Combined with good measurements and an understanding of what you are doing, they will get you a long way.
Well Im trying to understand the crossovers, but still I dont think I will be able to get it done without help...
I hope someone here i willing to help me, building the right cross over.
If the midbass goes nice up to 5kHz and the tweeter has a FS of 700hz, would not just a simple first order BW be sufficient, just one inductor and one capacitor? Or would still -3db be too much for a safe crossing of the tweeter?
I hope someone here i willing to help me, building the right cross over.
If the midbass goes nice up to 5kHz and the tweeter has a FS of 700hz, would not just a simple first order BW be sufficient, just one inductor and one capacitor? Or would still -3db be too much for a safe crossing of the tweeter?
Imagios, I just went through the same thing. Steep learning curve for me, but reading the page linked above by giralfino and just about anything else I could find on crossover design and the use of PCD I believe I have gotten somewhere reasonably close to a decent crossover ( we'll see ! ).
This page
https://sites.google.com/site/undefinition/simulated-measurements
describes some important steps prior to loading .frd and .zma files into Passive Crossover Designer. Not everything there is relevant to you such as the starting part about tracing graphs, but the part about using Frequency Response Modeller should be.
Another thing off the top of my head, did you enter in the relative acoustic centre offset between the woofer and tweeter in PCD ? It can have a significant impact on the crossover particularly with regards to phase and the way the two drivers combine.
I'm not comfortable giving advice as I'm just a beginner too, but I hope some of that at least leads you in the right direction.
This page
https://sites.google.com/site/undefinition/simulated-measurements
describes some important steps prior to loading .frd and .zma files into Passive Crossover Designer. Not everything there is relevant to you such as the starting part about tracing graphs, but the part about using Frequency Response Modeller should be.
Another thing off the top of my head, did you enter in the relative acoustic centre offset between the woofer and tweeter in PCD ? It can have a significant impact on the crossover particularly with regards to phase and the way the two drivers combine.
I'm not comfortable giving advice as I'm just a beginner too, but I hope some of that at least leads you in the right direction.
Thanks for the advice, the crossover is really the most complex part...at least for me.
I still have a lot of issues with all the SW from the net always something seems not to work.
PCD is crashing, FRD as well, might be my windows 7 and office 2010
Never the less I played around with the most easy setting, here is the result of just two first order crossovers, giving a +/- 1 db on the whole spectrum
(Still the main question and the big one is, would the tweeter survive just a first order crossover? Although the cross over freq. in this case is more than 10x his Fs)
I still have a lot of issues with all the SW from the net always something seems not to work.
PCD is crashing, FRD as well, might be my windows 7 and office 2010
Never the less I played around with the most easy setting, here is the result of just two first order crossovers, giving a +/- 1 db on the whole spectrum
(Still the main question and the big one is, would the tweeter survive just a first order crossover? Although the cross over freq. in this case is more than 10x his Fs)
adding just one point, since my amp is 8 and 4 Ohms, I consider to wire in series, since the calculation of the lowest impendancies of both drivers in parallel went on the edge for the amp.
According to the simple simulations it really seems I could do it with only one cap and one inductor. (sure a Zobel would smooth it up, but should I tomorrow get the cap and inductor, that might be the right start
According to the simple simulations it really seems I could do it with only one cap and one inductor. (sure a Zobel would smooth it up, but should I tomorrow get the cap and inductor, that might be the right start
Hi Imagios,
I don't know if the tweeter will survive, but I know it will make large excursions at 700 Hz and reach the limit at moderate listening levels. Therefore I would not go for a first order and choose a classic 2nd order filter at 2.2 kHz.
The peak down in your first diagram occurs, because the sound of woofer and tweeter have opposite phase and cancel. The usual fix is to reverse the tweeter polarity.
I don't think the simulations are right. The drivers seem to have a flat frequency response till infinity. You will get better results, if you import a .frd data file!
Good luck
I don't know if the tweeter will survive, but I know it will make large excursions at 700 Hz and reach the limit at moderate listening levels. Therefore I would not go for a first order and choose a classic 2nd order filter at 2.2 kHz.
The peak down in your first diagram occurs, because the sound of woofer and tweeter have opposite phase and cancel. The usual fix is to reverse the tweeter polarity.
I don't think the simulations are right. The drivers seem to have a flat frequency response till infinity. You will get better results, if you import a .frd data file!
Good luck
OK in the mean time I just already put the box together for testing, fortunately the tweeter survived
series wiring did not gave the good sound as parallel so the box is now running with a first order BW cross over, that means inside the box ist just the coil and cap, and well even without stuffing it sound quite good.
I have no measuring instrument to measure the responce, I did not closed the box, the back site is not glued so I can easily make changes, the cross over is definately not final, but, I must say Im quite impressed by the sound they give...
I dont even now have the bassreflex vent as improvisation the tooilet paper core is inside
any way my first self made self designed speaker (with a lot of help from all of you )I will keep you updated, and ask further with the crossover design that would fit...
0k here is the update,
the speakers should be already settled, so how do they play now, honestly flat, the treble is just too aggressive, it has some fatigue to listen to them. There one certain sound producing a bad distortion, its very rare I dont know the frequency but its there.
The trebble need to go about 3db lower, I played on the EQ even with 6db less and still they where very present. Till the bassreflex arrives, I just conected them to my PC sub. and now with the sub, it has too much bass, even when I set the bass at the sub to minimum... but for PC its OK.
So next step will really be the redesign of the crossover.
Here a pic. how it look right now
the speakers should be already settled, so how do they play now, honestly flat, the treble is just too aggressive, it has some fatigue to listen to them. There one certain sound producing a bad distortion, its very rare I dont know the frequency but its there.
The trebble need to go about 3db lower, I played on the EQ even with 6db less and still they where very present. Till the bassreflex arrives, I just conected them to my PC sub. and now with the sub, it has too much bass, even when I set the bass at the sub to minimum... but for PC its OK.
So next step will really be the redesign of the crossover.
Here a pic. how it look right now
Sorry I can not help with .frd files. But I have a proposal for the crossover. It may not be the ultimate solution, because the driver data are somewhat doubtful, particularly concerning the tweeter, but better than nothing...
My simulation is based on the data of Zaph (woofer) and the datasheet (tweeter). As I have heard of Tang Band tweeters having spl differences of 4 dB, you will probably have to adjust the resistor in the tweeter path.
My simulation is based on the data of Zaph (woofer) and the datasheet (tweeter). As I have heard of Tang Band tweeters having spl differences of 4 dB, you will probably have to adjust the resistor in the tweeter path.
Thanks a lot for that, I have to get your program
What about the high peak at the end, do I understand it right to adjust that with a resitor?
Btw. the bassreflex arraived today, just waiting till the glue gets dry
It will take me some time before I finalize it, since I would like to have the final crossover on a custom pcb. but from what your simulation shows it seems decent, I have just only expected a better lower freq. performace, as from WinISD simulations.
I will make the next update after playing the box with the vent
I could let you have the frd I created by tracing Zaph's measurements with SplTrace if you're interested ? It should be quite accurate, I was careful and checked it tallied with the original. I also have the zma for the W5 704D if you need it. Sorry, I can't help with the tweeter, I'm using a different one.
It'd be the "raw" frd I'd provide rather than the one's I'm actually using in PCD. Those are the raw driver responses modified with Response Modeler to take into account the effects of my specific baffle and enclosure (this is pretty much a necessity for predicting a decent crossover with PCD. The link I gave earlier in the thread explains how to do this ).
Thanks I will run some new simulations tomorrow, in the mean while the bassreflex is in place, it helped a lot, I have downloaded a freq. generator and found some unpleasant resonances in higher volumes, I will continue tomorrow to analyse it, but clear is the first order BW just by its self is not the right solution.
Hi Imagios,
The resistor in the tweeter section determines the general level of the tweeter. If the tweeter is too loud, increase the resistor to bring it in balance with the woofer.
The peak at 23 kHz is another issue. You will have to decide yourself, whether to apply a notch filter or not. But don't expect the peak to be exactly as in the datasheet, as it will probably be different in your box.
The low frequency simulation of WinISD is a little bit optimistic, as it does not consider the influence of baffle step and serial inductors (of driver and crossover). This leads to different results like 50Hz and 58Hz. Further I was not sure about your volume and vent tuning and assumed 12.78 litres with a port of 45 x 120 mm. But don't be disappointed, the result is ok in relation to a volume of less than 13 litres!
The resistor in the tweeter section determines the general level of the tweeter. If the tweeter is too loud, increase the resistor to bring it in balance with the woofer.
The peak at 23 kHz is another issue. You will have to decide yourself, whether to apply a notch filter or not. But don't expect the peak to be exactly as in the datasheet, as it will probably be different in your box.
The low frequency simulation of WinISD is a little bit optimistic, as it does not consider the influence of baffle step and serial inductors (of driver and crossover). This leads to different results like 50Hz and 58Hz. Further I was not sure about your volume and vent tuning and assumed 12.78 litres with a port of 45 x 120 mm. But don't be disappointed, the result is ok in relation to a volume of less than 13 litres!
I have run yesterday some test with the freq. generator, I believe I don´t have to be afraid of the end high pear, since just to hear a sound of 15 kHz in a not absolutely quiet room was at least for my ears not possible, I dont thing I could hear a 18kHz sound, at least not the generated at decent levels.
After putting the real frl file of the woofer into the system, that quite changed the perspective, I will try to get the tweeter as well to show a complete picture, but just now I know the cap is too high.
I will keep you updated
OK I did not get the frd file for the tweeter, but I downloadet the one for the Tang Band 25-1166S.frd from rbjaudio.com, according to the datasheet both thweeters are alsmost the same, except mine has this peak at 20kHz
never the less here is the horible result:
I will play around with the crossover to make that more smooth
never the less here is the horible result:
I will play around with the crossover to make that more smooth
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.