Im back
Hi! after a couple of years im now back to pick up where i left this project. I've learned a great deal about speakerbuilding since last time, and i now feel ready to give this project another try.
The speakers will be used mostly to play heavy metal - loud! So i need good punch in the mid and upper-bass area.
I desided to go for the 8" 4ohm silverflute bass, tb w4 mid and hivi k1 tweeter. both C/O's are relatively high up in frequency - 600hz and 6000hz. This way i hope to get the most spl out of the mid and the tweeter.
You guys think this setup could give me some proper output?
Another Q: I found this kit: http://www.madisound.com/catalog/product_info.php?products_id=8525
You think i could use these drivers together with the silverflute 8"? I'm just not sure if the vifa woofer is able to play loud enough without distorting.
Hi! after a couple of years im now back to pick up where i left this project. I've learned a great deal about speakerbuilding since last time, and i now feel ready to give this project another try.
The speakers will be used mostly to play heavy metal - loud! So i need good punch in the mid and upper-bass area.
I desided to go for the 8" 4ohm silverflute bass, tb w4 mid and hivi k1 tweeter. both C/O's are relatively high up in frequency - 600hz and 6000hz. This way i hope to get the most spl out of the mid and the tweeter.
You guys think this setup could give me some proper output?
Another Q: I found this kit: http://www.madisound.com/catalog/product_info.php?products_id=8525
You think i could use these drivers together with the silverflute 8"? I'm just not sure if the vifa woofer is able to play loud enough without distorting.
Hi skogs, long time no see 
Well, 3way is a pain, no doubt
I have made lots of changes on mine
Basicly the crossover is much simpler now
Though I have achieved great improvement, the phase issues is still a struggle
Mind you, even if you do good simulations, what you actually get could be far from the simulated
Thats the ironic part of simulations, you still need measurements of drivers in actual box
Simulating is just a useful tool to make measurements easier, with less trial and errors trying xo designs
I have tried to make it 18db/third order
But phase seems to go crazy
Be aware that even 6db may result in 24db acoustical rolloff, if placed close to drivers own rolloff
That may be a problem with your 6khz xo point
Anyway, I wouldnt use anything higher tha 12db without measurements
In the future I will do some 2ways passive
But any 3way will probably be active
For what its worth, my xo looks like this at the moment
If only my mids had better sensitivity
The whole speaker and xo curcuit are exstremely sensitive to even so small changes, one could wonder if its measureable at all
The other day I removed a bypass cap on one channel
The other channel I only disconnected the cap, andleft it connected with one leg only
So, it shouldnt do anything at all
But I could actually hear that the channel with the loose hanging cap was a bit off
A bit strange
btw, if you havent bought a woofer, MIVOC have some nice cheap ones
Anyway, to take a 2way design and make it into a 3way is asking for trouble
Well, 3way is a pain, no doubt
I have made lots of changes on mine
Basicly the crossover is much simpler now
Though I have achieved great improvement, the phase issues is still a struggle
Mind you, even if you do good simulations, what you actually get could be far from the simulated
Thats the ironic part of simulations, you still need measurements of drivers in actual box
Simulating is just a useful tool to make measurements easier, with less trial and errors trying xo designs
I have tried to make it 18db/third order
But phase seems to go crazy
Be aware that even 6db may result in 24db acoustical rolloff, if placed close to drivers own rolloff
That may be a problem with your 6khz xo point
Anyway, I wouldnt use anything higher tha 12db without measurements
In the future I will do some 2ways passive
But any 3way will probably be active
For what its worth, my xo looks like this at the moment
If only my mids had better sensitivity
The whole speaker and xo curcuit are exstremely sensitive to even so small changes, one could wonder if its measureable at all
The other day I removed a bypass cap on one channel
The other channel I only disconnected the cap, andleft it connected with one leg only
So, it shouldnt do anything at all
But I could actually hear that the channel with the loose hanging cap was a bit off
A bit strange
btw, if you havent bought a woofer, MIVOC have some nice cheap ones
Anyway, to take a 2way design and make it into a 3way is asking for trouble
Attachments
Hello tinitus! I hoped to get a reply from you, since you helped me so much the last time
Nice and simple xo! I notice that your low pass between mid and woofer is much lower than my current design, where is the xo point here? 2-300hz? why did you decide to cross over that low?
Are your mids polarity reversed?
I've now ordered the 2-way kit, only reasen i did was because it was so darn cheap for a couple of good drivers. The 2-way xo will probably be chopped up, so i can use the parts if i need'em. I'll try using both the tb w4 and the vifa tc14 (seperatly) and see which does the best job. Also ordered the SF 8" 4 ohm woofers.
Btw, i've have used your tips with placing a resistor in series with all paralell components, and with great success! This techique can be of great help when manipulating the drivers behaviour.
I use a quit advanced modelling software that allows me to simulate the xo's phase response. But does driver placement relative to each other be an issue here too? The phase response really is a bitch.
I havent taken the drivers natural roll-off into consideration yet, i've mainly worked with the phase, and rougly the spl-response. There will probably be alot of things that need to be altered for the final design.
Do you have a graph of some sort for your design? measured or simulated?
Nice and simple xo! I notice that your low pass between mid and woofer is much lower than my current design, where is the xo point here? 2-300hz? why did you decide to cross over that low?
Are your mids polarity reversed?
I've now ordered the 2-way kit, only reasen i did was because it was so darn cheap for a couple of good drivers. The 2-way xo will probably be chopped up, so i can use the parts if i need'em. I'll try using both the tb w4 and the vifa tc14 (seperatly) and see which does the best job. Also ordered the SF 8" 4 ohm woofers.
Btw, i've have used your tips with placing a resistor in series with all paralell components, and with great success! This techique can be of great help when manipulating the drivers behaviour.
I use a quit advanced modelling software that allows me to simulate the xo's phase response. But does driver placement relative to each other be an issue here too? The phase response really is a bitch.
I havent taken the drivers natural roll-off into consideration yet, i've mainly worked with the phase, and rougly the spl-response. There will probably be alot of things that need to be altered for the final design.
Do you have a graph of some sort for your design? measured or simulated?
Attachments
skogs said:
I dont know much about sim software, but Im sure theres a function for driver offset
If you have response of your drivers you can import them into your sim and get a better result
Do your sim take care of BSC
New simpler premature xo
I dont know where Im crossing, I only go for the best phase, and the rest comes by it self
Attachments
The modelling software isnt really a speaker design software, its mostly for advanced curcuit development, got it from my brother who is literatly a rocket scientist So i have no idea of how i can implement either driver offset or measured spl-response.
I was just reading about driver offset in "Vance Dickasons, Loud Speaker Design Cookbook". Quite interresting actually. I now understand better some of the reasons why simulated and actual spl-response can differ so much. Basically, when a tweeter and a woofer is mounted on the same baffel, there will in most cases be an "depth" offset between the drivers, where the woofers acustically center is deeper than that of the tweeter. This will cause a time delay, which will affect the phase at some frequencys, normally 1-3khz. Simulating this will be very difficult, so this is truly a trial and error aspect of crossover designing, in my view.
I use a dedicated BSC simulation program, "Edge", to simulated the baffels affect on the driver. I then modell the spl-response so it matches the losses caused by baffle-step. As you can see on my first graph (this year), the spl-response rising gradually towards the bass area.
would you advice me to cross over lower? Actually the mid woofer covers pretty much two octaves, i think im just using a bit of a weird algorithm for the simulating graph, compared to what people normally do, leaving the graph a little hard to read.
I like your way of designing your crossover, i notice you have omitted the values of some of your inductors. I guess you adjust them by ear, leaving you unaware of the value?
So i have no idea of how i can implement either driver offset or measured spl-response.I was just reading about driver offset in "Vance Dickasons, Loud Speaker Design Cookbook". Quite interresting actually. I now understand better some of the reasons why simulated and actual spl-response can differ so much. Basically, when a tweeter and a woofer is mounted on the same baffel, there will in most cases be an "depth" offset between the drivers, where the woofers acustically center is deeper than that of the tweeter. This will cause a time delay, which will affect the phase at some frequencys, normally 1-3khz. Simulating this will be very difficult, so this is truly a trial and error aspect of crossover designing, in my view.
I use a dedicated BSC simulation program, "Edge", to simulated the baffels affect on the driver. I then modell the spl-response so it matches the losses caused by baffle-step. As you can see on my first graph (this year
), the spl-response rising gradually towards the bass area.would you advice me to cross over lower? Actually the mid woofer covers pretty much two octaves, i think im just using a bit of a weird algorithm for the simulating graph, compared to what people normally do, leaving the graph a little hard to read.
I like your way of designing your crossover, i notice you have omitted the values of some of your inductors. I guess you adjust them by ear, leaving you unaware of the value?
skogs said:
Hi skogs
No, I cant say what you should do
That will be up to you to judge how your design works
But your design does demand more of the woofer,and maybe tweeter, depending
If drivers are up to it, and it sounds good, then it is good
But you use an 8" fit for a 2way, so its of less concern
Yes, I adjust by ear
Im too stupid to learn all that modern stuff
But sure, I would like to measure, simulate etc
But I dont think I would ever trust any of it 100%
I can masure caps and inductors, and do know the value
But I have to get them out of active curcuit to measure
When I show schematic here I really should remove all component values
Schematic works for me, and I will use it for another speaker as well, with adjustments
But it may not work for others
Its just meant to b an inspiration
Maybe also help in understanding each other better
The drivers have now been shipped, and im really excited! I've lowered woofer-mid xo and raised up mid-tweeter xo. Main reason was to make BFC easier to tweak, now i can probably just attenuate mid and tweeter to get desired BFC-slope. Raising tweeter xo is for better power handling.
Main reason for tweaking the filter at this point, is to get to now the filter better, so that i more easily can tweak it later, when problems occur with the real thing
Btw, i find the calculators on this site really helpful: http://www.diyaudioandvideo.com/
Main reason for tweaking the filter at this point, is to get to now the filter better, so that i more easily can tweak it later, when problems occur with the real thing
Btw, i find the calculators on this site really helpful: http://www.diyaudioandvideo.com/
Now i have recieved all the drivers, and built enclosures for them. I used 19mm mdf and lots of bracing. The woofer got 25L to play in and the mid has 6-ish L. Baffle is 22,5x110cm.
I'm still waiting for the parts for the crossover, but i have measured the drivers in the box. I have a big problem in the 100-200hz area, i think its both baffle-step and room modes that causes the problem. See the attatchment.
Any ideas how i should deal with this problem? I believe this area in the frequency range is quite vital, especially since im a heavy metall\hard rock fan.
I'm still waiting for the parts for the crossover, but i have measured the drivers in the box. I have a big problem in the 100-200hz area, i think its both baffle-step and room modes that causes the problem. See the attatchment.
Any ideas how i should deal with this problem? I believe this area in the frequency range is quite vital, especially since im a heavy metall\hard rock fan.
I have now designed a crossover filter that more or less removes the dip. The problem now is that the impedance decends towards 3ohm below 250hz. The reason for this is that the system is now more a 2.5-way, as the "mid" now only has a lowpassfilter.
I will be driving the speakers with a high quality Harman Kardon 340 reciever, do you think the low impedance will cause problems?
EDIT:
Nevermind the above, the powerhandling of the mid would totally ruin the consept of this speaker.
I've now simplified the filter. But a large series inductor with the woofer and lots of bsc causes low sensitivity. Probably +- 85db/1.83V.
I will be driving the speakers with a high quality Harman Kardon 340 reciever, do you think the low impedance will cause problems?
EDIT:
Nevermind the above, the powerhandling of the mid would totally ruin the consept of this speaker.
I've now simplified the filter. But a large series inductor with the woofer and lots of bsc causes low sensitivity. Probably +- 85db/1.83V.
So, you are testing 2 different speaker pairs?
WO: Silver Flute W20RC38-04 ohm 8"
Mid: Tangband W4-657SC 8 ohm 4"
Tw: Hi-Vi K1 6 ohm 1"
WO: Silver Flute W20RC38-04 ohm 8"
Mid: Vifa TC14WG49-08 ohm 5"
Tw: Vifa BC25TG15-04 ohm 1" or Hi-Vi K1 6 ohm 1"
Note: also check diyAudio Forums - W4-657sc
http://www.diyaudio.com/forums/showthread.php?postid=608950
WO: Silver Flute W20RC38-04 ohm 8"
Mid: Tangband W4-657SC 8 ohm 4"
Tw: Hi-Vi K1 6 ohm 1"
WO: Silver Flute W20RC38-04 ohm 8"
Mid: Vifa TC14WG49-08 ohm 5"
Tw: Vifa BC25TG15-04 ohm 1" or Hi-Vi K1 6 ohm 1"
Note: also check diyAudio Forums - W4-657sc
http://www.diyaudio.com/forums/showthread.php?postid=608950
Puh, i have now worked on the XO for this speaker ALOT, several hours a day, every day since i got the drivers.
I have now learned that simulating filters is of no use, what so ever. The only way i could do it, was trial and error (mostly error).
The biggest problems so far is first of all the woofers sensitivity, which is MUCH lower than anticipated. The tweeter is attenuated by 6-7db, And the mids sensitivity is lowered a little by the XO, probably because LP and HP rolloffs are relatively close. And still the speaker sounds to bright. XO's are 700hz and 2800hz. Bass\mid got 2. order acustical xo, and mid\tweeter got 3. order electrical xo.
Otherwise, its really hard to give a subjective impression of the speaker, because of the major issue at 100-150hz. When turning up the volume, the speaker doesnt seem to distort, but it gives more of that annyoing kind of loudness, where you just want to turn down the volume again. Seems like it just gets to bright, even though i attenuated the tweeter by 7db.
Im using a Radioshack digital db-meter, hooked up to my computer without proper calibration in the upper frequencies, but i've used one of "Troels Gravesens" designs as reference, and my speaker is supposed to have a SPL-response that drops a little more towards the higher frequencies, and still it sounds brighter (more annoying) than the reference speaker.
Any ideas of how to get a warmer sound from these speakers? Could the cold tone balance be all because of the 100-150hz issues, or could it be some kind of distortion that i havent taken into consideration?
I'm probably going to hate myself for a while, for ordering 2 pcs of the 4ohm 8" and not 4 pcs of the 8ohm 8". Double 8"'s would probably kick **ss compared to single.
I have now learned that simulating filters is of no use, what so ever. The only way i could do it, was trial and error (mostly error).
The biggest problems so far is first of all the woofers sensitivity, which is MUCH lower than anticipated. The tweeter is attenuated by 6-7db, And the mids sensitivity is lowered a little by the XO, probably because LP and HP rolloffs are relatively close. And still the speaker sounds to bright. XO's are 700hz and 2800hz. Bass\mid got 2. order acustical xo, and mid\tweeter got 3. order electrical xo.
Otherwise, its really hard to give a subjective impression of the speaker, because of the major issue at 100-150hz. When turning up the volume, the speaker doesnt seem to distort, but it gives more of that annyoing kind of loudness, where you just want to turn down the volume again. Seems like it just gets to bright, even though i attenuated the tweeter by 7db.
Im using a Radioshack digital db-meter, hooked up to my computer without proper calibration in the upper frequencies, but i've used one of "Troels Gravesens" designs as reference, and my speaker is supposed to have a SPL-response that drops a little more towards the higher frequencies, and still it sounds brighter (more annoying) than the reference speaker.
Any ideas of how to get a warmer sound from these speakers? Could the cold tone balance be all because of the 100-150hz issues, or could it be some kind of distortion that i havent taken into consideration?
I'm probably going to hate myself for a while, for ordering 2 pcs of the 4ohm 8" and not 4 pcs of the 8ohm 8". Double 8"'s would probably kick **ss compared to single.
I have now finished both speakers and crossovers, worked alot on placement and fine-tuning the phase between subwoofers, horn and speakers. The dip in the 150hz area has been more or less eliminated.
And deeeam this speakers play loud!! Really punchy in the "upper midbass area". "slightly heavy metal" becomes heavy metal Placing them in-wall really helped (maybe 2kw subs helped a little as well). Playing +-100db at sweetspot now, and i cant hear any distortion at all.
(When looking at the graph, remember that my spl-meter's sensitivity drops slightly towards the upper octaves.)
And deeeam this speakers play loud!! Really punchy in the "upper midbass area". "slightly heavy metal" becomes heavy metal
Placing them in-wall really helped (maybe 2kw subs helped a little as well). Playing +-100db at sweetspot now, and i cant hear any distortion at all.(When looking at the graph, remember that my spl-meter's sensitivity drops slightly towards the upper octaves.)
Attachments
Thanks alot for the help tinitus! Playing with resisitors in series with paralell components seems to be very helpfull indeed, when designing a proper filter.
The distortion that bothered me earlier may have been some kind of peak of some sort in the mid's upper register, because it more or less disappeared in the final XO, where the mid got an extra cap in series(3rd order).
Could someone list different kinds of distortion that can occur in a badly designed construction\crossover?
The distortion that bothered me earlier may have been some kind of peak of some sort in the mid's upper register, because it more or less disappeared in the final XO, where the mid got an extra cap in series(3rd order).
Could someone list different kinds of distortion that can occur in a badly designed construction\crossover?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.