Inductors for Heresy I's tweeter & squawker?

[Erik Mandaville]

And last idea experiment for attenuation: I was interested in working out a system that made use of the autoformer, but where one could toggle back and forth (by way of an external switch) between different taps on the autoformer. This one simply jumps between three autoformer settings, which I wanted to try without the simultaneous change in value of the input capacitor. It worked fine, and I did not notice any anomoly in performance other than the difference in attenuation of the driver. Again, one could use a swamping resistor to swamp the impedance so that capacitor values can remain the same. If one didn't want to use the resistor for whatever reason (I know some of the reasons), it's not difficult to include the change in value of capacitance, as well. One would just need the required number of caps to correspond with the number of switch positions. You choose the one you like best.

Erik

edit (apologies) well, okay. this is the type A, plus Bob's little inductor, plus swamping resistor. Anyway, you get the idea. There is a little switch box seen that I just put on top of the cabinet to experiment. I tried the same thing on a standard type A for the 6kHz crossover and without a swamping resistor. Swamping resistors have been around forever, by the way. I even used them on the Lowthers, which don't use a crossover.

I'm glad I can type quickly! Anyway, this is to show that there are possibilities regardless of method of attenuation (autoformer or fixed or variable resistive L-pad) They both do the same thing, though go about it differently. 3dB down is 3dB down regardless of chosen weapon -- at least that was the case for me. Some prefer to do that by heating a resistor, others by presenting a slightly annoying impedance match for the amplifier used. In the end, we really have the capacity to choose what we like best -- as with any aspect of this hobby and its chosen components.

[kg4guy]

Very cool networks.What inductors would you use for the Type B crossover I have for the Tweeter & Squawker.Thanks Eric

[Erik Mandaville]

They are examples of different networks used for 'very known' drivers. They were intended to show possibilities of attenuation for the midrange based on both the autoformer and more common resistive L-pad. I wanted to show both fixed and variable versions of both. The inductors you use depends on a number of things, but I'm hoping some others might offer their help, as well.

Inductors are not necessarily what you may be looking for. As a matter of fact, there are some schools of thought that would suggest that more inductors in a crossover are a compromise. Yet others will maintain that their sonic benefit is without question, and strive to design the higher order networks that require them. There are always different sides to a story.

My feeling is that you MAY have a perfectly good woofer, but need to balance it with the other drivers. You don't necessarily need more inductors to do that; and depending on what amplifier/s you're using, it might be a better choice to try to attenuate the HF and midrange first without adding extra parts to your crossover. Some here are very familiar with the type E network, and might be able to help with what would be needed to attenuate both the tweeter and squawker to your liking. If the T2A autoformer does not have quite the right level that you think you need, there is another autoformer available that should. Before you start adding inductors and changing the order of your existing networks, I would work with what's available to you right now.

Good luck!

Erik

[Erik Mandaville]

For Illustration purposes:

This is what a more conventional 1st order 3 way network looks like: There is a single series capacitor on the tweeter, the value of which is determined by the impedance of the driver and the desired crossover frequency. Incredibly simple.

The network for the squawker is called a 'band pass' (some refer to it as the pass band), the series capacitor and inductor values are determined again by the impedance of the driver and the upper and lower limit crossover points.

And there is the single series inductor on the woofer. So again, extreme simplicity. I've said many times that it's the simplicity of networks such as the E and the A that make a direct translation to a true band pass much more straightforward. These earlier networks did not include other compensation networks like notch filters, impedance equalizers, etc., and we generally know the behavior of the drivers in question (if they are the original Klipsch drivers).

The actual values you would need for part of the bandpass and the woofer inductor would be not far from what is shown, however I'm not familiar with the test performance of the woofer you're using. It may be that your woofer could benefit from a circuit that would boost its lower bass response, which consists of an inductor and resistor in parallel with one another and in series with the load.

I'm so used to getting this information from tables and charts in books I have, but this makes it easier. I've found some calculators that I prefer over others, but this gives you an idea. Note: There is NO attenuation in the crossover shown. You can calculate the desired reduction a couple of different ways, but also use a variable L-pad on both tweeter and squawker to find the exact volume you want. It does not matter to me in the slightest whether you prefer an autoformer or single or pair of resistors -- I have my own interests and preferences. I'm showing what would be needed to do this without the autoformer, which my preference (my own tastes derived from my own experiments).

You can see, though, that the connections are no more difficult than the autoformer, and no guess-work about what tap is going to do to the reflected impedance (which is an important part of calculating the needed values of capacitance and inductance.

Have a look for interest's sake -- just remember that your crossover point is not 750Hz, but 700 (at least that's what I remember as far as the type E).

Erik

Well -- I just tried to post this schematic and it didn't work. I have to write a test for my students right now but I can draw the thing out of my head later when I have time. Apologies......

edit: Bob, who already posted above, created a modification for the type A and AA that uses a lower crossover point using his new tweeters. It incorporates the use of an inductor in series with the squawker after the autoformer (since you did ask about that), as well as a change in capacitor value, and I don't know whether he has done the same for the type E you're using. I use an even lower crossover point on my Klipschorns, but I also don't use the autoformer. I prefer the use of resistors, so I'm going to defer modifications to autoformer networks to them. It's not too different, really, as long as one is keeping track of the numbers that come into play with different tap settings.

[Erik Mandaville]

Lastly this evening: If you have access to test data for your woofer (can't remember which one you're using, and I'm too tired to go back up to check), but you might see if their is an impedance curve available for it. If there is a rise in impedance, you can use a formula to design and impedance equalizer which may offer an improvement over what you have right now. Or maybe not. You need to know a couple of values related to the voice coil, which is often published by the manufacturer. Once you have those, the formula is easy, and the installation of the needed parts just as easy (you already know how to solder, after all!). I think I even found an on-line calculator for this, and beyond that you punch in a couple of numbers and let the computer do the rest. A snap. Piece of cake, even!

Erik

[kg4guy]

According to Eminence, the 15C is 98.1 db and claims 63 to 2800 frequency response.  Mine are based on the K-33 with sensitivity of around 96 db.

Bob  

The 15's I have are 94.4 dB 8 ohms what kind of increase can I expect from your 4 ohm 96dB's ?

[BEC]

According to Eminence, the 15C is 98.1 db and claims 63 to 2800 frequency response.  Mine are based on the K-33 with sensitivity of around 96 db.

Bob  

The 15's I have are 94.4 dB 8 ohms what kind of increase can I expect from your 4 ohm 96dB's ?

About 4.5 db. 3 db of that from going to 4 ohms.

Bob

[kg4guy]

Lastly this evening: If you have access to test data for your woofer (can't remember which one you're using, and I'm too tired to go back up to check), but you might see if their is an impedance curve available for it. If there is a rise in impedance, you can use a formula to design and impedance equalizer which may offer an improvement over what you have right now. Or maybe not. You need to know a couple of values related to the voice coil, which is often published by the manufacturer. Once you have those, the formula is easy, and the installation of the needed parts just as easy (you already know how to solder, after all!). I think I even found an on-line calculator for this, and beyond that you punch in a couple of numbers and let the computer do the rest. A snap. Piece of cake, even!

Erik

Thanks again Eric for all of your help.

[kg4guy]

According to Eminence, the 15C is 98.1 db and claims 63 to 2800 frequency response. Mine are based on the K-33 with sensitivity of around 96 db.

Bob

The 15's I have are 94.4 dB 8 ohms what kind of increase can I expect from your 4 ohm 96dB's ?

About 4.5 db. 3 db of that from going to 4 ohms. Bob

Thanks Bob

[Erik Mandaville]

Sure -

Hey, you may have already done this with your woofer cabinet, but some soft 'Acoustastuf' or poly fiber filling might be something worth trying. It can provide very real (meaning 'obvious') increases in efficiency, and this might help if your thinking bass efficiency or output could use a little help. Not sure if you're using a vented/ported reflex enclosure or a sealed one, but the way those are treated in terms of fill, varies. I used to use long-fiber natural wool, which worked really well, but tended to settle with time rather than stay fluffed up within the enclosure. Acoustic foam on the back and other walls might work well, too. To try either, you can go to those sort of home craft stores, which would have this stuff (cheap pun) in the sewing, fabric, etc. dept.

Erik

[Erik Mandaville]

kg4:

The above filling is something I tried in a very early pair of Heresies we own. That cabinet is really rather smaller, particularly in front-to-back depth, than the more common model I and II versions. I both lined the walls, and lightly filled the inside, and it helped. It was just 'tacked' in temporarily to try it out, but it's in the enclosures for good. This model Heresy, as well as Model Is seem to not use a gasket between the back panel and the surface to which they are screwed against, and I have wondered it this was done intentional as a way of providing a mild resistive vent effect, as a way of slightly increasing efficiency -- sometimes referred to as 'aperiodic enclosures.' They use a filled or damped port, which results in a sort of combination of the benefits of both vented and and sealed designs. Dynaudio used to sell add-on damped ports like this (I think I used to see them in the paperback Madisound catalog I had years and years ago). Maybe they still sell them. DIY versions consist of a small port cut in the cab, with something like 1/2" wire fabric on either side which holds the damping matrial in place.

Anyway, the point is that some internal acoustic treatment and stuffing might really help your situation -- if you haven't already tried it.

Erik

[kg4guy]

kg4:

The above filling is something I tried in a very early pair of Heresies we own.  That cabinet is really rather smaller, particularly in front-to-back depth, than the more common model I and II versions.  I both lined the walls, and lightly filled the inside, and it helped.  It was just 'tacked' in temporarily to try it out, but it's in the enclosures for good.  This model Heresy, as well as Model Is seem to not use a gasket between the back panel and the surface to which they are screwed against, and I have wondered it this was done intentional as a way of providing a mild resistive vent effect, as a way of slightly increasing efficiency -- sometimes referred to as 'aperiodic enclosures.'  They use a filled or damped port, which results in a sort of combination of the benefits of both vented and and sealed designs.  Dynaudio used to sell add-on damped ports like this (I think I used to see them in the paperback Madisound catalog I had years and years ago).  Maybe they still sell them. DIY versions consist of a small port cut in the cab, with something like 1/2" wire fabric on either side which holds the damping matrial in place.

Anyway, the point is that some internal acoustic treatment and stuffing might really help your situation -- if you haven't already tried it.

Erik

kg4:

The above filling is something I tried in a very early pair of Heresies we own.  That cabinet is really rather smaller, particularly in front-to-back depth, than the more common model I and II versions.  I both lined the walls, and lightly filled the inside, and it helped.  It was just 'tacked' in temporarily to try it out, but it's in the enclosures for good.  This model Heresy, as well as Model Is seem to not use a gasket between the back panel and the surface to which they are screwed against, and I have wondered it this was done intentional as a way of providing a mild resistive vent effect, as a way of slightly increasing efficiency -- sometimes referred to as 'aperiodic enclosures.'  They use a filled or damped port, which results in a sort of combination of the benefits of both vented and and sealed designs.  Dynaudio used to sell add-on damped ports like this (I think I used to see them in the paperback Madisound catalog I had years and years ago).  Maybe they still sell them. DIY versions consist of a small port cut in the cab, with something like 1/2" wire fabric on either side which holds the damping matrial in place.

Anyway, the point is that some internal acoustic treatment and stuffing might really help your situation -- if you haven't already tried it.

Erik

These are sealed and I think I am going with the 4 ohm Eminence Kappa 15C as it has a SPL of 98 and the woofer I have in is 94. SPL should make a nice change. I will make a powered sub out of the 8 ohm 15.that will for sure give me serious bass.

[Erik Mandaville]

Sounds great!

I hope it works out for you,

Erik

[IB Slammin]

Image above shows an autoformerless approach using fixed resistor-based attenuation. This is the mixed-order network I built for my Klipschorns with home-made coils (which I could wind for the EXACT amount of inductance I wanted -- since the precise calculated value isn't available). I know of others who solve that problem simply by buying a larger value, and then unwinding it as needed to achieve the inductance they want. The network above uses a 4000Hz crossover point for the tweeter at a slope of 18dB/octave, a very focused band-pass on the squawker, and a simple but very robust choke in series with the woofer.

This next one shows what a variable L-pad looks like. It's a network that in design is much like the ALK Universal (I built it with parts I had on hand as an experiment), but uses the variable 50 Watt L-pad shown with a black knob. I tried this with an extremely low powered output-transformerless amplifier, and the results were terrible. With a more powerful amplifier, it was much much better. The crossover point in this one is 6kHz.

Just to show you some alternatives. Since this type of control maintains the impedance of the driver (16 ohms) that's the value of impedance that's used in calculating the capacitance needed at the chosen crossover frequency. In other words, no need to change caps when all you want to do is turn down the volume on the squawker.

I gots ta ask. What is a fixed L-Pad? Makes me think of........ jumbo shrimp.

Thanks,

tc

[Erik Mandaville]

"I gots ta ask. What is a fixed L-Pad? Makes me think of........ jumbo shrimp."

Sure: A fixed L-pad consists of a set value of resistance in the form of a resistor -- like 25 ohms, 40 ohms, and so forth. The autoformer on the older Heritage networks also uses fixed, individual steps of attenuation, but goes about the process of attenuation differently. Each tap is essentially a fixed step in attenuation, but what changes is the reflected impedance load seen by the amplifier rather than simple resistance.

A variable (as opposed to fixed) L-pad operates much like a volume control, which is why it's referred to as variable. It's continuously variable in that the user is able to dial in the exact amount of attenuation needed/wanted and then stop. See the picture below? The two white rectangular components on the left, literally in the shape of an upper-case letter 'L' together form a 'fixed' amount of attenuation for the squawker. They perform the same function as the autoformer. I had used them in the past on other systems I built, but never on the Klipschorns. With both the A and AA designs, I preferred the resistors, which doesn't mean they are inherently better -- just that it's what I prefer.

It's possible to also use a single resistor in series, which is still a fixed value.

Erik

[Erik Mandaville]

Fixed L-pad detail:

[Erik Mandaville]

As opposed to a Variable type: This is a traditional 12/dB/octave Butterworth network I made a few years ago for our Klipschorns.

[Erik Mandaville]

And here is a sort of combination of both -- a kind of variable-fixed.

Some had commented on the difficulty of changing the amount of reduction on the midrange, which requires a fair amount of work, and, depending on how it's done, a change in value of capacitance.

This is a type 'A' network that uses a 3-position switch to enable switching on the fly between three different levels of attenuation on the autoformer. The switch box is kind of hard to see, but it's connected to the long leads that have been laced. This actually worked just fine. Despite the increased flexibility, the actual steps of attenuation are fixed.

This was also posted above.

slight correction: This is the type 'A' with Bob Crites series inductor modification included with the crossover point 4.5kHz. The swamping resistor is in place to offset impedance variations, which would normally require changes in value of the capacitor in series between the autoformer and input to the crossover.

[Deang]

That last network is a doozy, you actually use that thing?

[kg4guy]

Sounds great!

I hope it works out for you,

Erik

Thanks I ordered the Eminence Kappa-15c that should do the trick.