Crossover Theory Qs

[JohnA]

I recently designed a new crossover for a Peavey FH-1 and MFX-1 (treble

horn in box) for my brother. By all accounts the design works

very well.

The interesting thing about the original design is Peavey reversed the

locations of the caps and inductors in the crossover they sold with the

FH-1/MXF-1. IOW, the low pass had a cap in series with the voice

coil and an inductor in parallel for a 2nd order filter. I'm sure

this reduced the inductance seen by the amp, but,

a - How do I calculate the reduction in inductance and the effect on the filter's response, and

b - How do I calculate the size of the components, if I try to reverse the locations like they (Jon Risch, really) did?

[DrWho]

When wiring up a crossover that way, you don't really get a perfect

slope...ever taken an analog signal processing class? If you knew the

impedance and phase charts for the driver you could come up with some

pretty close equations to model them and then just plug and chug.

Or you could just go with a classic approach and not worry about the wierd effects []

[Al Klappenberger]

John,

I think what you are derscribing is a "sereis" configuration as compared to a "parallel" configuration. If true, the filter theory involved is called the "dual" of each part. If you show me a schematic of what you want I could find what you need. Basically, the dual replaces each part with one of the opposite type in the opposite connection. For example, each parallel inductor becomes a series capacitor.

AL K.

[hwatkins]

Al covered what I was thinking (actually he put it into much fewer and concise terms).

I have recently had to perform something very similar to get my High Definition reception from cable to consistently work well on both tuners in my DVR (MOXI). I had little more than migraines (the migraines grew to epic proportions when I actually had a cable tech on site) until I decided to treat the digital signal transport (the cable) like it needed a crossover to work with the tuners- this reality check happened while pondering why the cable card seemed just fine with the same signal. So, making an assumption that there was clearly a difference in the respective drivers (tuners) I began an experiment that, in effect, played with combinations of resistance in 'series' that included a couple of 'parrallel' solutions that were in series. Measurements of SNR and AGC were my guideposts for changes.

Whew - the point. Maybe our friends at Peavy had a similar problem. The design said one thing and the box created a situation that needed a bit of tweaking to fix....It worked for me.

[WMcD]

John,

You should post a schematic. I'd think that Al. K is correct in his interpretation of what you're describing. Al. K is always correct in what he does and says (smile) but interpreting without a schematic may well be beyond his powers (smile again).

Please recall that in the parallel feed every driver is fed directly by the amp. Or actually the input to their crossovers are fed directly. So they see the same voltage from the amp.

The individual crossover filter allows power through (acting like a short circuit, through), and the amp sees the impedance of the driver.. Outside the working freqs the input to the individual crossover filter looks like an open circuit. If it is an effective open circuit, no power goes through outside the operating freqs.

The last point explains how we have three drivers and crossover filters in parallel but the impedance across the full range is about 8 ohms, or whatever the driver presents (save for auto transformer effects).

In the series set up we have, say two or three drivers in series and they all get the same current. The current through a pure series is the same in every component.

Here if we need fiters, just about all we can do is put the filter in parallel with an an individual driver. Here the filter shorts out a given driver, outside the working freqs.

There is a problem with the series set up. It is that the performance of every individual crossover depends upon the impedance of the other drivers and crossovers. I think that is why they never became popular.

Gil

[Deang]

"There is a problem with the series set up. It is that the performance of every individual crossover depends upon the impedance of the other drivers and crossovers."

Unless I'm not understanding you correctly, I believe Master Yoda would say that this is always the case.

[sfogg]

One of the claimed advantages of first order series crossovers is that

the crossover points sort of self adjusts to changing driver

parameters. (from VC heating... or whatever)

Such that is say the woofers impedance changes from heating and the low

pass to the woofer goes up in frequency in a series crossover the

tweeters high pass will also move as part of the whole crossover.

Shawn

[Klewless]

If it is a true series type xover where the inductor shunts lows around the tweeter, then there exists a failure mode to get excited about. Should the inductor open, the tweeter would then face first hand all the bass currents for about a millisecond before it bites the dust.

I once had a book that gave the conversion rules to convert between series and parallel networks. Should actually still have that book someplace around here.