Low impedance -- ideas to fix?

[arc2v]

Let me see if I understand this correctly:

If the phase angle is zero, then the load is the impedance as a resistive load.

If the phase angle is low at the lowest point, then the amp has to struggle the most. If the lowest point comes at somewhere in the middle (like 30 degrees), that's better.

Is negative angle capacitive and positive angle inductive? Or vice versa?

It's been a long time since I studied circuits in depth, and even then it was more current/voltage theory, not filters and amplifiers.

thanks.

 

[AndrewT]

Hi,
if the phase angle of the load is zero degrees (=resistive) then the current and voltage in the output devices is in phase and the stress in the devices is least. Due to this low stress the devices (in a good amplifier) can usually handle a resistive load of half it's rated impedance.
As the phase angle of the load increases the load and current in the output become more out of phase and the stress in the output devices increases enormously. In addition an inductive load increases the phase margin of the output, but a capacitive load decreases the phase margin. If the reduced phase margin allows the amplifier to become slightly unstable then the voltage oscillations add to any phase induced stress and the output devices become intolerably hot.

If your amp is safe with 4ohm 60degree phase angle then great. But if your loading has anything more severe than this (2r0 zero phase angle is less stressful) then you must check the amp out thoroughly at the problem frequencies. These problem frequencies may not coincide with the lowest impedances in your plot of impedance vs F
The problem areas to look for are high phase angle AND lowish impedance.

 

[arc2v]

Thanks, I will definitely keep all that in mind when I redesign the crossover.

As of now, I have had good luck simulating 2nd order for the high crossover point (i.e. the impedance does not drop as much).

I may also try a different tweeter, as the combo of notch filter and other components make this the hardest one to get right.

Thanks again.

 

[arc2v]

Well, success of sorts. I don't have speaker measurements yet, because it got late, however, I think I came up with a much better design.

First, I got much better measurements after including the amp and mic calibrations in the driver measurements. This led to a much easier and more robust crossover. I can make the speaker brighter (tweeter), forward (midrange), or boomy (woofs) with much less adjustment.

Second, I changed the crossover frequencies to 600 Hz, and 4 kHz (2nd L-R and 4th L-R). With the better measurements, these points just seemed better and gave smoother responses (I wasn't fighting the dipole baffle effects as much, but rather rolling off with them).

Also, my two lowest impedance drops are 3 Ohm at zero phase. This is, of course, simulated, but I built the crossover and music sounded pretty good. I heard a hint of hollowness, usually from the tweeter being too low. I will measure tomorrow after work and readjust.

My lowest phase angle is 50 degrees, at 8 Ohms, at 375 Hz.

So my impedance problems are fixed. So far it sounds good.

I may just be a few tweaks away from having this thing done.

Thanks again for all the help. I was good to know about the capacitance and impedance phase angle. When I saw big dips like that with low phase angle, I knew to start with the (oversized) capacitor, lower its value and then tweak the inductor value until I got it closer to the slope. That worked across the board several times to keep the impedance well behaved.

I'll post the results as soon as I get them.