input on mod'ing ak-3 k-horn x-overs

[CECAA850]

i appreciate the offer but i think i'd like to keep the stock ak-3's incase i ever want to go back to original.

I thought you didn't like the originals? Max has made a great offer IMHO.

[Axz Hout]

z

[Max2]

ok lets keep this simple for the uneducated folks like myself. sorry but i cant follow the discussions above. so the links dean posted are not what i want? is there a change of opinion on the best way to get an extra 2 or 3 db reduction of the mids? i would prefer to just add caps as "mach-1" mentioned if thats teh easiest way.

max-2: i appreciate the offer but i think i'd like to keep the stock ak-3's incase i ever want to go back to original. i may be interested ion your ak-2's if the pirce is right... remeber i'm on a pretty tight budget so you might get top dollar posting them here or ebay. PM me details if you want. on a side note, i have read that for the ak series x-overs the 3's are supposed to be the "better" sounding ones. aside from the 4 or 5's. so the differences between the 2 & 3's is just 2 or 3 db's in the mids? wonder why they increased the 3's so much?

Thats no problem. Im not in any rush. I really don't know what the AK-2 upper Network and the bass caps are worth for an outright sale, but I will get a photo of them. Maybe Dean will know what a fair price is. They do already have the female terminals for your drivers so you could just plug them in on the second set of tabs on the drivers. You would however have to cut the wires on your current driver tabs to remove the upper network because they were soldered from the factory.

[001]

i appreciate the offer but i think i'd like to keep the stock ak-3's incase i ever want to go back to original.

I thought you didn't like the originals? Max has made a great offer IMHO.

because i dont like them doesnt mean i dont want to keep the originals. i do like the speakers as a whole & am aware they will probably sound much better in a bigger room or different gear etc. mainly, if i ever sell them i'd want the ability to return them to original. i agree it was a great offer...

thanks for posting the history info, nice to see when they went through changes. sounds like the driver output changed in 1989 so they made the db increase on the x over to compensate.

good to know something can be done, maybe an eq will be a quick non permanent way to play with the mid levels. i agree i dont like flat settings on the k-horns. they donet seem to be balanced as well as others, actually im fine with the "direct" (flat) mode on my fortes, they still have balanced bass.

[erik2A3]

There's another way of attenuating the output of any driver -- which will not change the reflected impedance and subsequent values of capacitance required in order to maintain the same crossover point. The single part needed (one for each channel) for this would be installed after the autoformer, using the same lead that right now connects the autoformer to the driver. The output lug of the device to which I'm referring is then connected to the driver, and one additional connection after that needs to be made. Obviously I'm talking about a variable L-pad, which would be chosen to match the impedance of the driver voice coil. One using this sort of attenuator just needs to be comfortable with this specific approach to attenuation, but in my view it's at least worth a try given the fact that in this case higher power amplification is being used and it seems only a mild amount of attenuation is needed.

We discussed the merits (or, depending on subjective opinion, lack thereof) of resistive variable L-pads for years and to exhaustion in the past, and I bring it up as a legitimate way of attenuating to output of tweeter and midrange drivers. Whether one decides they are effective in a given application would be a purely personal choice. It's used just as one uses a potentiometer to control volume on a preamp or integrated amp...and yes, P.W.K preferred to use autoformers in order to intentionally alter the reflected impedance (and thus calculated capacitance for a desired crossover point at THAT impedance) as a way to balance outputs of the mid and high frequency drivers of the three-way heritage speakers. Other companies chose variable L-pads. One could (and we sure have!) argue on behalf of one or the other.

[erik2A3]

Just for interest, this is one of many sites on the subject. Another possibility would be if you know the desired amount of attenuation you want (measured in Dbs) you can make a simple fixed L-pad designed for the impedance of the mid-horn driver. That's what the autoformer is doing right now, but the level of attenuation is not enough to suit your taste. So, you could make a small fixed L-pad with non-inductive resistors to pad the gain of that driver. The advantage of a variable L-pad (which may be a hassle for some) is that you can find exactly the amount of attenuation you want, and then measure the L-pad resistances at THAT setting in order to build a fixed L-pad.

Or use another autoformer with more taps from which to choose, and follow up with the cap values needed to maintain the same crossover frequency for the new reflected impedances -- as Dean mentioned.

Here's an example of a variable L-pad and an associated explanation:

http://www.colomar.com/Shavano/lpad.html

edit: note the example is for an 8ohm driver. They are also available for 16 ohms.

Edited March 15, 2014 by erik2A3

[001]

i do like the idea of a variable L pad, seen them on many speakers over the years & they seem to be a good way of accomplishing what im trying to do. if they dont affect the sound quality that would be a good option & could be easily removed if i didnt like it or wanted to return the speakers to original.

do any caps or other parts need to be changed with a variable L pad?

Edited March 15, 2014 by klipschfancf4

[erik2A3]

The L-pad on a tweeter or in your case midrange driver is not really part of the crossover network -- it's an added element used only to attenuate the output of the driver it's used with. It comes after the actual crossover components and will maintain the same impedance regardless of the position of rotation, high or low. You won't change values of capacitance since the signal will be seeing the same load. If on the other hand you removed the autoformer altogether (my understanding is that Klipsch no longer uses them on current production Heritage speakers....I may very well be wrong about that) and replaced it with a resistive L-pad, you WOULD need to adjust values of capacitance in order to achieve the same crossover frequency because the load impedance will have changed by the removal of the autoformer (Something I have done on several occasions). It works very well providing one compensates for the impedance change.

In your case, I thought of doing an initial experiment where you could get the desired attenuation without making any other changes.

As to sound quality, I've seen many arguments both for and against an autoformer. Some swear by them, others I've spoken to cringed at the idea of putting anything like it in front of a signal as a way of adjusting gain. I've done it both ways many times, and both were good enough sonically to use permanently. They are different, but I thought in your case, and based on your description, that having the ability to turn the level up or down while 'on the fly' would be helpful. If you didn't care for the sound, go the adjustable autoformer route.

But you're right. Many companies as well known as klipsch use both fixed and variable L-pads -- Altec, JBL, etc.

Edited March 15, 2014 by erik2A3

[Deang]

For years I've wanted to try that Fostex unit from Madisound, but at $180.00 each ...

I don't know if putting an l-pad between the driver and the autoformer tap works the way you think it does, I just know I started twitching uncontrollably when I read that. : ) Seriously, I don't know, so if he goes this route, I for one would greatly appreciate that since you suggested it, that you complete what you started and guide him through the process.

[001]

good to know the variable L pad wont change impedance or need other parts changed. any suggestions for good quality L pads yet at a reasonable price?

i'm still weighing all 3 options discussed here, seperate EQ is the easiest quick way to see if dropping the mid freq a few db helps, but i have read its the least desirable way to do this. the new auto transformers as dean suggested may be the best for SQ but requires the most work & buying many new parts. the variable L pad seems quite simple & if the theory will work as described i may try that if one can be found for a decent price. i like the ability to adjust "on the fly" or simply remove it to retain originality.

decisions decisions.

[Axz Hout]

x

[Axz Hout]

x

[erik2A3]

Dean, the L-pad should appear to the output of the auto-former similar to how it sees the driver without the L-pad between the two -- just as in for example, a tweeter circuit, where the L-pad is across the tweeter terminals but AFTER the crossover. Since the L-pad is a constant load that matches the nominal impedance of the driver, the technique I described above will work for the owner to get an idea of what lowering the output of the squawker will sound like without having to invest, perhaps more the once, in all the capacitors that would be required to correctly find the ideal setting on the auto-former.

[001]

" Any suggestions for good quality L pads yet at a reasonable price? "

http://www.parts-express.com/speaker-l-pad-attenuator-100w-mono-3-8-shaft-16-ohm--260-261

16 ohm? will that work for my 8 ohm speakers?

erik: any suggestions on the best value L pad to buy?

[erik2A3]

mach-1

Good, thank you for finding that part. It is what I was going to suggest. It's a pretty robust unit, and I used them on the board in the same location as the mounting location for the auto-former. Since you are experimenting for the time being, I would carefully desolder the connection from the auto-former output (that goes to the driver terminal), and replace it with a new lead long enough to allow the L-pad to be on top of the cabinet where you can work with it easily. By the way, you do not need to use heavy gauge speaker cable for this. Something a bit lighter and easier to work with will be fine. You have to use the correct #s 1, 2, and 3 L-pad connections, which is not a big deal. I'm glad you can solder!

Dean: what Fostex driver are you talking about?

[erik2A3]

Klipschfanc4f:

I'm sorry, what are you asking? all of what I mentioned was specifically related to a higher impedance driver. What is your question? Are you interested in trying an L-pad? I honestly did not mean to turn this into an L-pad frenzy, but.....for an 8ohm mid or HF driver, use an 8ohm L-pad.

[Axz Hout]

s

[erik2A3]

There you go. I don't recall the AK-3 network, but it may invert phase on the squawker. Just match the connections keeping that in mind.

[001]

mach: ok thanks for clarifying that, was thrown by the 16 ohm, didnt realize the k55 driver was 16 ohm.

erik: was just asking if you had a suggestion for the best valued L pad. looks like the one linked at PE is the best bet.

so besides confirming the phase of the squaker, it should be an easy mod. will update this whenever i get around to attempting it. thanks again for the input & advice.

Edited March 15, 2014 by klipschfancf4