The B&C DCM50. Can it replace the K55 driver?

[jwc]

I will have to agree that using an autoformer to attenuate to a "large degree" isn't satisfactory. I've had two experiences in which I was trying to attenuate a High SPL driver to match a crites tweeter. One was the Altec 902-16B. I was having to attenuate 6-7 dB and the sound quality of the mids seemed to deteriate a little. This was a while ago so don't remember the details.

The second time was very noticeable. This was to attenuate a JBL 2446 on a 223 Tractrix horn to match output with a crites tweeter. This did not work. The sound from the mids were poor. I switched to an active setup and triamped the system. That sounded very good. You could say that the active system was just gonna sound better anyway.....I don't know. (BTW...the Crites tweeter sounded very good with an active setup.)

The above stated is obviously unscientific by any means. Most people have autoformers on there networks around here, so I guess its role will play a part.

I have no tendency as of now thinking one driver is "better" than the other. The focus of discussion seems to be a battle of drivers. Again my intial point is one of Horn vs Horn......instead of driver vs driver. I'm just curious of what qualities that Trachorn 220 will possess that will outperform the V-trac. You may have points that don't need to be shared and keep secret....I can appreciate that. I'm just curious of the pluses....such does it have a better throat to match the driver?, is the baffle depth or construction "better"?, or is just the whole construction/build gonna make the difference? Is there a thought of shape expansion or other design inputs to improve polars? maybe it is a combination of all.

Also again.....this isn't a post of criticism......just trying to understand.

jc

[Coytee]

Probably a dumb question but wouldn't the problem of attenuating that driver be mitigated if someone was using an active crossover?

I know there are many fans here of various passives (me one of them as I have Al's ES networks in my LaScalas) but it seems to me like there is finally some glacially slow movements of people and conversations about going to larger horns and now, larger drivers. By this, I mean we now have Al's Trachorns, Greg's horns, Gothovers horns (am I forgetting anyone?). Now, Al is coming out with Trachorn II so the choices are getting even better.

I would think an active crossover would help this problem of mixing & matching hyper sensative drivers to others, would it not? Isn't the idea of injecting an active crossover into the mix sort of the next step of this (taboo by some) evolution?

For the record, no disrespect meant to crossover buyers, designers, tweakers, sellers or anyone else. No active crossovers were hurt nor disassembled during the process of me writing this post.

[]

[Rudy81]

Al, no conclusions here, just exciting to see a light at the end of the tunnel. My knowledge of drivers and crossovers has been increasing so I am taking great interest in these discussions. I can readily see the effects of the various driver efficiencies reflected in my tri-amp setup. My woofer amp is about halfway in sensitivity, my mid amp is almost at minimum sensitivity setting and the tweeter amp is wide open. Granted, not a scientific test, but it just showed me how difficult it must be for a passive designer to get the correct balance between the three drivers.

Now that I have an active setup, I look forward to having more options in my setup. Not that it is bad now, but part of the enjoyment is trying out new drivers and horn.

Thanks for your work on this.

[Al Klappenberger]

Guys,

Yes, the problem of cutting the sensitivity back is getting bigger as we move to more sensitive drivers. There is only a couple ways around it. One is to add a restive attenuator to the transformer. That one stinks. You might as well just use the resistors and forget the transformer if you are going to do that! Another way is by cascading two transformers. That stinks too! The best way is to have another transformer designed that has its taps set at lower levels. That's the best option technically, but I don't think that's going to happen!

An 8 Ohm driver needs a tap 3 dB lower than if it had a 16 Ohm voice coil for the same output level. That says to pick a 16 Ohm driver over an 8 Ohm if you can. That's not much of a help though. I am just hoping that the B&C driver will have a bit lower sensitivity than the others that have been used and give us a little more adjustment room. The spec sheets say that it does. That seems to depend on the horn. We will know when I do the tests.

As to comparing my new horn with the V-Trac horn, I am simply not going to get into that on the forum. I am obviously way to biased about that subject! I want only to stick to the technical aspects of hypothetical horns and these three drivers. I will be using my own horns to do the tests. They are all very similar and will provide a constant test-bed to compare the various drivers.

Al K.

[Guest David H]

Al, when switching to an 8 ohm driver, you are going to get some electrical gain correct? If so, I see another possible issue, if you go lower on the taps, are you not also going to need to adjust the value of the swamping reisistor?

Lets say you have to hypothetically attenuate the B&C -3db, so the tap choice is 0-4, the driver will be showing 16 ohms through the transformer, then add that to the parallel 10ohm swamping resistor, I suspect the network will see 6 ohms.

Even being first order in the Univeral network, I suspect this may be audible. Mathmatically stepping up to a 16ohm swamper might stabilize the network for an 8 ohm driver.

Just a thought, Dave

[PrestonTom]

Probably a dumb question but wouldn't the problem of attenuating that driver be mitigated if someone was using an active crossover?

........

For the record, no disrespect meant to crossover buyers, designers, tweakers, sellers or anyone else. No active crossovers were hurt nor disassembled during the process of me writing this post.

No, it is not a dumb question!

Point #1: An active crossover allows a great deal of flexibility (choice of crossover point, steepness of filter, adjustment of relative gain, a bit of equalization etc). It also allows time-alignment of drivers, which is not typically practical for horn set ups (using passive crossover components).

Point #2: We should not need to apologize for pointing this out since this Forum is not "Garage Sale" but rather "Updates and Mods". I think Active setups should always be considered when folks are updating and modifying.

[Al Klappenberger]

Dave,

Al, when switching to an 8 ohm driver, you are going to get some
electrical gain correct? If so, I see another possible issue, if you
go lower on the taps, are you not also going to need to adjust the
value of the swamping resistor"

Actually no. The only ting that will give you gain is a amplifier! Gain without a power supply is the long sought after perpetual motion machine! Impedance is the ratio of voltage to current. Power is voltage TIMES current. The product stays the same but the ratio changes between 8 and 16 Ohms. 16 Ohms is higher Voltage. 8 ohms is higher current. The power stays the same. The hard part to realize is that the impedance reflected back to the primary side of the transformer stays the same when you move the tap down 3 dB for the 8 ohm driver. So the swamping resistor does not need to change.

" Lets say you have to hypothetically attenuate the B&C -3db, so
the tap choice is 0-4, the driver will be showing 16 ohms through the
transformer, then add that to the parallel 10ohm swamping resistor, I
suspect the network will see 6 ohms."

Again no. The reason is your assumption of 0-4. When you move down to 0-1 on the taps, like we have to do with these super sensitive drivers. the impedance at 5-0 goes up, but it can go no higher than the 10 Ohm resistor! Parallel connect 10 Ohms with 1 meg Ohm and what do you have? 10 Ohms! That is only a 1.25:1 mismatch to the filter connected to it. Anything below 1.5:1 is considered ok. The is the beauty of the swamping resistor idea! If I know a high efficiency driver is going to be used with a network, I will guild the Lilly by changing the resistor to 9.1 Ohms, but it isn't a big deal at all. That makes the error 1.14:1 maximum. The danger is if a low efficiency driver gets connected and the taps move to 0-4, then the load to the filter may get too low going out of the 1.5:1 limit. I suppose that could happen.

Coytee,

YES! An active crossover and bi-amping will totally solve the problem. The faster slopes of 24 dB / octave that some of these provide solves other problems as well. The attenuation problem is simply one dumped on us poor passive crossover users!

Al K.

[Guest David H]

The hard part to realize is that the impedance reflected back to the primary side of the transformer stays the same when you move the tap down 3 dB for the 8 ohm driver. So the swamping resistor does not need to change.

You are correct, this is hard to understand, because when I measure the impedence through the swamped autoformer I see a change in impedence. I am going to have to take your word for it.

Thanks, Dave

[Al Klappenberger]

Dave,

The change is probably because you moved the tap using the SAME LOAD. Try connecting an 8 ohm resistor to 0-3, then a 16 Ohm resistor to 0-4. I betcha the Zo you see at 0-5 will be nearly the same. The next thing to try is to apply a signal to 0-5 and measure the voltage across each resistor. Using ohms law (power = (E X E) / R) . The power will be nearly the same. You can do all this with a pencil and paper using tap ratios too.

Al K.

[Guest David H]

I will try that, I was using parallel resistance calculation. ie, 1/ 1/r1 + 1/r2 = R-total, and verified by measuring esr through the swamped autoformer with an 8 and 16 ohm loads. With a pure resistive load the results were constant at 1khz and 10khz.

Thanks again, Dave

[Al Klappenberger]

Dave,

I can't follow what you did, but It's very easy to get confused when you are doing that sort of thing. The Parallel resistor formula only comes into the picture when you are figuring the swamping resistor. That's step two. Understanding the impedance reflected back through a transformer with different loads and the power it represents is a different idea. Forget about the swamping resistor while your figuring that out.

Al K.

[Al Klappenberger]

Before we go to far off the subject, there a couple of points I would like to explain about crossover network order and horn size. Talking to DeanG on the phone earlier got me to thinking that I should explain it.

I made the statement a while ago that a smaller horn is best because the higher you go above the low cutoff (Fc) the higher the distortion becomes. This says the higher you make the Fc, or the smaller the horn is, the lower the distortion will be at some given frequency well above the Fc. The problem becomes protecting the driver from energy below the crossover as the horn is made smaller. I think it was DJK who said the rule of thumb was that the Fc should be 1 octave below the crossover. That would be 200 Hz for a 400 Hz crossover. That’s a pretty big horn. I assume this is to allow the horn to provide loading to the driver below the cutoff and above where the crossover network starts to significantly reduce the lows going to the driver. Any deep lows getting to the mid-range drive would just sound like a buzz. The slope of the filter in the crossover determines how much attenuation you get below the crossover. I am going to make a few assumptions about this because I really don’t know what amount of power it would take to damage a diaphragm or to cause distortion. I am going to assume what I figure is worst case. Maybe 1 Watt into the driver at 200 Hz with 100W input. The woofer should be doing all the work down there. That means the highpass filter in the network should be providing 20 dB of attenuation at 200 Hz.

You usually see filter plots for 1,2,3 and 4th order with the 3 dB points all lined up and the slopes getting tighter and tighter as the filter gets more elements. I made up a set of plots with the 20 dB points lined up instead showing where the 3 dB points (the crossover frequency) would need to be to hit 20 dB attenuation at 200 Hz. There’s a little ‘+’ mark at the 20 dB point and another mark at the 400 Hz 3 dB point. That’s where we want to crossover the Khorn. The 1st order filter (Like in my Universal network) has to crossover so high I couldn’t plot it! N=2 needs to cross over at 640 Hz. N=3 at 430 Hz. The 24 dB / octave or N=4 can cross over at 355 Hz. This is what I meant when I said the active crossover also solves other problems besides attenuation. Anyhow, I think you can see why driver makers want 12 or 18 dB / octave filters! That won’t give you 20 dB at 200 Hz but it will keep the driver from joining its ancestors!

I also plotted a 3rd order elliptic function filter, which has the same number of parts as the N=4 filter but lets you cross over at 320 Hz. If you use that sort of filter designed for 400 Hz instead, your 20 dB point will move well above 200 Hz letting you use the smaller horn safely. I also plotted the ES400 network filter, It hits 20 dB at 350 Hz. Well above the 310 Hz Fc of both my Trachorns (the vertical green line). Notice how little extra attenuation even a 4th order filter gives you between 200 and 310 Hz over the 2nd order. This says the big horn isn’t the way to safeguard the driver. A more aggressive filter is.

Now will come the comment again about my 120 dB / octave networks being overkill. For driver protection, it is. There’s more to it than just protecting the driver and reducing distortion. There is the matter of interference between drivers that are not time-aligned. If only one driver is making a given sound, there is nothing to align! If you assume you will get interference between to sounds if they are within about 10 dB of each other, then the frequency range where both drivers are less then 10 dB down will be what I call the “interference window”. I am not sure 10 dB is the right number. It’s just an assumption. With the ES400, it’s only 37 Hz wide at 400 Hz. It’s about 200 Hz wide with a 4th order filter. With a 6 db / octave, 1st order filter, it is as wide as a barn door! All this is why I don’t like 1st order filters!


Al K.

[Guest David H]

Very nice, and I could be wrong, but I thought the rule of thumb for horns is 1/2 octave above the fc for the low freq cut and 4 or 4 1/2 octaves above for the high frequency.

Dave

[Al Klappenberger]

Very nice, and I could be wrong, but I thought the rule of thumb for horns is 1/2 octave above the fc for the low freq cut and 4 or 4 1/2 octaves above for the high frequency.

Dave

You might be right. I'm working from memory and my memory stinks big time! If that's really the case, a 200 Hz horn truly is to blasted big for a 400 Hz crossover! For the purpose of illustrating the point about the choice of crossover slopes, the full octave assumption makes the point even if it is wrong!The assumption of needing 20 dB at 200 Hz is probably wrong too considering not to may of us will be pumping 100W into their speakers at 200 Hz. If they were, we wouldn't be using first order crossover networks very long! The K55 is tough but not THAT tough.

Al K.

[hklinker]

Well guys, my brains have officially exploded!

Time to grab the gin and vermouth.

[Al Klappenberger]

I'm not just trying to show off. This stuff is actually fun for me. I did it this morning, just to kill some time, as if I didn't have more important things to do! Here's an actual buildable elliptic function crossover network at 400 Hz that gives about 16 dB protection to the driver at 310 Hz. The 1 mHy inductor at the output of the low filter represents the K33 voice coil inductance. It has 1 more inductor than the classic N=4 24 dB / octave crossover. That's the only down side to it. It gives you about 22 dB at 200 Hz.

Back to reality.. BORING! ...

Al K.

[boom3]

Hey! Hey! Me, in the back of the room!... maybe this is irrelevant, but is the point of this discussion:

a. Replacing the K55 in the current Klipsch top end (still using the K401) ?

or

b. Finding the right combo of non-Klipsch horn and non-K55 driver?

Reason I ask, over the years I have heard some say that the old K-5-J horn was better than the K400/401 largley because it had a bigger mouth. Paul needed to reduce mouth size to fit the K77 on the same baffle without increasing the top hat size. So he reduced the K400 mouth size and fitted a baffle (he called it a flange) to compensate. However, not every bought that concept, as we've seen from the many replacement mid horns folks have constructed. Sooo...are we discussing (a) or ( here?

lapsing back into silence....

[Al Klappenberger]

Boom3,

I think it has to be "a" because the K401, or my horns for that matter, can't use a 2 inch driver and it seems there is just no driver that can directly replace on the K55 in a 1-inch format and be a real improvement. I think the main reason for that is simply that the K55 just isn't that bad! A good 1-inch format driver can be made to survive with a sharp enough skirt on the network filter, but what driver? I know of none that is designed to operate down to 400 Hz. How will a driver made to reproduce 800 Hz sound at 400 Hz even if it survives? The JBL 2426 specs say it can be used to 500 Hz, but that's still higher than 400 Hz. I think it's a small point, but one to consider since everybody is only considering the big BMS driver as if it were the only 2-inch driver available. It's not! There is a HUGE list of possibilities once you swallow the idea of moving to a 2-inch horn.

About the K-5-J, I don't know, but PWK quite making it for some reason. Like the K400, it's an expnential horn. The modern trend is Tractrix horns.

Al K.

[Pete H]

For the record, no disrespect meant to crossover buyers, designers, tweakers, sellers or anyone else. No active crossovers were hurt nor disassembled during the process of me writing this post.

Always the diplomat.[]

[deafbykhorns]

Al, what about developing custom voice coils made for the atlas, JBL or other motor assemblies that will meet the 400hz, impedance and sensitivity requirements? I mentioned a website in another thread that does this custom work.They also have some VC's that will fit the vintage JBL currently in production. I believe you initiated the custom autoformers used in most of the aftermarket xovers. So why not come up with some voice coils to fit the specifications? Or is this a job for Bob Crites given your work load?