How 2 Calculate (theoretical) crossover point from a schematic

[rongon]

OK, maybe a dumb question. But I hope somebody can help me out...

I've read some speaker design books, looked at a couple of crossover design spreadsheets, but I haven't found anything that would let me plug in values from a previously designed crossover and figure out what the theoretical crossover point will be. The calculators all seem to start from driver parameters and a desired crossover point, and calculate which component values you should use for this or that type of crossover.

My example is a Klipsch KG4.5 crossover. The high-pass section is a 3rd order, like this:



How do I figure out what the crossover point is for this part of the filter?
___________________________________

Here's the low-pass section of the crossover (2nd order filter):





Thanks in advance...

[DrDyna]

Speaker Workshop can do that I think, but you'd be assuming resistor-flat impedance for the drivers, without seeing what the drivers impedance and frequency response charts look like, there's no way to know for sure.

[rongon]

Thanks for your reply.

you'd be assuming resistor-flat impedance for the drivers, without seeing what the drivers impedance and frequency response charts look like, there's no way to know for sure.

I'm not looking for precision, just a ballpark idea. For instance, it would be useful to me to know if the theoretical (electrical only) crossover point is in the neighborhood of 1000 Hz, instead of 2500 Hz. The 2nd order filter should be straightforward, what is it, 1/2pi(LC)? Or is that for 1st order only...

[DrDyna]

Can't remember the formula off hand, but I usually visit this page if I need a formula.

[Andrew Eckhardt]

I'm sure there was a time when certain people looked down upon others for using a slide rule. There's nothing wrong with using the equations manually, but I'd just download LTspice, plug in the circuit, and run an AC analysis. Then you'll have a powerful tool sitting on your hard drive that can answer a lots of other questions too.

(open in an image viewer actual size)

[rongon]

Thanks Andrew. DL'd LTspice, gave it a whirl, didn't get a good result the first time, but I'll keep at it.

--

[infinia]

Originally Posted by Andrew Eckhardt

I'm sure there was a time when certain people looked down upon others for using a slide rule. There's nothing wrong with using the equations manually, but I'd just download LTspice, plug in the circuit, and run an AC analysis. Then you'll have a powerful tool sitting on your hard drive that can answer a lots of other questions too.

(open in an image viewer actual size)

+1
I do this , but using the drivers Le and Re and taking the output across Re.
Also using inductors Rdc can give very good 1st order approx of the electrical transfer function. Once you get good you can enter driver sensitivity as well. but remember acoustic response is important too esp if cross is near woofers linear or HF capabilities.

[Conrad Hoffman]

There are simple formulas for common crossovers that you can rearrange to get the frequencies. You have to assume a fixed resistance driver for those. For an arbitrary crossover or one that includes driver reactance you're talking circuit analysis. This requires complex values (j, the square root of -1) and you can do it by hand or you can do it by Spice, like LTSpice. IMO, it's not a bad idea to do it by hand once or twice, but you'll quickly start using LTSpice and never look back. If you want to learn how to do it by hand, get a text on basic-to-intermediate electronics like Boylstad's Introductory Circuit Analysis. I like the 5th edition, but it's up to the 11th and I see a lot of complaints. The later edition spends more time with Spice and the earlier edition does it by hand. No idea which is clearer, but definitely go with a cheap used copy. There may also be a download available. IMO, some version or another is an extremely good reference for the audio electronics DIY bookshelf. As Jim Williams would have said, "No home lab library is complete without a copy."

[infinia]

yes but doing "it" by hand can give different results for the breakpoints esp for BSC compensation. For Example in 'The Econowave' thread at AK they were talking about the 2 crossover frequencies? till I showed where it was really crossing over using LTspice.

[Conrad Hoffman]

Then somebody isn't doing "it" right! Are you saying that the response of complex circuits couldn't be calculated correctly until Spice was invented? I'd have some doubts about that