Live Edge Dipoles - #1 at Parts Express 2023 Speaker Design Competition - Updated Design

I guess 680uF is going to do that, especially in combination with only 1.8mH.
well, per the StarWars quote "Its a trap!" there is always going to be a Z null at the series resonance. the drivers are nominally 8 ohms. I'm lowpassing the woofer at around 100 Hz and high passing the coax at around 400 Hz and those are the values I end up. Its very similar to what Perry did.
 
I'm lowpassing the woofer at around 100 Hz and high passing the coax at around 400 Hz and those are the values I end up. Its very similar to what Perry did.
Yes, indeed. I think the values you've ended up with are more challenging than Perry's, though. He has 200uF on the Live Edge dipoles (according to post #1) and 360uF on the Bitches' Brew (see post #11). I think 680uF is a lot of capacitance to shunt a woofer with.
 
Here is the Vituix screenshot
View attachment 1451266
As best I can tell, you have two 8" mids and one is canceling the other to achieve a desired polar pattern. The cancellation + baffling reduce the sensitivity of the pair by something like 6-9dB below baseline, so you designed a series crossover that compensates for that by drawing more current from the amplifier at 160hz.

Typically I would never present an amp with a 1 ohm load. You can always increase the L and decrease the C while keeping the frequency the same and that will lower the output and raise the impedance. I suggest aiming for a minimum impedance of 4 ohms.

Then you compensate by adding more gain over that ~160Hz frequency range in the DSP so you're asking the amp to deliver voltage instead of current.
 
Thanks Perry! I was sorely tempted to keep that 1 ohm load and find an amp sturdy enough to handle it but I think I can raise the minZ to 2 ohms.

I have a coax mid with side slots which do the cancelling to create the cardioid-like directivity. I added the 8RS340 front woofer just below it to carry the load from 100 Hz to roughly 250 Hz to limit excursion of the 5208C mid. The low pass slope of the side subwoofers you saw in the system simulation is carefully tailored in amplitude and phase/delay to provide the cancellation between 100 Hz and 200 Hz and to augment it above 200 Hz. Initially I had a brute force XO with FIR filtering on 4 channels. I eventually learned to do the side subs with IIR plus DSP delay allowing it to fit within the tap limits of MiniDSP flex. Now I'm targeting Hypex FA-253, which requires a discrete XO to eliminate 1 amplifier. Yesterday was my first stab at that.

I apologize for presenting the problem in an unexplained system context. Here it is again in isolation where I took your advice and played with L & C to get a higher Z. That resulted in C decreasing and L increasing and moved the low pass on the 8RS up in frequency. The blue dashed line is the result after FIR correction to target XO slopes. I end up with a minimum Z of 2 ohms with the peak apparent power roughly half the 2 ohm rating of the FA253's amps. I still have to check 5208C excursion and see directivity in the system sim but I think this is the best I can do with a first order series XO. I may be asking too much of its narrow slopes, but as you said they make for a good blending of the outputs of two drivers.

You can find the Power Dissipation pop up under the view button onthe Vituix main screen. I only just found it there myself.
SeriesXOscreenshotV2.jpg
 
  • Like
Reactions: perrymarshall