To Jimmy154,
When you say you have the same speaker, do you mean that you purchased the woofers alone or that you actually have bought a Zetag Kinima G4. I know that Graham and I think his wife sell them on EBay. While I am terribly unimpressed with the Kinima G4, I would absolutely recommend purchasing just the woofers. I think Graham sells them really cheap. However, he says they are good up to 3KHz, and if you look at the response curve above, unless you have a very high order cross-order, there is no way you can use them in a quality set up to 3KHz IMOP.
I am still in the process of refining the cross-over, and will post it shortly. I used AudioAnalyser to do the tests of the units in the box, then actually used PSpice to model the basic cross-over. I am more of an electronics hobbyist, but enjoyed the process, and may invest in Calsod. I have downloaded speaker workshop, but have not had a chance to play with it.
I purchased all the components from SOLEN, but you could go to Madison, Parts Express, etc. I would like to try out one of the flat-film wire inductors to see if their is an improvement.
In terms of ingorance of Graham on cross-overs, I can't comment on that. I can say that in addition to the obvious response issues that I have, that the tweeter was not wired opposite in phase to the woofer as would be typical in a second order LR cross-over configuration. Of course, given that the responses are screwed, I guess that is not a problem.... 🙂
Alvaius
When you say you have the same speaker, do you mean that you purchased the woofers alone or that you actually have bought a Zetag Kinima G4. I know that Graham and I think his wife sell them on EBay. While I am terribly unimpressed with the Kinima G4, I would absolutely recommend purchasing just the woofers. I think Graham sells them really cheap. However, he says they are good up to 3KHz, and if you look at the response curve above, unless you have a very high order cross-order, there is no way you can use them in a quality set up to 3KHz IMOP.
I am still in the process of refining the cross-over, and will post it shortly. I used AudioAnalyser to do the tests of the units in the box, then actually used PSpice to model the basic cross-over. I am more of an electronics hobbyist, but enjoyed the process, and may invest in Calsod. I have downloaded speaker workshop, but have not had a chance to play with it.
I purchased all the components from SOLEN, but you could go to Madison, Parts Express, etc. I would like to try out one of the flat-film wire inductors to see if their is an improvement.
In terms of ingorance of Graham on cross-overs, I can't comment on that. I can say that in addition to the obvious response issues that I have, that the tweeter was not wired opposite in phase to the woofer as would be typical in a second order LR cross-over configuration. Of course, given that the responses are screwed, I guess that is not a problem.... 🙂
Alvaius
I bought the 8" mids on ebay because like you stated they were very good speakers for $70 USD a pair. I didn't really care that they didn't have a name of a company attached to them. I then checked out his other auctions and found the Zetag G3 for sale. I decided to build the G4 speaker and I e-mailed zetag for box and/or cross-over specs. They referred me to graham and I bought the drivers and cross-over for the G4 from him. I put them in a 46 liter vented box tuned to 43 Hz, doesn't sound to good, so I put straws in the port (taking Graham's advise) and I'm going to tune it to 38 Hz.
There are 2 versions of the woofer that I have right now. The ones that are sold on ebay are the ones Graham doesn't want (kind of an obvious statement I know); model # 24 10 00. But he says he accidently sent me the ones he uses in his speakers now; model # 24 07 00. And even before he told me this I noticed that they looked a little different and they sound a little better. One speaker has the good ones and one has the alright ones so it's easy to tell using the balance on my amp. Another set of good ones are on there way he says.
I was thinking of getting 2 of these http://www.tb-speaker.com/detail/w4-657s.htm drivers and using a 1st order cross-over in the speaker and build a sealed box. But I'm not sure that I want to build another set of boxes. Also, the sensativity of these and all other 3" or 4" woofers I've looked at is at least 3 db lower than that of the tweeter and woofers.
I can't seem to figure that speaker workshop. I think I need to invest in a microphone and do some measurements. Some body in this forum wrote that you can build a good one for $20-30, I think I will try that eventually. Anyway let me know what happens with your cross-over.
There are 2 versions of the woofer that I have right now. The ones that are sold on ebay are the ones Graham doesn't want (kind of an obvious statement I know); model # 24 10 00. But he says he accidently sent me the ones he uses in his speakers now; model # 24 07 00. And even before he told me this I noticed that they looked a little different and they sound a little better. One speaker has the good ones and one has the alright ones so it's easy to tell using the balance on my amp. Another set of good ones are on there way he says.
I was thinking of getting 2 of these http://www.tb-speaker.com/detail/w4-657s.htm drivers and using a 1st order cross-over in the speaker and build a sealed box. But I'm not sure that I want to build another set of boxes. Also, the sensativity of these and all other 3" or 4" woofers I've looked at is at least 3 db lower than that of the tweeter and woofers.
I can't seem to figure that speaker workshop. I think I need to invest in a microphone and do some measurements. Some body in this forum wrote that you can build a good one for $20-30, I think I will try that eventually. Anyway let me know what happens with your cross-over.
Jimmy,
Could you look in the area of the surround on the woofer and tell me what number\letter you see there and which woofers it matches up with. I am curious as to which woofers are in my units.
Could you look in the area of the surround on the woofer and tell me what number\letter you see there and which woofers it matches up with. I am curious as to which woofers are in my units.
The both have the Visaton logo and then "178B," right across from that there's a "3" on the alright ones and a "6" on that ones that sound better. I'm not sure how much better they sound though, I guess I will have a listen. Listening to them again, I thnk they sound almost the same I would say, it's hard to tell cause the ports sound so bad. I still say the "6" is just a little bit clearer and maybe has less of a peak at 3,750 like you were saying. The visible difference between the 2 is the "3" is more orange and the "6" is more pink or copper colored. Graham told me it's because the "3" is a less pure form of copper on it, which makes it more orange. I forget the name of the element that makes it orange but it begins with a "t" and ends in "ium," not titanium. The other visible difference is the "6" is about 2 mm bigger than "3" only cause the sides of the front flange are flared, I guess is the word. That is, not perpendicular to the surface they're mounted on.
Jimmy,
I am almost there on the cross-over. Parts should come in any day for the final version. I am using Solen CAPs on first pass, but will upgrade to Musicaps once I am happy with the way it works.
The thing that has me stumped is what to do with the second woofer. A 6 db boost at low frequencies is going to make it sound boomy. That is just way too much baffle step compensation in all but the largest rooms with the speakers well away from the back walls. So I want about 3db of boost at low frequencies. That means a 2.4ohm in series with the bottom woofer. There goes my nice low impedance amp out the window, though for what it is worth, the inductors in the speakers as delivered were over 1ohm of DC resistance!
I will probably build them today with the attenuation, knowing the difficulty, and then convert them to an active design for the lower woofers when I have the time.
Anyone else have any thoughts?? dual 8" woofer - indentical, 1" tweeter. Want baffle step compensation with lower woofer at about 400Hz, but only about 3db. The woofers have almost exactly the same efficiency as the tweeters. Upper woofer crossing over to tweeter at 2.45KHz. Must be a better way of doing this.....
Alvaius
I am almost there on the cross-over. Parts should come in any day for the final version. I am using Solen CAPs on first pass, but will upgrade to Musicaps once I am happy with the way it works.
The thing that has me stumped is what to do with the second woofer. A 6 db boost at low frequencies is going to make it sound boomy. That is just way too much baffle step compensation in all but the largest rooms with the speakers well away from the back walls. So I want about 3db of boost at low frequencies. That means a 2.4ohm in series with the bottom woofer. There goes my nice low impedance amp out the window, though for what it is worth, the inductors in the speakers as delivered were over 1ohm of DC resistance!
I will probably build them today with the attenuation, knowing the difficulty, and then convert them to an active design for the lower woofers when I have the time.
Anyone else have any thoughts?? dual 8" woofer - indentical, 1" tweeter. Want baffle step compensation with lower woofer at about 400Hz, but only about 3db. The woofers have almost exactly the same efficiency as the tweeters. Upper woofer crossing over to tweeter at 2.45KHz. Must be a better way of doing this.....
Alvaius
I'm not familar with musicaps. I thought Solen fast caps were considered to be very good.
I don't know if 1 ohm DC resistance is a problem. I thought it was only a problem when you didn't calculate it into the nominal resistance of your woofer. If your inductor is going to be about 1 ohm and woofer is 8 design it so it will be 9 ohm resistence. You would need a mathmatical equation to get it very exact, but this would be a pretty accurate way of doing, if not it's a good start or the third option and most likely, I don't know what I'm talking about.
The Auriga on this link http://home.hetnet.nl/~geenius/ is a design similar to the G4, except the top woofer is sealed and the bottom is ported (wish I thought of that). He says this about the cross-over for the bottom woofer: "The woofer rolls off smoothly with a nice and simple 1st order crossover using a single transformer type inductor and a RC-network to correct the impedance rise and tailor the output in the midrange.
The RC-network acts as a fine-tuning instrument to blend in perfectly the woofer with the mid- woofer." I guess it's called an impedance equilization circuit, not sure if this solves your problem though. I'm not even sure what impedance is. That all I got for ideas so far.
I don't know if 1 ohm DC resistance is a problem. I thought it was only a problem when you didn't calculate it into the nominal resistance of your woofer. If your inductor is going to be about 1 ohm and woofer is 8 design it so it will be 9 ohm resistence. You would need a mathmatical equation to get it very exact, but this would be a pretty accurate way of doing, if not it's a good start or the third option and most likely, I don't know what I'm talking about.
The Auriga on this link http://home.hetnet.nl/~geenius/ is a design similar to the G4, except the top woofer is sealed and the bottom is ported (wish I thought of that). He says this about the cross-over for the bottom woofer: "The woofer rolls off smoothly with a nice and simple 1st order crossover using a single transformer type inductor and a RC-network to correct the impedance rise and tailor the output in the midrange.
The RC-network acts as a fine-tuning instrument to blend in perfectly the woofer with the mid- woofer." I guess it's called an impedance equilization circuit, not sure if this solves your problem though. I'm not even sure what impedance is. That all I got for ideas so far.
" The only sound way to design a xover is to perform acoustic measurements on the drivers in the box. "
Dick, I agree with you. There is no other way. Theoretical computations will only give an order of magnitude. Commercial Xovers are usable only with specified drivers *and* enclosures.
Sufficiently accurate measurements can be done with correct electret microphone, and a PC with standard sound-card.
Regards, Pierre Lacombe.
Dick, I agree with you. There is no other way. Theoretical computations will only give an order of magnitude. Commercial Xovers are usable only with specified drivers *and* enclosures.
Sufficiently accurate measurements can be done with correct electret microphone, and a PC with standard sound-card.
Regards, Pierre Lacombe.
You may can eliminate the resonace with a R/C
combination i borrowed also for my actual speakers.
On this page you can see the 5,1R 0.33ü is killing
a resonance with a simliar shape than yours.
http://members.aol.com/spkpreise/al17proj.htm
Since your resonance is at lower fequency a 0,47
or 0,68 will just do fine.
Wombat
combination i borrowed also for my actual speakers.
On this page you can see the 5,1R 0.33ü is killing
a resonance with a simliar shape than yours.
http://members.aol.com/spkpreise/al17proj.htm
Since your resonance is at lower fequency a 0,47
or 0,68 will just do fine.
Wombat
Sorry my last post didn't make any sense cause I get the words resistance and impedance confused all the time. I remember now what the RC or zobel network flattens off the upper range of an impedance graph of a driver.
Looks like I have to design my own cross-over.
Wombat or anyone,
How do you know what combination of R/C to use for this type of filter? I've never heard of this type of filter (in parellel, not zobel). How do you know I should use a .47 or .68 uF capacitor
with a 5.1 ohm resistor.
One more thing . . . Does anyone know how to measure how many ohms a resistor is. I got a bunch that aren't labeled. I tried to use a voltmeter that has a ohms switch on it, but I read a resistance of infinity. Does the resistor have to have power running through it so I can measure the resistance?
Wombat or anyone,
How do you know what combination of R/C to use for this type of filter? I've never heard of this type of filter (in parellel, not zobel). How do you know I should use a .47 or .68 uF capacitor
with a 5.1 ohm resistor.
One more thing . . . Does anyone know how to measure how many ohms a resistor is. I got a bunch that aren't labeled. I tried to use a voltmeter that has a ohms switch on it, but I read a resistance of infinity. Does the resistor have to have power running through it so I can measure the resistance?
The resistor ist just to prevent a short at higher frequencies.
The correct value i need i simple measure.
When i remember right, 0.47 works around 5 kHz and 0,68 should
be around 4kHz like you need it.
There is sure a way to calculate but i don´t need it so i don´t
have a formula.
The man on the page i linked to uses Speaker Pro 7.0 or HR97
(whatever that is). They used the "advanced universal filter"
to simulate this.
Sorry for not having more info.
The correct value i need i simple measure.
When i remember right, 0.47 works around 5 kHz and 0,68 should
be around 4kHz like you need it.
There is sure a way to calculate but i don´t need it so i don´t
have a formula.
The man on the page i linked to uses Speaker Pro 7.0 or HR97
(whatever that is). They used the "advanced universal filter"
to simulate this.
Sorry for not having more info.
I think I need it at 3,750 Hz. I guess I will try around 0.73 uF. If I try this kind of cross-over. I haven't found anything on the web explaining this kind of filter. I translated that link you gave me the best I could into English. He seems to have 2 different cross-overs for the same speaker. The one you're referring to was changed to the one lower on the page, according to my translation of that page.
On his page the below is a different project from
the factory Visaton. His suggestion with the 5,1 Ohm
resistor should be safer to use. It prevents swinging
of your amplifier at high frequencys around 100-200kHz
that can even break it.
Did you try a translation at worldlingo.com?
http://www.worldlingo.com/products_services/worldlingo_translator.html
Some funny results but some things make sense!
the factory Visaton. His suggestion with the 5,1 Ohm
resistor should be safer to use. It prevents swinging
of your amplifier at high frequencys around 100-200kHz
that can even break it.
Did you try a translation at worldlingo.com?
http://www.worldlingo.com/products_services/worldlingo_translator.html
Some funny results but some things make sense!
No I didn't try worldlingo.com. I got the same result for altavista.com. I can kind of understand some things, but I didn't pick up on him revising the x-over.
I think I will have to measure the frequency response of my drivers to design a proper cross-over, which I've been trying to do, on and off, for a lot of months with no success. This is very dumb, when I desided to build these speakers this summer, I didn't know it was gonna be all this.
I've been trying to get this microphone to work for months now. They say that if you solder it and it gets to hot that it will get ruined. So I don't know if the mic is bad or I'm missing something with the Speaker Workshop. Speakerwork shop is the only program I found that measures frequency response and I can't even design a dumba** box using that nerd program. The soldering points are so small on the mic that I don't know if the wires are touching. I've once again lost what little patience I have, with the microphone and/or Speaker Workshop.
I will also try to figure out exactly what kind of filter it is that he uses, so I can figure out what value I need at 3,750 Hz.
I think I will use a 12 db-octive x-over for these speakers at 2,500 Hz because I have 1 mH inductors already and all I have to get is 4 - 4 uF caps. I will use the filter you gave me to filter out that peak, as soon as I figure out what value I need for the uF and what it does to the frequency response (besides filtering the peak).
I guess I will LP the bottom woofer at 400 Hz using 3 - 1 mH inductors, to give me 3 mH. I don't know if this will be a problem for the top midrange woofer because it is in the same sealed enclosure (I put an old shirt in each port that was tuned to 38 Hz cause it sound awful) as the bottom woofer. It's in a 47 liter enclosure and has a VAS of 42.23 liters. So if I'm listening to something that has frequencies above 400 Hz it will act as if it's in free air. I don't know if this is good. I would think not. Does anybody know?
I think I will have to measure the frequency response of my drivers to design a proper cross-over, which I've been trying to do, on and off, for a lot of months with no success. This is very dumb, when I desided to build these speakers this summer, I didn't know it was gonna be all this.
I've been trying to get this microphone to work for months now. They say that if you solder it and it gets to hot that it will get ruined. So I don't know if the mic is bad or I'm missing something with the Speaker Workshop. Speakerwork shop is the only program I found that measures frequency response and I can't even design a dumba** box using that nerd program. The soldering points are so small on the mic that I don't know if the wires are touching. I've once again lost what little patience I have, with the microphone and/or Speaker Workshop.
I will also try to figure out exactly what kind of filter it is that he uses, so I can figure out what value I need at 3,750 Hz.
I think I will use a 12 db-octive x-over for these speakers at 2,500 Hz because I have 1 mH inductors already and all I have to get is 4 - 4 uF caps. I will use the filter you gave me to filter out that peak, as soon as I figure out what value I need for the uF and what it does to the frequency response (besides filtering the peak).
I guess I will LP the bottom woofer at 400 Hz using 3 - 1 mH inductors, to give me 3 mH. I don't know if this will be a problem for the top midrange woofer because it is in the same sealed enclosure (I put an old shirt in each port that was tuned to 38 Hz cause it sound awful) as the bottom woofer. It's in a 47 liter enclosure and has a VAS of 42.23 liters. So if I'm listening to something that has frequencies above 400 Hz it will act as if it's in free air. I don't know if this is good. I would think not. Does anybody know?
First: Good luck with your microphone. Calculating is just relevant to
some degrees. You need all parameters of the speakers within a program
that can handle this to come only near to the real world.
Mounting, geometry, and even the material of the boxes will influencs all
things you can´t calculate correctly. There are many more factors.
I only calculate to have starting points for not using to wrong filters.
Afterwards i try to flatten the frequency response to an acceptable point.
then the real tuning with pure listening starts. Parts only vary a bit from
there on but do much, much influence!
I never calculated very near the finished boxes.
Sorry to make it sound even more complicated - but this is how it goes.
Another thing is putting 3 inductors in series. When they are not really low
resistance all calculations for the volume go to hell if these Ohm values are
not into the calculation.
Just some thoughts, others here hopefully will help you further.
some degrees. You need all parameters of the speakers within a program
that can handle this to come only near to the real world.
Mounting, geometry, and even the material of the boxes will influencs all
things you can´t calculate correctly. There are many more factors.
I only calculate to have starting points for not using to wrong filters.
Afterwards i try to flatten the frequency response to an acceptable point.
then the real tuning with pure listening starts. Parts only vary a bit from
there on but do much, much influence!
I never calculated very near the finished boxes.
Sorry to make it sound even more complicated - but this is how it goes.
Another thing is putting 3 inductors in series. When they are not really low
resistance all calculations for the volume go to hell if these Ohm values are
not into the calculation.
Just some thoughts, others here hopefully will help you further.
Definitely the fourth-order filter is the way to go. The notch filter will introduce phase shifting that will manifest itself smack-dab in the middle of the upper mid-range. You will be able to hear it.
When you measure this filter it only works on
a very limitted range.
4th order passive will do much more trouble from only
choosing the parts i guess.
But with only calculating - ok - less space for errors.
a very limitted range.
4th order passive will do much more trouble from only
choosing the parts i guess.
But with only calculating - ok - less space for errors.
Wombat said:
I'm not sure what you're saying in this statement.
A 4th order x-over will give me more trouble because it's hard to choose the right parts. Are you saying once you have bought your parts it's harder to refine or change things?
When you say calculating you're talking about calculating the x-over point.
Also you think the second order x-over with the filter is better option than 4th order with no filter. One more thing I just realised, to select what compnents I want for the x-over I use the impedance at the x-over point not the nominal impedance right?
The way I saw 3 inductors of 1mH wired to make a 3 mH inductance. 1 is parellel on the posative wire, 1 is parellel one the negative wire, and 1 is in series.
It is very hard and expensive to get good sounding crossover parts for a 4th order.
Every part influences the sound itself. Especially 2 in series -
what you need for 4th order will get complicated or expensive.
This is my experience.
When i say calculating a 4th order you eliminate more problems
with that kind of spikes like your chassis have with the fact that
they are more suppressed from the filter itself when you aren´t aware
of this kind of problem.
When you exactly can localize problems and are able to eliminate this with
a lower order filter it sure will sound better.
Less parts on the way thru the switch is always to favour.
?? to get 3mH from 3x1mH it has to be in series,
contrary to capacitors!
Every part influences the sound itself. Especially 2 in series -
what you need for 4th order will get complicated or expensive.
This is my experience.
When i say calculating a 4th order you eliminate more problems
with that kind of spikes like your chassis have with the fact that
they are more suppressed from the filter itself when you aren´t aware
of this kind of problem.
When you exactly can localize problems and are able to eliminate this with
a lower order filter it sure will sound better.
Less parts on the way thru the switch is always to favour.
?? to get 3mH from 3x1mH it has to be in series,
contrary to capacitors!
Okay, I think I'm almost ready to order the parts . . .
I have a few more questions.
I'm going to copy alvaius and go with the cross-over point at 2.3 KHz cause he knows better than me what it should be. And the lower woofer will be LP at 6 db/octive at 400 Hz coppying alvaius. I will use the R/C combination Wombat gave me at 5.1 ohms and .68 or .73 uF cap. Where or how can I find out the exact value? Any body every hear of this kind of filter?
When I got the values for my x-over I use the impedances at the x-over frequencies, right? Not the nominal impedances. 7.09 ohms for tweeter and 13.6 ohms for woofer at 2.3 KHz. And for bottom woofer at 400 Hz it's around 7 ohms I would say. Do I need a zobel network for the bottom woofer? Impedance was taken by alvaius at this thread http://www.diyaudio.com/forums/showthread.php?s=&threadid=4205&highlight=zetag
As you might have guessed I plan to compensate for the baffle step with the bottom woofer. The baffle is 11" (27.5 cm) wide, with the corners angle cut I think it will be like 10" (25 cm). So I think the baffle step will be like 500 Hz (don't know exactly yet, any one know a good website for baffle step). My question is how many db down should the bottom woofer be at 500 Hz? I was thinking the bottom woofer should be at least 3 db down at 500 hz.
Also I have a bunch of 18 gauge inductors that I might use is it bad to use these for sound quality or is it okay as long as I factor in the resistance of these inductors into my calculations?
And I'm not going with resistors for the tweeter cause the woofer and the tweeter are both around 91 db sensativity. Is this a good idea? Do I need resistors or attenuation circuit?
Does anybody know how to measure the impedance of resistors? I used an ohms/volt meter, but I got an impedance of infinfity. Does it have to have power running threw it when you measure the impedance?
Does it matter that the top and bottom woofer are now in the same sealed (with T-shirts in port) enclosure? If playing top woofer frequencies (above 500 Hz only) and lower woofer isn't moving the top woofer will act like it's in free air, I think. 1 - woofers VAS is 42.22 liters and the box is 47 liters. This is going to be bad right? What should I do about this? Try to make the top woofer sealed (might be to late though)?
As you can see I don't know exactly what I'm doing, so any help is appreciated.
I have a few more questions.
I'm going to copy alvaius and go with the cross-over point at 2.3 KHz cause he knows better than me what it should be. And the lower woofer will be LP at 6 db/octive at 400 Hz coppying alvaius. I will use the R/C combination Wombat gave me at 5.1 ohms and .68 or .73 uF cap. Where or how can I find out the exact value? Any body every hear of this kind of filter?
When I got the values for my x-over I use the impedances at the x-over frequencies, right? Not the nominal impedances. 7.09 ohms for tweeter and 13.6 ohms for woofer at 2.3 KHz. And for bottom woofer at 400 Hz it's around 7 ohms I would say. Do I need a zobel network for the bottom woofer? Impedance was taken by alvaius at this thread http://www.diyaudio.com/forums/showthread.php?s=&threadid=4205&highlight=zetag
As you might have guessed I plan to compensate for the baffle step with the bottom woofer. The baffle is 11" (27.5 cm) wide, with the corners angle cut I think it will be like 10" (25 cm). So I think the baffle step will be like 500 Hz (don't know exactly yet, any one know a good website for baffle step). My question is how many db down should the bottom woofer be at 500 Hz? I was thinking the bottom woofer should be at least 3 db down at 500 hz.
Also I have a bunch of 18 gauge inductors that I might use is it bad to use these for sound quality or is it okay as long as I factor in the resistance of these inductors into my calculations?
And I'm not going with resistors for the tweeter cause the woofer and the tweeter are both around 91 db sensativity. Is this a good idea? Do I need resistors or attenuation circuit?
Does anybody know how to measure the impedance of resistors? I used an ohms/volt meter, but I got an impedance of infinfity. Does it have to have power running threw it when you measure the impedance?
Does it matter that the top and bottom woofer are now in the same sealed (with T-shirts in port) enclosure? If playing top woofer frequencies (above 500 Hz only) and lower woofer isn't moving the top woofer will act like it's in free air, I think. 1 - woofers VAS is 42.22 liters and the box is 47 liters. This is going to be bad right? What should I do about this? Try to make the top woofer sealed (might be to late though)?
As you can see I don't know exactly what I'm doing, so any help is appreciated.
- Status
- Not open for further replies.