AA-to-A network with L-pad midhorn attenuation

[Erik Mandaville]

I just finished installing high wattage, variable L-pad attenuators on a pair of AA-to-A networks I worked on some months ago. We have discussed whether this is a suitable and hopefully easy and effective way of controlling and attenuating the output on the midrange horn, more often done by changing secondary taps on the autotransformer. As untraditional for Klipsch networks as this approach may be, it is a fairly common practice.

I can't not do something based on it's diverging from more traditional or 'accepted' approaches, and this might work very well. Then again, it might sound terrible. I'll report for those interested, as well as include a picture (which I've alrady taken, but can't figure out how to post it with my wife's camera. I'll try again!

Erik

[Deang]

I'm interested.

You don't need to defend your decision to try it Erik. I never had a problem with dropping the autotransformer, only the use of the pots -- and I did qualify the statement by saying it was based on my experience with one lone speaker. From what I remember, I don't think anyone in those threads was really critical of you giving a whirl. I think it's cool, and I hope it works it.

[Erik Mandaville]

Dean:

Thank you! I sincerely appreciate what you've said about this! Kind of makes my day, really...

Erik

[Deang]

Did you really think I was THAT critical? I'm sorry if I came off that way.

What do I have to be critical of anyway -- look at all the Klipsch networks I've ruined.

[Erik Mandaville]

Sorry it took so long to get this picture out here - but here it is!

[Erik Mandaville]

Yuk! sorry about that. It looked clear to me this morning in my workroom. Any way, the idea is there. I took the back plate off the attenuator, and it is securely screwed to the board, with the pad reattached and clipped in place afterward.

I have tried to hear a difference between it and the overall sound of the autotransformer and, if it is, I just can't really tell. Maybe the L-pad is slightly brighter sounding (???), but I can't confirm that. It works as I hoped, but caution and care is needed in blending the mid-horn with the tweeter. On other systems I've built, I just put my ear between the two drivers and turn up or down until the two seem balance, or provide the desired amount of attenuation. It's very easy to use, and it also makes it clear how important the midrange is.

As you can see despite the fuzzy picture, these are pretty robust controls, and have cast ceramic insulation for the wire-wound internal structures. They are heavy for their size.

I think more ears are needed for the final verdict as to whether they sound as good, while being easier to use, as the autoformer.

I have the system playing right now, and all I hear is a little more air on the top end, because I turned the midrange control down just slightly.

Erik

Oh! (I'm also now a proud owner of a Radio Shack SPL meter, which is going to be used with the Lexicon tomorrow!)

*I enjoy the Peach more by the hour!

[Deang]

Man, do I have to drive down to Texas and show you how to use that camera? What kind of camera is it?

[Deang]

With such flexibility from the variance in the controls, it's probably going to be close to impossible to make a direct comparison to the stock board with the autotransformer. Do you hear any "grit"? I guess not, or you would have mentioned it. So -- it works!!

[Erik Mandaville]

I've got more hi-hat and brushes, now! lots of air, and which must have been the result of an overly prominent midrange clouding tweeter response a bit. Doubtless the same thing can be achieved by changing the taps on the autoformer -- this just seems to sound as good to me (so far, at least), and is lots easier to experiment with!

[Deang]

To drop the midrange output correctly, you also have to replace the 13uF cap with a 6.5 cap. Your way is certainly less trouble.

Is it hard to balance the outputs -- or is it kind of intuitive?

[Erik Mandaville]

You're right, Dean -- a pretty lame shot (my students use 'lame' all the time to describe things that....well....sort of suck!)

Anyway, I don't hear grit, and you're exactly right about the intuitive adjustment. I'm just bringing the volume up from total silence to where it seems to evenly blend with the tweeter -- and then STOP!

Thanks to Born2Rock for getting us these! I think you're going to like this Craig L.!!

[Erik Mandaville]

just little things I notice --

Background noise is black as night! strange..... just nothing at all except for the slightest hiss from the tweeter.

[mikebse2a3]

Hi Erik

FWIW: One problem I've ran into with wirewound L-pads is over time the wiper contact can vary due to oxidation between the wiper contact and wire which just working the control from end to end will temporarly clear up.Yours look like really good quality and the ones I used where fairly good quality also but once they are set and not used over time the connection tends to oxidize over time since the cleanig action of using the control doesn't happen once a setting has been chosen.Maybe you will have better luck with them than I've had.

I like to replace the L-pad with resistors. Once I've settled on the setting I'm going to leave it at. I just take it loose and measure it for the resistor values I'll need to replicate it with.

I'm using the ALK crossover with the variable autoformer.Its really an interesting learning experience how just a small variance in the squawker setting will show up different things in the recordings by not just the tonal balance shifts but there is some masking effects that show up or nearly disappear depending on the settings.I've also noticed the same effects by how I use room treatments.

The center channel speaker you been experimenting with also is an eye opener about how we hear. If you think about it, by adding the center speaker you are changing things more than a person might would think.You will be changing the first arrival times due to the different locations/designs of the speakers(The lexicon if it has center channel delay will change this to some extent and I will be really interested to see how this changes your perception of the sound). You will have constructive and destructive interaction between the speakers.The speaker room interaction changes also.What I've noticed so far using a center speaker is something more natural sounding especially on vocals happens even when the spread isn't very wide (the wider the spread the more obvious the benefits of the center speaker are).

Thanks for posting all your experiments!

mike

[Erik Mandaville]

Hi, Mike!

Thanks for posting your thoughts on the L-pads and center channel. Sure, I agree with you on the issue of oxidation. Any potentiometer -- volume control, tone controls, balance control, etc. can become less effective because of oxidation. The point you made though, was that once the level is set, the wiper is stationary and will not provide the regular 'cleaning' action one would normally see in something that was more often used, such as a volume control. The same thing would happen, I would think, on input level controls of and amplifier or preamp. Once those are set, they would stay at that position, and only the 'master' control would be used.

With that in mind, I have 3 pair of speakers that I built between....lets see....1991 and '95 I think it was. All of those speakers are still used in separate systems -- well, wait....no, only two of them are being used. They all have fairly efficient tweeters that need attenuation to match the output of the mid-bass drivers. My intention was to one day change the variable L-pads to, as you said, fixed resistors, but they have been working so reliably that I just never bothered to do it. The pads I installed yesterday are also far more robust and better built than any I've used in the past, and look like they will provide long-lasting performance.

I can't help but think you have a very good point, though! I think the ideal thing to do would be to change the L-pads to fixed resistance once the 'best' output level has been found. And this last point has really to do with the primary reason I am using variable rather than fixed resistance. It's going to take a little bit of time and experimentation to find what the best level is. Once that has been done, I can always use a single pole, three position switch to provide three different levels of 'suitable' series resistance. Knowing my own behavior, though, I'm fairly certain that the crossover experiment is going to very soon be moved to a back burner for something else I've become fascinated with -- stereo listening with a center channel.

Once again, I appreciate the constructive way you went about making your point, Mike. That the center channel would in theory bring about some curious interaction between it, the room, and the other two speakers is likely. It's probably true, too, that careful measurement would reveal some oddities or other characteristics that might not 'look' so great on paper. However: I have no means to test our room response or sound quality other than my ears, which are of course ultimately what will decide on what sounds good or not so good. In this case, the 'good' brought about by the inclusion of a center channel is to such a degree that I have no intention of ever going back to two channels alone -- regardless of what theory or testing might disclose. What I have discovered is a sound so complete, lifel-like, palpable, etc. that it's almost uncanny! Last week I did an experiment where I turned the center channel off just after my wife left the listening room for a glass of water. When she came back, she looked at me, looked around the room, and said, "Did you do something?" "This sounds bad!" The sound she was referring to was the Klipschorns running by themselves in stereo-only mode.

Your point about constructive and destructive contributions of the center channel is exactly right. There are seemingly an endless number of compromises in all of this. I have also been told by non-horn believers that my system is compromised simply because of the fact that I'm using horn speakers rather than cabinets plugged with dynamic drivers. I have built speaker systems using some of the best drivers available -- Vifa, Audax, SEAS, etc., and I prefer the sound of high efficiency horns by far.

Thanks again for your thoughts, Mike!

Erik

[Daddy Dee]

Erik,

Looks like a good solution. I'd think the ability to adjust the mid output by less than 3db increments would be handy.

It is a robust looking element for use in the crossover. I'd suppose that all LPads are not created equal and that lesser quality could result in less pleasing results. But that comparison would be too much like work. Thanks for the post and doing the listening work for comparison.

[sfogg]

Erik,

"That the center channel would in theory bring about some curious interaction between it, the room, and the other two speakers is likely.  It's probably true, too, that careful measurement would reveal some oddities or other characteristics that might not 'look' so great on paper. "

Actually the opposite occurs when you steer a center channel out. You reduce the comb filtering (frequency response interaction) between the L/Rs as they no longer reproduce as much of the same central material and it instead goes to a single speaker. Also because the distance between the speakers gets closer (between L/C and C/R) where they do comb gets pushed up to a higher frequency and becomes less audible.

There are also other things that change to. In real life when you listen to a single person talking in front of you it is one source with basically two arrival times to your ears that are just about the exact same time if the person is front and center. If you listen to a reproducing of that person talking there are two sources playing with four different arrival times that do not match up. The same amplitude from L/R is what makes us hear this as coming from in front (even though it is playing off to the sides) but the ear/brain still hears the differences in the arrival times and is what makes it sounds less real. A single speaker front and center can mimic the arrival times of an actual person there, L/Rs can't do that.

Shawn

[Erik Mandaville]

Great response, Shawn. Thanks once again for your help with this! In crediting Mike's opinion and viewpoint, I mentioned that all I have heard can only be described as positive -- not just mildly positive, but unqquestionably so. I have heard differences in amps, and described those differences as not being subtle. The quality of sound we get with the well-balanced center channel takes that phrase (the difference isn't subtle)to entirely new level! It's extraordinarily good.

What you indicated as being an improvement in response is actually why I referred to the current arrangement we have as being stereo with center channel support or reinforcement. In other words, it's something that, while not calling attention to itself unless turned up too high, makes two channel stereo far better than it is on its own. The center channel literally seems to vanish, leaving the listener with nothing less than a very complete, tightly woven musical fabric. It doesn't seem to simply put a patch on what's missing, but incorporates the Left and right channels in a fashion that appears seamless. Pretty amazing, really, and totally unexpected!

I have a couple of questions I would like to ask about the Lexicon, Shawn. I spent time with the owner's manual early this morning, and I am bookmarked on the section describing channel balancing with the SPL meter. We are going out for some errands and a coffee, and I'm going to set the levels after we get back.

This little box has so many features is scare's me! (hey, I heard that! someone just mumbled to himself, "Yeah, it's got so many features I'll bet it doesn't do any of them very well..."

Somehow I very much think that is not going to be the case! The Lexicon appears to be very well built, and the digital domain of 0s and 1s seems naturally to be very different from analog. This 'might' imply that information can be modified and changed in a way that is virtually loss-less, meaning that 'sound shaping' can evidently be performed without compromising the essential integrity or quality of the signal. I can't wait to try it!

...and I'm going to stop with this!

Erik

[sfogg]

Erik,

"I have a couple of questions I would like to ask about the Lexicon, Shawn. "

Fire away here or through e-mail. Or send me your phone number and I'll give you a call.

As far as the channel balancing goes it is pretty straight forward. Set the meter to 70dB, C weighting and slow response and stand behind your normal listening position. Hold the meter about where you head normally is pointing up or 45 degrees forward.

Keep in mind the menus are just trees that you move through with the arrow keys. If you get lost just keep hitting the left arrow till you are all the way out of the menus.

With the Lex. on press the right arrow key... that will move you into the main menu. Move down to Setup and hit the right arrow again to move into the Setup menu. Go down to Output Levels and hit the right arrow to move into that menu. This will bring up Internal Noise Test... hit the right arrow again to move into it. I think it will start playing the internal noise test at this stage or you might need to hit the right arrow once more to confirm.

It will start playing the noise on your left speaker. When the noise is playing on your left speaker press the right arrow again to move into adjusting its output level. Now use the up down arrows to raise or lower that speakers level. Get it to 75dB on the meter. When it is at 75dB press the left arrow to move out of its adjustment.

The noise will move to the center channel. Again hit the right arrow to move into its adjustment and use Up/Down to get it at 75dB. Left arrow to get out of that adjustment and it will move to the right speaker. Repeat the above processes for any/all channels you have connected.

When you are done with this you left arrow to work your way back out of the menu tree. You can just keep hitting left arrow till you are all the way out.

This calibrates your Lex. to the proper SPL level at the listening position which it uses for a few of its processes. One of the cool things about calibrating like this is if either of us is listening to the same piece of music and I tell you I'm listening at -7 on the main volume control if you set your Lex. to -7 we are listening at the same volume at the listening position regardless of the differences in our amps,speakers, and distance from the speaker to the listener. The volume also gives us an idea of what the max SPL per speaker could be. If the main volume control is set to 00dB that means each speaker could hit 105dB SPL at the listening position for a 0dB Full scale signal on a digital input. Every dB you lower the main volume control is 1dB below 105dB. So IOW if you set your volume control to -10dB that means your peaks per speaker are 95dB (105 - 10)...-20dB means peaks of 85dB per speaker (105 - 20)..etc..etc. It gives us a very good frame of reference for volume.

If you want to tweak the level of the center relative to the L/R speakers do not do it in this menu. In each of the 'Effects' in the 'Effect Adjustment Menu' there are settings to tweak the centers level and also to adjust how it is derived from the L/Rs. That is where you would adjust the centers levels beyond the calibrated setting.

BTW, if you have just L/C/Rs hooked up right now the Lex. is all set to do the channel balancing. If you hooked up a pair of rear or side speakers you need to enable them before you do the level balancing. To do that enter Setup -> Speaker Config -> Custom Config then change either the rear or side (whichever you have your amp hooked to) and change it from None to Small and set the crossover point for them at 40,80 or 120hz. I'd probably just go with the 120hz as that will keep them from having to work hard and it will let the K'Horns get the most material possible for you to try the 'Bass Enhance' processing when you get to that step.

Also be sure to go into Setup-> Listener Position -> Speaker Distance and set the distance to each speaker for the time alignment to be effective.

"This 'might' imply that information can be modified and changed in a way that is virtually loss-less, meaning that 'sound shaping' can evidently be performed without compromising the essential integrity or quality of the signal."

It isn't lossless but well done DSP (with enough bits of precision) can do a heck of a lot before it really starts to compromise the integrity or quality of the signal.

Shawn

[mikebse2a3]

Erik said:

It's probably true, too, that careful measurement would reveal some oddities or other characteristics that might not 'look' so great on paper. However: I have no means to test our room response or sound quality other than my ears, which are of course ultimately what will decide on what sounds good or not so good. In this case, the 'good' brought about by the inclusion of a center channel is to such a degree that I have no intention of ever going back to two channels alone -- regardless of what theory or testing might disclose. What I have discovered is a sound so complete, lifel-like, palpable, etc. that it's almost uncanny! Last week I did an experiment where I turned the center channel off just after my wife left the listening room for a glass of water. When she came back, she looked at me, looked around the room, and said, "Did you do something?" "This sounds bad!" The sound she was referring to was the Klipschorns running by themselves in stereo-only mode.

---------------------------

Yes Erik!

Good test using the unsuspecting wife!!

I do similar things with my girlfriend. I like trying something and then without telling her what I've done or what changes I've noticed in the sound I'll get her observation. Since she doesn't get into this critical mode of listening(which is something that at some point in the experiment I try to turn off and start listening more relaxed(natural)mode like at a real event) its a real help hearing her observations against my own.

I believe just like you do that the ear will be the final judge!

Even though their might be things that don't look good in a measurement the center channel is providing needed information that without it the sound isn't perceived as natural as when the center is providing this information.

I(like you) am always trying things for the personal experience/knowledge that can be gained.

Like I said I'll be very interested in your observations with the Lexicon unit.

mike

[Erik Mandaville]

Wait!

How dumb am I, anyway!? (don't answer that!) I've got a pair of perfect speakers for the rear, and can use my small ASUSA 4 watt stereo amp for the rear channels. Neat!

BTW: Those Radio Shack jacks WERE going to be too tight, Shawn. Each ground barrell has four little collar/taps, and I neatly broke off one of those on each jack using needle nose. They fit fine, now.

Ok, I'll read your response when we get back, and will post what will probably amount to silly-sounding questions right here! Give me a potter's wheel, antique clock movement, or tube amp and I'm ok; computers and Lexicons frighten me!

I'll give you a call, too, if I hit a snag -- which Marie will probably be able to figure out for me. I just get sort of bewildered with complex remotes

the same way a caveperson probably would with a cell phone. Actually, I have problems with my cell phone, too! I accidentally speed-dialed (sped dialed?)our home number at school last week (didn't know it!), and Marie picked up! A student heard someone whistling on the other end (Marie trying to get my attention), and said, "Mr. Mandaville, somone is whistling on your phone!" I took the phone our of it's holder and said, "Hello?!" I was met with the sound of someone laughing so hard she couldn't talk!