Intact Audio

Hello all,

I would like to try an RL high-pass filter on my compression drivers instead of the series caps (15uf) currently in place. The caps I have tried range from the "bright and chalky" to the "wet-and-slippery" to the "warm-and-plasticky"...and I would like not to have a cap in there at all, as it is my experience that any cap is some combination of said attributes. The cap is followed by an l-pad (crappy 8 ohm rotary Fostex, sounds tinny and anemic though this is supposed to be a temporary arrangement and perhaps an autoformer should be an option as well!).

It appears that whatever series resistance is chosen as part of the RL filter will be added to the series "leg" of the l-pad, influencing the impedance of the network and therefore the crossover point.

How do I take this into account or counteract this?

I am obviously a "newbie" and would very much appreciate someone chiming in with a circuit diagram, preferably drawn in crayon...many thanks!

Hey,

How much attenuation are you getting from your L-pad?

dave

something like this?

assuming the 8dB cut, the source impedance driving the speaker is below. (Green is cap, Blue is the inductive)

Thanks very much, Dave.
I wish I could say with any degree of accuracy or confidence what attenuation I am actually getting with the l-pad in place. I have measured the series R and it is somewhere in the 7-7.5 ohm area, the parallel leg somewhere in the 2.5-3 ohm range (multimeter dances around these numbers).
However, according to the calculator I am using this only results in a 12db or so drop, which can't be right...The l-pad is almost entirely "closed" (say 90% attenuation as far as the rotary angle of the knob) and this is a 118db driver in a horn matched more or less to a 96-97db low-mid driver in a box. It seems that I must be attenuating at least 20db?!
In the circuit you put up, what is the role of that series inductance (L2)? With my limited understanding it looks like a low-pass filter (?) with only a very small resistance value?
When I had only a series resistor in line with the compression drivers, I remember it was in the 150 ohm range to match with the midrange...
So I am in a fix on all counts....

Hey,

The two series inductors are on a common core and tapped. The overall value works against the 8R series resistor to form the RL filter. The 8R series resistor makes sure that the amp sees at least 8R at all frequencies and the tapping of the inductor provides attenuation AND lowers the source impedance driving the speaker.

dave

What would this look like if you needed to do a 3rd order filter?

You would have to sim it but this should work... you may need a shunt resistor across the entire choke to get things stable since the impedance curve can get wonky with the stepup from the tapped choke.

dave

Is there no way to do it purely as an RL circuit?

Not that I can come up with for this situation. The RL filter requires a series resistance and a shunt inductor and the closest I can come is using the output impedance of the amplifier as the source Z and a shunt inductor but that will unduly load down the source at low frequencies. If say 6dB of HF attenuation is also needed then a 8Ω resistor could be placed in series with an 8Ω load and the shunt inductor could be sized for the crossover frequency.

Actually.... since the device being high passed typically also needs attenuation the "Rf" can also be part of the pad and then a shunt inductor added as the filter element. Assuming around 6dB here is a LR compared to a typical capacitor + autoformer. The big difference between the two will be the damping factor of the driver. In the LR case it will be 1 and in the autoformer case it will be much higher.