Since I wanted to run my small 2 way MEH (https://www.diyaudio.com/community/threads/ikea-kallax-meh-unity-synergy-speaker.382400/) with a single 300B amplifier, I built a simple first order crossover. The sensitivity of the planar tweeter is almost matching the sensitivity of the LF part and the overall FR is equalized with a DSP (basically a low Q cut of a few dBs to make the midrange flatter). Sounds fine and measures reasonably well.
I remember reading that for a MEH it is beneficial for the tweeter to have an inductor in parallel so that it is effectively shorted at low frequencies - to protect the tweeter from the pressure caused by the LF part. For the powers involved (300B SET), it does not probably matter much - but would there be any benefit in using a series crossover, where the inductor is in parallel with the tweeter?
And if I extended this to my larger 3 way builds, I could save 2 channels of amplification by using a first order crossover between HF and MF. I can see some potential issues with the upper end of midrange above the cancelation notch.
Would it be worth using higher order passive filters (parallel) with this hybrid approach?
I remember reading that for a MEH it is beneficial for the tweeter to have an inductor in parallel so that it is effectively shorted at low frequencies - to protect the tweeter from the pressure caused by the LF part. For the powers involved (300B SET), it does not probably matter much - but would there be any benefit in using a series crossover, where the inductor is in parallel with the tweeter?
And if I extended this to my larger 3 way builds, I could save 2 channels of amplification by using a first order crossover between HF and MF. I can see some potential issues with the upper end of midrange above the cancelation notch.
Would it be worth using higher order passive filters (parallel) with this hybrid approach?
To report back, with two shelf filters in the DSP, I get a pretty flat response. I will order some components and try the series crossover, there is a negligible difference when changing from series to parallel with the same component values. And the next step would be FIR filter maybe.
Yes, XSim s what I used - I "tuned" the values not to get any impedance dips and at the same time to get a good summation. I ended up slightly off from a "textbook" crossover. By comparing the serial and parallel crossover, only minor deviations were seen. I will try to post some pictures later.
What I meant was you can see the impedance looking both ways.. from the amp or back from the speaker. The latter would show you how well the driver is shorted by each type.