😉 I'm stubborn and rather determined at times. I don't like giving up. Great offer but I'll persist. I'm using Visaton drivers currently as no need to trace any graphs. It loads 4 curve files as soon as the driver is loaded.
So the select transformer appears to change the optimisers goals. So added one of it's calc'd Z corrections. Optimised and it did improve it. Then added a fairly random RC to the circuit. The effect is shown dotted, the higher F slope. Double peak close to the first optimisation. Solid what the optimiser did when the add on RC was included. All of the previous values where changed. This is with box.
I've just added an L pad to remove 8dB. Result Z is now very close to a straight line at 8ohm. This one is mid range. The woofer has to be 4ohm. I'm wondering if I should switch this one to 4. Or try and bring it down from 8 with the L pad.
So the select transformer appears to change the optimisers goals. So added one of it's calc'd Z corrections. Optimised and it did improve it. Then added a fairly random RC to the circuit. The effect is shown dotted, the higher F slope. Double peak close to the first optimisation. Solid what the optimiser did when the add on RC was included. All of the previous values where changed. This is with box.
I've just added an L pad to remove 8dB. Result Z is now very close to a straight line at 8ohm. This one is mid range. The woofer has to be 4ohm. I'm wondering if I should switch this one to 4. Or try and bring it down from 8 with the L pad.
There is no reason or benefit to making it 4 ohms.. however the L-pad itself does help to bring the impedance closer to flat so make good use of it.
I tried correcting the woofer again. i did look at using an RLC but it wanted 2,5mF so didn't try it. This time I did an optimised it, That gave a sensible value for C. Then added the RC circuit, Result a pretty flat 4ohm line across the whole F range. The other 2 are 8ohm and 10ohm.
Amp sees ~2ohm all connected. I've no idea what crossovers will do to that. So next step to see.
Amp sees ~2ohm all connected. I've no idea what crossovers will do to that. So next step to see.
That comment interested me. Does again now I'm redoing it all.
I hear LR crossovers are fantastic. Bit of a problem. Boxsim can show the theoretical curves for these and Butterworth. If I add an LR the filtering is nothing like it should be. Phase is usually sort of weak S shape, cuttoff slope way out and the calculated cutoff F doesn't hold. Always out. A 2nd order LR doesn't really offer anything over a 2nd order Bessel. In this case load is pretty flat 4ohms dished a bit to 3.8 at cross over F.
Butterworth. Calced cut off is close and gets a lot closer to showing the correct slope. A 3rd order even better. Phase - pretty straight saw tooth.
Linkwitz Q 0.5 and Butterworth both calc'd via Boxsim
😉 It can show Bessel too but doesn't offer a calculator. The posted one was the result of optimisation. Phase optimisation seems to involve changes to cross over F. That seems to be a valid method on multiway speakers.
Sometimes LR is what you need, sometimes Butterworth, sometimes something in between.
The first issue is getting there. Standard component values won't do that. It can help to know what they look like and aim for that. It helps if you have a plot overlay to follow.
The first issue is getting there. Standard component values won't do that. It can help to know what they look like and aim for that. It helps if you have a plot overlay to follow.
