Electrical Schematic Question

[whatatrip]

In the attached schematic, I would like to convert the impedance EQ circuit into two parallel legs circuits. Is the attached schematic equivalent to the following?
Leg1: Ce + Req
Leg2: Cm + Lm + Rm + Req = Cm + Lm + Rm + R(m+eq)

[Pete Schumacher]

Re: Electrical Schematic Question

In the attached schematic, I would like to convert the impedance EQ circuit into two parallel legs circuits. Is the attached schematic equivalent to the following?
Leg1: Ce + Req
Leg2: Cm + Lm + Rm + Req = Cm + Lm + Rm + R(m+eq)

The problem you have in breaking the circuit up is because the action of Ce on the input to the second leg is going to be missing if you move it.

What is it you're trying to accomplish? You don't need to flatten the impedance. The crossover can be made to account for the impedance to produce the proper acoustic profile.

[whatatrip]

Re: Electrical Schematic Question

The problem you have in breaking the circuit up is because the action of Ce on the input to the second leg is going to be missing if you move it.

What is it you're trying to accomplish? You don't need to flatten the impedance. The crossover can be made to account for the impedance to produce the proper acoustic profile.

I'm trying to figure out what Xover Pro is doing and how to translate that to PCD. I can usually take a PCD design and put it into Xover Pro and come up with very similar system results but going the other way is sometimes hard to do because of the way Xover Pro handles resonance peaks (the LCR). It has always bothered me why Xover Pro does it the way that it does and what the difference is between that an running parallel legs. If I use the conversion that I came up with, PCD does model very much like Xover Pro for this design but that doesn't mean my conversion is correct.

Without the RC and LCR there is a hump about about 500-600 Hz with the mid range dome.

[Wolf]

Re: Electrical Schematic Question

Without the RC and LCR there is a hump about about 500-600 Hz with the mid range dome.

The Fs of a mid-dome a lot of times is compensated in a speaker. Otherwise it can sound a little harsh if you xover to close to the Fs. The zobel to flatten the impedance rise helps in the range of the mid to tweeter, and can ease the transition with less trouble.

I would keep the LCR for the mid's impedance, and go from there, but it's not required for a speaker. You can get by without it sometimes.

Good luck on the model-comparison!
Wolf

[dano]

Go

I think Xover Pro uses electrical and mechanical parameters to model crossovers. Can you model crossovers accurately with out frequency response curves and phase information?
________
Lamborghini Lm004 Specifications

[Pete Schumacher]

Re: Electrical Schematic Question

I think Xover Pro uses electrical and mechanical parameters to model crossovers. Can you model crossovers accurately with out frequency response curves and phase information?

In a word, no.

[whatatrip]

Re: Electrical Schematic Question

I think Xover Pro uses electrical and mechanical parameters to model crossovers. Can you model crossovers accurately with out frequency response curves and phase information?

Xover Pro can use FRD files. I use the very same FRD files for PCD and Xover Pro. In PCD I use ZMA files exported by WT3 and in Xover Pro I use the WT3 parameters for the electrical parameters. In this case, since I don't have drivers on hand, I use the manufacturer's T/S parameters and trace their graphs. Xover Pro can use just T/S parameters but if you import FRD data, the system response graphs change, phase graphs do not, impedance graphs do not and group delay graphs do not --- as far as I can tell. That tells me Xover Pro is getting its phase from the electical side but I am fairly new to all of this and phase hasn't sunk in yet.

I am thinking about a system with the RS270, RS52AN and the RS28F. The RS270's were on sale, now I am waiting for the others to go on sale. This is the first time I modeled from manufacturer's data. PCD uses driver offsets and Xover Pro does not. Small changes on the z axis, in particular, produce relatively substantial changes in PCD's system graphs. Even so, it is notIn as though there are great differences between the graphs of PCD and Xover Pro but there are differences particularly in the region where the woofers z axis comes into play.

With the last couple of designs I measured the drivers in the enclosures and used that data for PCD and Xover Pro with very similar system response graphs. In fact, I have posted both PCD and Xover Pro system graphs of the same systems in past posts. If you measure drivers in the enclosures, offsets are not needed as long as you don't want to model off axis.

Xover Pro tends to lead me down the path of more components than are really needed. That is one of the reasons why this thread is here and why I asked the question that I did. I can reduce components in this system by 7 if I delete impedance compensation on the woofer and mid but at the expense of a little less flat response and much larger capacitors. I still have to consider Wolf's response. This is a learning experience for me and I thank all for thier input. I have always tended to get source information from 2 or 3 places and then form my own opinion. The same is true for software. I figure if I get similar results in 2 or 3 programs, I might be on the right track.

[Jeff B.]

Re: Electrical Schematic Question

Xover Pro can use FRD files. I use the very same FRD files for PCD and Xover Pro. In PCD I use ZMA files exported by WT3 and in Xover Pro I use the WT3 parameters for the electrical parameters. In this case, since I don't have drivers on hand, I use the manufacturer's T/S parameters and trace their graphs. Xover Pro can use just T/S parameters but if you import FRD data, the system response graphs change, phase graphs do not, impedance graphs do not and group delay graphs do not --- as far as I can tell. That tells me Xover Pro is getting its phase from the electical side but I am fairly new to all of this and phase hasn't sunk in yet.

I am thinking about a system with the RS270, RS52AN and the RS28F. The RS270's were on sale, now I am waiting for the others to go on sale. This is the first time I modeled from manufacturer's data. PCD uses driver offsets and Xover Pro does not. Small changes on the z axis, in particular, produce relatively substantial changes in PCD's system graphs. Even so, it is notIn as though there are great differences between the graphs of PCD and Xover Pro but there are differences particularly in the region where the woofers z axis comes into play.

With the last couple of designs I measured the drivers in the enclosures and used that data for PCD and Xover Pro with very similar system response graphs. In fact, I have posted both PCD and Xover Pro system graphs of the same systems in past posts. If you measure drivers in the enclosures, offsets are not needed as long as you don't want to model off axis.

Xover Pro tends to lead me down the path of more components than are really needed. That is one of the reasons why this thread is here and why I asked the question that I did. I can reduce components in this system by 7 if I delete impedance compensation on the woofer and mid but at the expense of a little less flat response and much larger capacitors. I still have to consider Wolf's response. This is a learning experience for me and I thank all for thier input. I have always tended to get source information from 2 or 3 places and then form my own opinion. The same is true for software. I figure if I get similar results in 2 or 3 programs, I might be on the right track.

If you need to pad the midrange down too, then you can usually flatten the impedance with nothing more than a pair of resistors used for the padding. Below is a graph of the impedance of my PHL midrange using this method.

[Pete Schumacher]

Re: Electrical Schematic Question

Xover Pro can use FRD files. I use the very same FRD files for PCD and Xover Pro. In PCD I use ZMA files exported by WT3 and in Xover Pro I use the WT3 parameters for the electrical parameters. In this case, since I don't have drivers on hand, I use the manufacturer's T/S parameters and trace their graphs. Xover Pro can use just T/S parameters but if you import FRD data, the system response graphs change, phase graphs do not, impedance graphs do not and group delay graphs do not --- as far as I can tell. That tells me Xover Pro is getting its phase from the electical side but I am fairly new to all of this and phase hasn't sunk in yet.

I am thinking about a system with the RS270, RS52AN and the RS28F. The RS270's were on sale, now I am waiting for the others to go on sale. This is the first time I modeled from manufacturer's data. PCD uses driver offsets and Xover Pro does not. Small changes on the z axis, in particular, produce relatively substantial changes in PCD's system graphs. Even so, it is notIn as though there are great differences between the graphs of PCD and Xover Pro but there are differences particularly in the region where the woofers z axis comes into play.

With the last couple of designs I measured the drivers in the enclosures and used that data for PCD and Xover Pro with very similar system response graphs. In fact, I have posted both PCD and Xover Pro system graphs of the same systems in past posts. If you measure drivers in the enclosures, offsets are not needed as long as you don't want to model off axis.

Xover Pro tends to lead me down the path of more components than are really needed. That is one of the reasons why this thread is here and why I asked the question that I did. I can reduce components in this system by 7 if I delete impedance compensation on the woofer and mid but at the expense of a little less flat response and much larger capacitors. I still have to consider Wolf's response. This is a learning experience for me and I thank all for thier input. I have always tended to get source information from 2 or 3 places and then form my own opinion. The same is true for software. I figure if I get similar results in 2 or 3 programs, I might be on the right track.

Your driver selection had an example in the project showcase. I don't see it there now, but if you just build that system, you would be all set.

http://techtalk.parts-express.com/showthread.php?p=1633852#post1633852

He picked some pricey capacitors, but you don't have to stick with those types as long as you keep the values the same.

One approach you could take is to use Tony's XO to derive the transfer functions. Then you could replicate the transfer functions in a parallel XO, and likely save quite a few components in the process.

[whatatrip]

Re: Electrical Schematic Question

Your driver selection had an example in the project showcase. I don't see it there now, but if you just build that system, you would be all set.

One approach you could take is to use Tony's XO to derive the transfer functions. Then you could replicate the transfer functions in a parallel XO, and likely save quite a few components in the process.

The "Black Box" project is what caught my eye a couple of months ago. I wanted to do something similar to compare to the Dayton DC380, DC50 and DC28 system I made which turned out very nice. It has plenty of SPL so I thought I might try higher end drivers only with a smaller woofer and also because they don't make an RS380. Up to now I had forgotten about the Black Box but I knew somewhere I got the idea for the all RS but I first wanted to go it on my own and then go back and compare. Thanks for the effort to get the link. I have looked at it and remember now where that idea for the all RS came from.

LOL, I forgot how expensive the xover was. At the time I though it was because of the small enclosure being used and that if I went with a larger, vented enclosure I could use less parts. The Black Box uses 23 xover components. MY first fun Xover Pro used 18 components and my last try excluding the impedance equalization circuits has it down to 12 components per speaker. Instead of 10-15 uF caps in a couple of places I now have 40-60 mF caps. I have been using the cheaper electrolytic caps with a smaller metalized in parallel on the larger cap requirements to cut down on costs.

Looking at the slopes is a good idea. Comparing the Black Box (BB) to what I came up with is mostly similar. Woofer slope is the same, mid range on BB is a bit steeper on the woofer side and much steeper than mine on the tweeter side. The tweeter slope is steeper on the BB. It looks like the BB Mid/HF is xovered at about 3k and I have acoustic xover at 4.5k. So, I don't know what is best. Besides, I don't have the drivers yet. One of my purposes was to do a design without the drivers and then with the drivers and see how close they come to each other.

I do have L-pads on both the mid and tweeter.

So in summary, I gather that there really isn't an equivalent parallel leg circuit that is equal to the LCR that intersects the RC as Xover Pro does it. I wonder why they do that? I could call them since Harris has always been good at answering my questions.

[Pete Schumacher]

Re: Electrical Schematic Question

The "Black Box" project is what caught my eye a couple of months ago. I wanted to do something similar to compare to the Dayton DC380, DC50 and DC28 system I made which turned out very nice. It has plenty of SPL so I thought I might try higher end drivers only with a smaller woofer and also because they don't make an RS380. Up to now I had forgotten about the Black Box but I knew somewhere I got the idea for the all RS but I first wanted to go it on my own and then go back and compare. Thanks for the effort to get the link. I have looked at it and remember now where that idea for the all RS came from.

LOL, I forgot how expensive the xover was. At the time I though it was because of the small enclosure being used and that if I went with a larger, vented enclosure I could use less parts. The Black Box uses 23 xover components. MY first fun Xover Pro used 18 components and my last try excluding the impedance equalization circuits has it down to 12 components per speaker. Instead of 10-15 uF caps in a couple of places I now have 40-60 mF caps. I have been using the cheaper electrolytic caps with a smaller metalized in parallel on the larger cap requirements to cut down on costs.

Looking at the slopes is a good idea. Comparing the Black Box (BB) to what I came up with is mostly similar. Woofer slope is the same, mid range on BB is a bit steeper on the woofer side and much steeper than mine on the tweeter side. The tweeter slope is steeper on the BB. It looks like the BB Mid/HF is xovered at about 3k and I have acoustic xover at 4.5k. So, I don't know what is best. Besides, I don't have the drivers yet. One of my purposes was to do a design without the drivers and then with the drivers and see how close they come to each other.

I do have L-pads on both the mid and tweeter.

So in summary, I gather that there really isn't an equivalent parallel leg circuit that is equal to the LCR that intersects the RC as Xover Pro does it. I wonder why they do that? I could call them since Harris has always been good at answering my questions.

With the RS28 and RS52, a crossover point between 2KHz and 3KHz would be ideal. The RS52 loses some dispersion above 2500Hz, so having the RS28 cover that range would be an improvement from that standpoint.

I'm picturing a single padding input resistor, Cap, Inductor, Cap for the tweeter.

The mid can be brought to meet the woofer and tweeter LR4 with around 6 components, a shunt L and C, a series L and C, and a padding resistor, with perhaps a small cap to shunt the series inductor to tame the 13KHz resonance.

I'd think no more than 4 components would be required to work the RS270 into shape, at most 6 if you build a notch, including the input cap.

[mgrabow]

Re: Electrical Schematic Question

The problem you have in breaking the circuit up is because the action of Ce on the input to the second leg is going to be missing if you move it.

What is it you're trying to accomplish? You don't need to flatten the impedance. The crossover can be made to account for the impedance to produce the proper acoustic profile.

Educate me please... I have seen this comment in a few threads that a crossover can be made to account for impedence. Isnt that what the schematic in this thread does (as well as account resonance peak)?

[Jeff B.]

Re: Electrical Schematic Question

Educate me please... I have seen this comment in a few threads that a crossover can be made to account for impedence. Isnt that what the schematic in this thread does (as well as account resonance peak)?

What he means is that in most cases a crossover circuit can be made to produce the desired transfer function without doing anything special or extra to equalize the impedance. This often results in much fewer parts and nets the same results. The best way to learn this is by using a good crossover simulation tool.

[whatatrip]

Re: Electrical Schematic Question

Educate me please... I have seen this comment in a few threads that a crossover can be made to account for impedence. Isnt that what the schematic in this thread does (as well as account resonance peak)?

You can try this for yourself with one of your own designs. Attached I have two Xover Pro schematics that yield very similar results eventhough the first uses 18 components and the second uses 12 components. Graphs are also attached. I'm not an expert and have little experience so be forewarned.

I did email Harris Technologies and got some answers. I asked if my conversion to two parallel legs was accurate and support said "not exactly" but close enough for practical purposes. As I stated earlier, when I put it in PCD, system reponse was again very close to Xover Pro. They did point out an error in that I had Rm in there twice which was not my intention. The **Rm** should not be in there.

Leg1: Ce + Req
Leg2: Cm + Lm + Rm + Req = Cm + Lm + **Rm** + R(m+eq)

I also asked if there was any advantage to their methodology. The reply was that their circuit topology was "promoted by the late Dr. Bullock, inventor of many modern crossover network and impedance EQ circuits". I think I saw his name in a book but can't be sure. The reply also stated "Impedance EQ circuits are primarily used to smooth the impedance in the crossover region so the filters in a crossover network behave as expected. If the impedance is smooth for a couple of octaves before and after a crossover point, then they may not be needed. Plus, impedance EQ is also used to flatten the inductive reactance rise that is typical at upper frequencies."

I do remember reading that in the manual so I've tried to keep them fairly flat except at the very low end. That was discussed with reference to one of your questions I believe when you had a high value capacitor on a woofer. I'm not sure what is meant by a smooth impedance near the crossover. How much of a rise is still considered "smooth".

Whew, there really is a lot to know about all of this.

I also asked about future releases and there is one in the works which will include, among other things, modeling of notch filters.

[mgrabow]

Re: Electrical Schematic Question

You can try this for yourself with one of your own designs. Attached I have two Xover Pro schematics that yield very similar results eventhough the first uses 18 components and the second uses 12 components. Graphs are also attached. I'm not an expert and have little experience so be forewarned.

I did email Harris Technologies and got some answers. I asked if my conversion to two parallel legs was accurate and support said "not exactly" but close enough for practical purposes. As I stated earlier, when I put it in PCD, system reponse was again very close to Xover Pro. They did point out an error in that I had Rm in there twice which was not my intention. The **Rm** should not be in there.

Leg1: Ce + Req
Leg2: Cm + Lm + Rm + Req = Cm + Lm + **Rm** + R(m+eq)

I also asked if there was any advantage to their methodology. The reply was that their circuit topology was "promoted by the late Dr. Bullock, inventor of many modern crossover network and impedance EQ circuits". I think I saw his name in a book but can't be sure. The reply also stated "Impedance EQ circuits are primarily used to smooth the impedance in the crossover region so the filters in a crossover network behave as expected. If the impedance is smooth for a couple of octaves before and after a crossover point, then they may not be needed. Plus, impedance EQ is also used to flatten the inductive reactance rise that is typical at upper frequencies."

I do remember reading that in the manual so I've tried to keep them fairly flat except at the very low end. That was discussed with reference to one of your questions I believe when you had a high value capacitor on a woofer. I'm not sure what is meant by a smooth impedance near the crossover. How much of a rise is still considered "smooth".

Whew, there really is a lot to know about all of this.

I also asked about future releases and there is one in the works which will include, among other things, modeling of notch filters.

OK, I am still missing something. Both of your x-over pro designs have the same order crossover, same drivers and the same crossover points yet the components values the software suggests are different. I see where the response is similar and you obviously selected NOT to have the impedance equalization and resonant peak filter but if I remove those from mine it just removes them. It does not re-calculate the original component values of the 2nd order crossover. So what are you telling x-over pro to do differently that your photographs are not showing ? Other than the L-Pad on the mid...


secondly, with many many drivers that I have run in x-over pro, I get huge peaks and dips in the FR if I do not include impedance equalization. When I get home tonight, I will fun one of those and post the pictures...