|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
70.209.67.93
I have been searching through my books and all over the net and can not fine the answers to these questions.
1. If you have a 1st order 6db network on tweeter and a 2nd 12db on the woofer does the tweeter see the cap on the woofer in parallel with its series cap? The reason I ask this is the only way I can get the numbers right on the diagrams is if the two caps are added together.
2. If you are running a 2nd order L/R on the woofer/ midrange what was the purpose of the resistor in series with the cap to ground? To explain better signal 1st thru inductor series to cap/resistor shunt to ground. There is no resistor in normal 2nd order crossover just a cap to ground. He also used a resistor on a inductor to ground on the tweeter high pass.
I have not found a any explanations on how to do the math when you use different slopes on high and low crossover?
I could sure use some help.
Thanks Tom
Follow Ups:
1. No, the two functions are separate and do not interact - assuming a normal circuit and a high-damping-factor power amp.
2. It sounds to me like you are describing a 6dB/octave crossover with Zobel networks (the series RC across the woofer and the series RL across the tweeter).
The function of a Zobel is to maintain a flat resistive impedance of the driver around the crossover frequency range, since drivers have significant deviations from flat resistance. It is not easy to design a good Zobel, because the driver-specific information you need is neither readily available nor easily measured.
Hi Paul
It is always great when you get involved. Because I know I will get someone how can think, hear, and tries things out.
The crossover I was referring to is a 2.5 way L/R 2nd order on woofer/midrange and a 4th order on tweeter. So he wanted the 6.5 scanspeaks to be somewhat separated. The bottom crossed at around 100hz and middle one 1300 hz.
So on the woofers we start with a 1.5 mh inductor in series then 22.5 cap in series with a 6 ohm resistor to ground to 1st woofer to ground then to pad resistor ln series to a 5 mh inductor to lower woofer.
So he called it a L/R 2 order which would seem right if there Was not a resistor under the cap to ground. So can you use a zobel network to create a second order or is he doing something else with that resistor?
I know I am asking a lot but I am confused with formulas and calculators not being what the people are saying the crossover networks are.
Here is one that came with my stock Newform 645s. It has a 10.3 if cap in series with the ribbon. It also has 2.0 mh inductor in series with a 2 ohm pad resistor to the parallel woofers. Also there is a 22.5 uf cap to ground.
So it seems to me it is a 2nd order on the woofers and a 1st order on ribbon. So where did the values come from? If you use a L/R 2 way 2nd order 1300hz will get you close to inductor value. 1600 hz will get you close to 10.3 uf cap value for ribbon. But the woofer cap to ground said it wants 6-8 if and the stock crossover is 22.5. I realize the calculators assume it is a second order on top and bottom and this is a 1st top and 2nd bottom but you said they do not interact.
My only guess at this point is they crossed the ribbon at 1600 hz the woofers at 1300 hz and maybe made the slope steeper by making the woofer cap larger.
I am all ears no body talks about this stuff including The loudspeaker design cook book.
Thanks Tom
I am confused about how many different crossovers you are talking about, and about why you use the labels "2nd order L-R" etc - i.e. are these labels your deduction or are they provided by the designer?
The 22.5uF in series with 6 ohms seems a reasonable design for a Zobel on a typical 6.5" woofer of nominal 8 ohms impedance, to correct for the voice coil inductance. Without more detailed data on the specific woofer in question, nobody can go further than that. (ScanSpeak makes many different 6.5" drivers, and there seem to be several versions of the Newform 645 speaker system.
Beyond that, I'll just note that the drivers themselves are rarely flat, so the acoustic crossover is a combination of the electrical crossover and the driver's response. This, combined with the non-flat impedance of most drivers, explains why theoretical calculations are almost always very different from real-world crossovers.
As an example, I have made a very good 4th order acoustic L-R highpass at 3200Hz, using a single capacitor with a calculated crossover frequency of 16kHz. (This was for a Selenium slot tweeter.)
Hey Paul
Sorry for the confusion. The example I gave with the 22.5 uf cap and 6 ohm resistor to ground was designed by a guy using soundeasy v5.0 10 years ago. He called it 2.5 way cross over. He said it was a 2nd order on low pass and 4th order on the high pass. The only thing that did not look like the normal text book was the 6 ohm resistor under the cap to ground. I follow you on the zobel network so could it be a 1st order with a zobel network and he did not realize it?
On the second cross over it was the stock one that came with the speakers. I just looked at the schematic and use crossover formulas. The only thing here was the numbers came up with 10.3 cap 1600 hz for high pass ribbon and 2.0 mh for 1300 hz high pass woofer. And the cap to ground on the woofer was 22uf the books say it should about 7uf.
So with the stock crossover the big number difference is the cap to ground on the woofers in parallel.(18w/8545-00) scanspeaks. Not sure what changes when you increase the size of this cap from 7uf to 22uf.
I am still confused let's say I wanted a 1st order on tweeter and a 2nd on woofer let's say at 1500hz. All the books and calculators are done with the same slope/order for woofer and tweeter. So if you use 1st order on tweeter it will be 1st order on woofer. So could you say use the values from calculator for a 2 way 2nd order crossover then just remove the grounded inductor on the tweeter.
I follow in the real world there are many variables which create less than a flat response curve. I am trying to understand how calculate a starting point for a 1500 hz crossover 1st order on tweeter and 2nd order on woofer.
Thanks Tom
OK, one more try :^)
For a first order ribbon tweeter crossover you start from the actual impedance of the tweeter in the vicinity of the crossover, assuming this to be fairly flat and free of reactive impedances. Then, given that resistance and the target crossover frequency, you can calculate the capacitor. In fact, given any two of those numbers, the third can be calculated. For example, if you have 1600Hz with 10.3uF, then the resistance you assumed must be 9.66 ohms. That seems an odd value to have chosen, but I have no data on the ribbon impedance.
The woofers are more tricky. For simple home-based calculations it is usual to first put a Zobel series RC network across the woofer to make it a constant resistive impedance, then put the calculated network in front of the [driver plus Zobel]. For commercial manufacture you could then use a computer optimization program to adjust the values for a simplified network, for example replacing the crossover cap plus Zobel with a single cap. It's not exact, but in combination with the driver's response you might easily get a good fit with the desired acoustic crossover function.
Finally, you said "I am still confused let's say I wanted a 1st order on tweeter and a 2nd on woofer let's say at 1500hz. All the books and calculators are done with the same slope/order for woofer and tweeter. So if you use 1st order on tweeter it will be 1st order on woofer. So could you say use the values from calculator for a 2 way 2nd order crossover then just remove the grounded inductor on the tweeter." No, the capacitor for a second order filter will depend on the Q of the filter, and will not in general be the same as the capacitor for a first order high pass filter. For second order filters, the Q determines the type of response - Butterworth, Linkwitz/Riley, Bessel, etc. You can instead use the low-pass portion of a second order design on the woofer, and the high-pass portion of a first order design on the tweeter.
Hi Paul, Can you tell me how to determine impedance for
a particular frequency when impedance vs frequency graphs
are not available ? I have EV SP15B's and find no such
info.Thanks for any help.
Clayton
Clayton
You can download Room Eq Wizard (REW) for free from the Home Theater Shack site (do a search), and it's available in both Mac and PC versions. With this, and a simple home made test box, you can generate your own impedance graphs. There's also a good instructional you-tube video on using REW. I have'nt tried an imp. graph with REW yet, but it has to be a lot less tedious than the old school way of using a test box, an audio oscillator, and a multimeter to make a paper graph.
And share the results here with the Electro-Voice Brotherhood!
One of the other Pauls
Thanks Paul, I will look into this.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: