Need some help with 3-way crossover network...

[Wigwam Jones]

I could use some help understanding my 3-way crossover network. This is mounted on a board in my vintage ElectroVoice EV Nine speakers, and I'm planning to recap. I know I can just replace the caps with like values, but I'd like to understand more about what's going on here. I'd appreciate any help anyone could give me. I apologize for the lack of photos/diagrams.

Parts:
3 inductors. Large, medium, and small. The two largest are solid-core, the smallest is air-core. No values marked on them.

4 capacitors. 1 uf, 4 uf, 12 uf, and 50 uf. All electrolytic. All marked 25 v N.P.

2 resistors. 1 wire-wound (rectangular block) marked 7.5 ohm 10%. The second is a carbon resistor color-coded to 40 ohms 5% (yellow/black/black/gold).

1 potentiometer.

NOTES:

* The 40 ohm resistor is connected across the legs of the pot.
* The 7.5 ohm resistor is connected in series with the 4 uf capacitor.
* There appear to be two ways to get to the negative terminal of the tweeter. Either from positive terminal to a 1 uf cap, or to the potentiometer, then to a 12 uf cap. Both end at the tweeter negative.
* There is a 50 uf capacitor connected between Woofer Negative and Woofer Positive on the crossover board.
* The woofer and midrange negative are tied together with the tweeter positive. It appears that the tweeter is phase-reversed with respect to the woofer and tweeter.

With that in mind, here is what I came up with:

Paths to Tweeter:
1a) Positive terminal -> 1 uf capacitor -> Tweeter Negative
1b) Positive terminal -> potentiometer -> 12 uf capacitor -> Tweeter Negative
2) Tweeter Negative -> air-core inductor (small size) -> Tweeter Positive & Negative terminal

Paths to Midrange:
1) Positive Terminal -> 7.5 Ohm 10% resistor -> 4 uf capacitor -> solid-core inductor (middle size) -> Midrange Positive
2) Midrange Negative -> Negative Terminal

Paths to Woofer:
1) Positive Terminal -> solid-core inductor (large) -> Woofer Negative
2) Woofer Negative -> Negative Terminal

I understand the purpose of this type of crossover; the inductors set the low-frequency crossover point, and the capacitors set the upper-frequency crossover point. What I don't understand are:

50 uf cap between pos & neg terminals for woofer?
40 ohm resistor across legs of potentiometer?
2 ways to get to the tweeter negative terminal - through the 1 uf cap or through the potentiometer and then through a 12 uf cap?
The air-core inductor appears to run directly between the positive and negative terminals of the tweeter and not be in the path at all, really.

Can anyone explain this to a layman in simple terms and perhaps take a WAG at the crossover frequencies we're talking about here? I realize that without the Henry values of the inductors it is a crap-shoot, but what do you suppose the logical choices might be? I don't have any way to measure the inductance, I only have a multimeter.

Thanks!

 

[Wigwam Jones]

*bump*

Anyone?

 

[Wigwam Jones]

I have added this diagram, hope it helps! You can click on it to see it larger if it's too small.


electrovoice_ev_nine_crossover_network by Wigwam Jones, on Flickr

 

[dc270]

The xover frequencies are going to depend on the ohms at the xover point of the respective drivers, it looks like it is a -12db octave type network for the woofer/tweeter, which is a good steep slope. Not knowing the inductor values is a wild only guess. Perhps if you can offer the driver size them we can work backwords from that.
DC

 

[Wigwam Jones]

The xover frequencies are going to depend on the ohms at the xover point of the respective drivers, it looks like it is a -12db octave type network for the woofer/tweeter, which is a good steep slope. Not knowing the inductor values is a wild only guess. Perhps if you can offer the driver size them we can work backwords from that.
DC

Thank you!

The woofer is 10 inches (I said 12 inches previously, mistakenly). It measures 4 ohms across the terminals. The midrange is 5 inches with a sealed back, it is marked '8 ohms' and it measures 8 ohms across the terminals. The tweeter is 4 inches and it is marked '8 ohms' and measures 8 ohms across the terminals. Only the woofer is unmarked. They are all vintage alnico designs. The maker is unknown for the woofer, the mid and tweets are marked 'Made in Denmark' and the speculation is that they are Peerless units. All are paper-cone units.

Does that help?

 

[dc270]

From my experience with similar speakers I have found the following:

The sealed back 5" mid is a good clue- they do not cross over very low, this unit is crossed over with a simple 1st order network even at that. I would guess the mid/ woofer to xover at least 1000hz on the low end- at 1000hz the woofer is possibly at 8 ohms. The upper end I would say is in the 5000~7000hz range. Typical for a tweeter of that era and the effective upper limit of that mid. Hope this can help ya.
DC

 

[paymenow]

Are you sure that the air core inductor is not between the + and - leads? That is what I would expect to see. Also, are the + and - terminals identified as such on the drivers. I would kind of expect the mid to have its connections reversed but I'm not as sure of that as the air core inductor connection.

 

[Wigwam Jones]

From my experience with similar speakers I have found the following:

The sealed back 5" mid is a good clue- they do not cross over very low, this unit is crossed over with a simple 1st order network even at that. I would guess the mid/ woofer to xover at least 1000hz on the low end- at 1000hz the woofer is possibly at 8 ohms. The upper end I would say is in the 5000~7000hz range. Typical for a tweeter of that era and the effective upper limit of that mid. Hope this can help ya.
DC

It does, thank you! FWIW, I had an earlier post on the speaker and drivers themselves:

Woofer:


100_3103 by Wigwam Jones, on Flickr

Midrange:


100_3105 by Wigwam Jones, on Flickr

Tweeter:


100_3106 by Wigwam Jones, on Flickr

Given the information available - assuming you were willing to upgrade this speaker - would you attempt to engineer a newer/better crossover for this speaker, or just replace the caps and call it a day?

The reason I ask is that now that I have a parts list, I am about to place an order for the capacitors. But I also notice I could purchase a new pre-built crossover for not much more (perhaps less, depending on where I shop). Or, I could try to build my own crossover using the same speaker components, with perhaps a newer (hopefully better) design.

As a design exercise, which would you prefer? Thanks again!

 

[Wigwam Jones]

Are you sure that the air core inductor is not between the + and - leads? That is what I would expect to see.

Ah, you are correct! I'm sorry, my drawing is wrong. The air-core inductor goes to the terminals on the crossover board that serve the tweeter wires that are green and black.

Very sorry for having gotten it wrong.

Also, are the + and - terminals identified as such on the drivers. I would kind of expect the mid to have its connections reversed but I'm not as sure of that as the air core inductor connection.

Good point - they are not marked + or - on the speaker terminals. I had to make assumptions that black wires were 'common' or negative. The woofer wires are black/white. Mid is red/black. Tweeter is green/black. So I presumed that black was common. I noted that the terminal where the Tweeter green wire is connected is a three-way connection - it also goes to the woofer black and the midrange black.

Does that make any sense?

 

[dc270]

I admire your interest here. Reviving an older design is an admirable goal indeed. I would personally start with the caps- I have found them typically drifted way out of spec. Later if you wish to make a small investment in a LCR meter, measure your inductors and replace them with lower insertion loss types- however this coul possibly (as will the newer cap ) affect the voicing of this speaker! Start small, make gains you can live with and affoard then move along. I think you can impress yourself!
DC

 

[Wigwam Jones]

I admire your interest here. Reviving an older design is an admirable goal indeed. I would personally start with the caps- I have found them typically drifted way out of spec. Later if you wish to make a small investment in a LCR meter, measure your inductors and replace them with lower insertion loss types- however this coul possibly (as will the newer cap ) affect the voicing of this speaker! Start small, make gains you can live with and affoard then move along. I think you can impress yourself!
DC

Thank you very much!

My wife just returned from the local public library; turns out they had a couple of interesting-looking books there that she checked out for me: "Advanced Speaker Designs for the Hobbyist and Technician," by Ray Alden and "The Audiophile Loudspeaker Anyone Can Build," by Gene Healy. Glancing through the pages, I see they explain a bit on crossover design and how it works. I can see how I'll be spending MY Christmas!

I understand your point about incremental upgrades. I guess it can't hurt to start by re-capping the existing crossovers; I can always re-use the capacitors if I end up doing something else!

Where does one go about getting an LCR meter? Are they expensive?

And one website I found indicated that I need an 'audio tone generator'. They suggest building one for myself - I looked over the details and it appears to be way too much for me. Can one buy such a thing inexpensively?

 

[Wigwam Jones]

Oh, and I would appreciate it if anyone could educate me on a couple nagging questions I have.

What is the purpose of the 50 uf capacitor placed between the woofer black and white terminals? It appears not to be in the circuit other than that, so what does it do?

Why is the potentiometer jumpered with a 40 ohm resistor between the input and output? And why is there a 1 uf capacitor going from one side of the pot to the tweeter black wire and a 12 uf capacitor going from the other side of the pot to the same tweeter black wire? I don't understand this either/or circuit - like the circuit can flow either or both ways???

 

[GordonW]

Wiggy, there's an easier way to get a tone generator. You can download one of many signal generator programs, that just uses your computer's sound card as a signal generator. Here's on example:

http://www.dxzone.com/cgi-bin/dir/jump2.cgi?ID=17724


I haven't used it myself, but it looks to be about as straight-forward as could possibly be...

Re: The capacitor on the woofer... that converts the slope of the crossover, from first to second order. It makes the woofer roll off "sharper" (reduces the amount of high frequencies that will "overlap" with the response of the midrange).

Regards,
Gordon.

 

[Wigwam Jones]

Wiggy, there's an easier way to get a tone generator. You can download one of many signal generator programs, that just uses your computer's sound card as a signal generator. Here's on example:

http://www.dxzone.com/cgi-bin/dir/jump2.cgi?ID=17724


I haven't used it myself, but it looks to be about as straight-forward as could possibly be...

Brilliant! Thanks! That I can definitely do!

Re: The capacitor on the woofer... that converts the slope of the crossover, from first to second order. It makes the woofer roll off "sharper" (reduces the amount of high frequencies that will "overlap" with the response of the midrange).

Oh-ho! I am going to make a leap in logic here - since it involves a higher parts count, I presume that the 2nd-order design is not based on saving money in the engineering design, but is actually an improvement over 1st-order designs? Or is there some deficiency in the woofer that requires the use of a 2nd-order crossover design? Pardon the questions, I'm eager to learn here.

 

[dc270]

Brilliant! Thanks! That I can definitely do!



Oh-ho! I am going to make a leap in logic here - since it involves a higher parts count, I presume that the 2nd-order design is not based on saving money in the engineering design, but is actually an improvement over 1st-order designs? Or is there some deficiency in the woofer that requires the use of a 2nd-order crossover design? Pardon the questions, I'm eager to learn here.

The woofer probably extends a ways out beyond the xover point so a sharper roll off was in order by the engineers. The 1sy order, 2nd order, etc, is selected by the engineer based on driver characteristics and interaction with other drivers. Sometime you may want more interaction- other time not- depends on the engineers, power ratings, etc.....
DC

 

[dc270]

You can find basic LCR meters for $30+ on those auction sites! For the small investment they are well worth it- especially in trouble shooting and design. Gordon was right on about the tone genertator. You can also down load test tones off the net, hook up a good mic and get a good responce program or ocilliscope proram to help you as well.
DC

 

[paymenow]

To expand a bit on Gordon's answer re the function of the 50uF capacitor, a basic property of capcitor is that the present decreasing impedance as frequency increases. Inductors are the opposite. So the 50uF cacitor offers a lower impedance path for high frequencies than the inductor in series with the woofer or the woofer voice coil. Conversely, the inductor in series with the woofer presents a higher impedance to high frequencies thus attenuating them in the woofer circuit. By applying these concepts to the rest of the crossover, you can generally figure out what is happening.

 

[Wigwam Jones]

OK, so I redid my diagrams. This time, to better understand what is happening, I broke down the tweeter, midrange, and woofer sections of the crossover. I also added the connector strips for clarity.

This is the simplest, the woofer:


EV_Nine_Woofer_Crossover_Diagram by Wigwam Jones, on Flickr

As I understand it, the incoming signal on the positive speaker terminal must pass through the inductor; it doesn't have a choice.

Once it passes through the inductor, the signal can go one of two directions; to the woofer and then back to the negative (common) speaker terminal, or it can go through the 50 Ohm capacitor and then to the negative speaker terminal.

If I understand this correctly, signals of a higher frequency than the 50 Ohm capacitor represents are going to return to common. Those signals of a lower frequency than the 50 Ohm capacitor represents are going to go to the woofer and be converted into sound energy.

I am still not sure what the function of the inductor is in this circuit, though. All signals have to pass through it, right? What signals would not?

Moving on the next most complicated crossover, the midrange:


EV_Nine_Midrange_Crossover_Diagram by Wigwam Jones, on Flickr

In the midrange crossover, the incoming positive terminal passes signal through the resistor, the capacitor, and the inductor. Then it goes to the midrange red wire. The return wire from the midrange driver goes straight to the negative speaker terminal. So all signals must pass through this network, they don't have any 'choices' like with the woofer circuit.

Finally, the tweeter, which is the most complicated and hard for me to understand:


EV_Nine_Tweeter_Crossover_Diagram by Wigwam Jones, on Flickr

I won't even try to describe this one, as I simply don't understand what is happening here.

Thank you all for your patience. I'm sorry this is so difficult for me to wrap my head around!

 

[ironhead17]

cross-over

Wiggy - can you correct the drawing and provide a picture of your cross-over? I'm trying to capture some of the same knowledge and this topic is very helpful. Thanks

edit - Nice - i was away and you lapped me. Thanks

 

[GordonW]

The woofer crossover is just a plain jane 12dB/oct. lowpass. Normal stuff.

The midrange crossover is a "first order" bandpass... 6dB highpass (meeting the woofer) and 6dB lowpass (meeting the tweeter).

The tweeter crossover is a variant of a 12dB/oct. highpass... that extra cap and resistor acts as sort of a level control, but instead of just increasing the level as you turn it up, it actually reduces the crossover frequency as well. I wonder how well THAT works, in practice? It may well work, with the shallow slope (only 6dB/oct) of the midrange... it may function to create a "shelving function (where the tweeter is "shelved down" in level- i.e, there's a "step" in the response between the mid and tweeter- both are flat, but at different levels) to allow the tweeter response to be turned up and down, wholesale, in relation to the mid, without resulting in as much of a "discontinuity" in the response at high levels...

Regards,
Gordon.