I've played a bit with Xsim just with dummy drivers, as I don't have concrete measurements. I then made diagrams of different 2-way speakers to see what the components do and the different correction notch that people use. I am aware that the curve I get in Xsim is not the curve the speaker gets as I do not take into account the driver's characteristics. But playing with it has given me a lot of insight I didn't have before.
Often the tweeter has a higher sensitivity than the bass and needs attenuation. I see that some have the resistor before the XO itself and others have it after. But will it depend on the individual situation whether one or the other is an advantage. A filter is shown here. I've put the resistance both before and after the 3rd order filter, to see what it does if it's on "Value" or "short", that's a smart function in Xsim, so I can see the difference in the curves in Xsim.
But is it for impedance reasons that you do one versus the other or why?
Often the tweeter has a higher sensitivity than the bass and needs attenuation. I see that some have the resistor before the XO itself and others have it after. But will it depend on the individual situation whether one or the other is an advantage. A filter is shown here. I've put the resistance both before and after the 3rd order filter, to see what it does if it's on "Value" or "short", that's a smart function in Xsim, so I can see the difference in the curves in Xsim.
But is it for impedance reasons that you do one versus the other or why?
You should be able to get the same result either way. Putting the resistor near the driver is often done as it reduces the variations in the impedance vs frequency, perhaps making the design simpler.
On the other hand if you plan to experiment with the resistor value you might find that putting it before has a reduced effect on the filter varying while you do.
On the other hand if you plan to experiment with the resistor value you might find that putting it before has a reduced effect on the filter varying while you do.
Agreed, the resistor out front will not tilt the response down, whereas after it can. The impedance phase can also be minimal with the series resistor before the network, if that is something the designer wants to have less variation.
Since you are experimenting and learning what resistors do, do not place the shunt resistor out in front as well. This has to be after the xover, or there will be a resistive load across the amp running the full audio bandwidth. This means the resistors would get hot and possibly start a fire.
Since you are experimenting and learning what resistors do, do not place the shunt resistor out in front as well. This has to be after the xover, or there will be a resistive load across the amp running the full audio bandwidth. This means the resistors would get hot and possibly start a fire.
Every time I aks and get some answers, I just have to ask more 🙂 .
Shunt resistor what is it. The one I called R2 on my diagram is that not exactly a shunt? Ore is it af parallel resistor like en L-pad damping?
Shunt resistor what is it. The one I called R2 on my diagram is that not exactly a shunt? Ore is it af parallel resistor like en L-pad damping?
Shunt resistor is connected to ground, placed after the series resistor. A classic L-pad that will maintain a certain impedance.
Ive often seen a resistor installed after the coil to ground. Can anyone comment on the purpose of this?
Ive often seen a resistor installed after the coil to ground. Can anyone comment on the purpose of this?
Last edited:
It's a parallel resistor, meant for across the tweeter terminals. These must go after the xover next to the driver.
The resistor in series with the tweeter shunting coil is a function of the resistance in that leg. Usually, this resistance is shown as the DCR of the coil plus any added resistance. This can be varied to align the phase with the driver adjacent, to change the damping at the knee of the rolloff of driver+xover, or to change the gauge of the coil.
Just so I understand it correctly. L-pad before XO is a no go. Because the parallel resistor is a shunt. L-pad must always be after XO is it understood correctly. I don't think I've seen it before XO. It is only in cases where there is just a resistor in series with the tweeter that I have seen both before and after the XO just as I have shown. There must also be something to do with how much Watt the resistor can handle. I haven't played with that in Xsim yet though. But think that if it is first, it sees many more watts than if it is after XO.
Yes, I can see that if the resistance is before XO, in this case anyway, then there is a slope upwards from XO and up in frequency. When it is after XO, the frequency curve on the treble is more horizontal.
Since sound signal is AC and not DC, current flows both ways. Being that, the highpass will limit the bandwidth seen by the series resistors both fore and aft, and therefore power handling is not really an issue.
Yes, full L-PAD before is a no-go, but after is fine. Series can be before or after.
And to your more recent query- no, the resistor before will not exhibit the tilt and be flat if the driver was originally flat. If you place the resistor afterwards, the rising response tweeter will tilt down and become flatter. It's a matter of incorrect perception on your part here.
Yes, full L-PAD before is a no-go, but after is fine. Series can be before or after.
And to your more recent query- no, the resistor before will not exhibit the tilt and be flat if the driver was originally flat. If you place the resistor afterwards, the rising response tweeter will tilt down and become flatter. It's a matter of incorrect perception on your part here.
At first I didn't quite understand what you wrote Wolf-teeth, I also had to have it translated into Danish 🙂 , but now I've just tried Xsim both without resistance at all, and with resistance before and after. What tricked me into thinking the curve was flat if the driver was flat was that without resistors the curve is not flat at all, it rises sharply right from XO and up in frequency. Both resistance before and resistance after "knocks down" the slope. And having the resistance after the XO turns the slope down a lot and when the resistance is before the XO, the slope is turned down just a little. Is that corect?
A resistor in series with the inductor of a 2nd order high pass filter can vary the slope over what appears to be an extended band, and if you position that carefully you can effectively have a more shallow slope.
It all depends, you'll have to measure first and then model in e.g. Xsim. You might find changes in the phase overlap or the individual delay of the drivers. As @AllenB mentioned the slope can/will get affected which may be a good or a bad thing.
Measurements will show you the Z offset which will have a big impact on the crossover design.
IMHO: Without measuring each and both drivers with something like HolmImpulse and modelling in Xsim you have not got a clue what's happening.
Measurements will show you the Z offset which will have a big impact on the crossover design.
IMHO: Without measuring each and both drivers with something like HolmImpulse and modelling in Xsim you have not got a clue what's happening.
