.partsexpress.com/pe/pshowdetl.cfm?&Partnumber=290-659

[skywave-rider]

Wow, Mike. Did the crossover get that hot?

Welcome to the old-world of tube amps. : )

The ear percieves so differently at different volume levels, that I think shifting

values in attenuators are not significant; assuming they stabilize eventually.

We're supposed to "work" a speaker at

what, 85dBspl? 90? Too loud for me these days.

At a certain echelon in pro sound, everything is active anyway.

I know u guys know that.

We're dealing with old world stuff here, but we like it.

I also note u guys love arguing. []

I'm still catching up.... []

[skywave-rider]

Isn't that heat/resistance thing the way that the lightbulb tweeter protection circuit works?

Probably.

http://www.partsexpress.com/pe/pshowdetl.cfm?&Partnumber=290-659

Urei used bulbs as fuses; but I never had to change one.

EDIT: Actually, if I remember correctly, the bulbs were only WARNING LIGHTS,

standard fuses were in line, tweeter and woofer. Sorry. Can't find the manual....

BULBS!

(fifth type)

[Dflip]

I've used them on the autotransformers from Universal Transformer of Texas. The typical 1/4 inch connector is the correct size for them. (Others might be familiar with the Al. K. design and the flexibility of using various taps on the autotransformer to adjust the level.)

I also fool around with RS 70 volt line transformers, using the primary 4-8-16 ohm taps as an autotransformer. These take a smaller push on connector which I found at an automotive "speed shop".

There is some need for caution. Sometimes the fit is too loose or too tight and the push on connector has to be fiddled with. If to loose, there is not a snug connection. If too tight, there is the danger that pulling the connector off will pull out the windings. I haven't ruined anything yet.

They sure are handy. Soldering and resoldering might damage the connections between the "tab" and the windings too. The push on connector is a nice compromise for the experimenter.

Least I forget. The cludgy automotive world has a push on "female" with two "males". These can be used when you have to connect two wires to the same tap. They cost a couple of bucks in a blister pack at the speed shop. No need to order from Parts Express.

I am a visual learner, I like pictures to help it make sense. I assume the items below are what you are talking about? I checked the listing at Parts Express just for the pictures for the connection to the autoformer. What size wire do you use for the crossovers?

Disconnect terminals:

for 18 - 22 awg wire for 14 - 16 awg wire

Spades come in #6, #8 size or #10 size, which is preferred, or will fit? These go on the terminal strips.

#6 18 - 22 awg wire # 6 same size wire #8, 14 - 16 awg wire

Thanks in advance for your assisance.

Don

[Guest " "]

Al K sells an autoformer with a circuit boad and a jumper block built in so it's very easy to change between 0,3,5,7,and 9 dB of attenuation.

since a new link was posted to this old thread, let me make a correction. The refered to ALK autoformer on a circut board with jumper blocks.....according to Al...is a transformer and not an autoformer.....I have a pair...and sent an email off to Al to inquire and he so replied.