Resistor error in thinking...

[davidB]

Try putting a low bass note through a low-to-mid impedance shunt before the network.

[DeZZar]

I have a sneaking suspicion that the power dissipation plot in Xsim on the fore resistor is exaggerated. I've not sat down and tried to test my theory, but something tells me this is the case. The one after is in series with the tweeter, so they share the load together. But, the fore resistor is still highpassed and should not have a problem in reality. I've never had a fore resistor get warm if using sandcast or ceramic resistors. The one resistor that gets a little warm at any rate is the metal oxide variety, but that's not due to position in the circuit.

So I setup a quick test this morning using the above mentioned crossover schematic as a basis.



With an 8Khz test tone and the volume knob arbitrarily set the input saw 7.43Volts before the first resistor. I didn't have exactly 2.7 so I used the closest I had @ 2.97.

After the resistor the voltage dropped to 4.86v. This is with 1.97amps across the 8ohm resistor pretending to be a tweeter. After a couple of min the resistor did warm up quite a lot (about a hot cup of coffee temp).

With a 500hz tone the voltage only drops by around 0.07v with virtually no amp draw at the "tweeter". So I stand corrected the resistor wont see the "full range" as such positioned here due to the amp draw.

As mentioned earlier, under real world music conditions and reasonable listening levels I've never had resistors get hot and certainly cant see how it would become so hot that your speakers would...burst into flames!

[Billet]

I also spent about an hour working on a post, only to delete it. This is a complex topic for sure!

[Wolf]

I have a sneaking suspicion that the power dissipation plot in Xsim on the fore resistor is exaggerated. I've not sat down and tried to test my theory, but something tells me this is the case. The one after is in series with the tweeter, so they share the load together. But, the fore resistor is still highpassed and should not have a problem in reality. I've never had a fore resistor get warm if using sandcast or ceramic resistors. The one resistor that gets a little warm at any rate is the metal oxide variety, but that's not due to position in the circuit.

[rpb]

How loud is the tweeter playing at those power levels?

[rpb]

I have some old Radio Shack L-pads if anyone wants them.

[rpb]

Bill, I'm okay with a resistor before the xover. In fact, I normally use a split Lpad with the series before, and the shunt after. I'm trying to debunk the myth that seems to be popular thought.

There is no bad reason to use a resistor in series before, and it's actually an improvement in impedance phase as a result. This also does not tilt the response like the one after can. The 2 best ways to attenuate and not change the spectral shift are as my split example.

Later,
Wolf

I spent an hour trying to figure out how to word my post, because I was unsure of what you were saying. This clears it up. We are in complete agreement as far as I can tell.

[Wolf]

You are using my preferred method of Lpad, and there should be no problem with it.
Wolf

[Wolf]

Bill, I'm okay with a resistor before the xover. In fact, I normally use a split Lpad with the series before, and the shunt after. I'm trying to debunk the myth that seems to be popular thought.

There is no bad reason to use a resistor in series before, and it's actually an improvement in impedance phase as a result. This also does not tilt the response like the one after can. The 2 best ways to attenuate and not change the spectral shift are as my split example.

Later,
Wolf

[tktran]

Can you recommend an adjustable L pad product/model?

[DeZZar]

ok....maybe I read the question wrong?

According to simulations of this xover this resistor is at risk of dissipating more than 5W from 20W+ input, 10W from 35W+ input and so on. So depending on the power rating of the resistor - it could be at risk of over extending itself. Sorry if that's not what was being asked?

[Steve Lee]

AEIOU - The suggested technique is still valid even if the quality of your L-Pad (Potentiometer) is not.

[AEIOU]

I've attempted to do that, the problem being is that the usual variable (adjustable) wire wound L-pad is horrible, doesn't maintain a very constant impedance at all. The ones I've measured vary from 7-11 Ohms and this was with an 8 Ohm non-inductive resistor in place of a tweeter, so there was very little reactive component.

[AEIOU]

Yes is the wrong answer.

[billfitzmaurice]

I agree with Wolf. I can't think of any good reason to place a resistor before the high pass filter, and there's a few why it's a bad idea. I also can't think of a good reason to use anything but an LPad after the filter. Since you can't know exactly how things will balance out without audio testing I use an adjustable LPad temporarily wired in place, test the system, adjust the LPad to the degree of desired attenuation, then measure the LPad itself to determine the resistor values to use in a fixed value LPad.