Just across the wiring terminals of the speaker. You can solder the C and the R on the L and connect the circuit on the output of the HP filter.
See below a crossover design and tranfer function The LCR create the small ~2 dB dip around 12 Khz. This crossover is from Jay and use 830883+810921. You might find this old thread interresting :
http://www.diyaudio.com/forums/multi-way/114409-my-experience-using-peerless-hds-810921-tweeter.html
Attachments
Last edited:
can you describe what the operation of the LCR and why it produces that little dip in the response.
Can you also describe the operation of the circuit parts that create the rise in response as frequency falls?
Can you also describe the operation of the circuit parts that create the rise in response as frequency falls?
OK...first of all I don't understand this 100 UF in that schematic...if you can clarify that would be awesome.
Also, I finally got a chance to ad the LCR. I decided to try that before modifying the xover further. Wow, what a difference! It's not quite there yet, but now I feel like I am getting somewhere.
I went with the LCR you recommended Pascal on one of the first pages of this thread. I used 1 uf, .1 mh and I didn't have the 6.5 ohm resistor so I used a 5.6 ohm. That was too much so I changed that to 18 ohms, and much better.
I also read that whole thread you linked to above. First of all, is the graph above his xover, since the image no longer exists that he originally posted?
If so, again what the heck is 100UH? I understand it is a type of inductor.
Oh awesome...thanks 5th...
I am going to listen them for a bit now, but I am really feeling encouraged. Thanks for all the help and suggestions, I'll keep everyone posted.
I am going to listen them for a bit now, but I am really feeling encouraged. Thanks for all the help and suggestions, I'll keep everyone posted.
I dont want to be a pain, but Ive heard many complaints about troels designs.
Maybe you will not be able to make it sound the way you want.
Looking at the high xo for the midbass and the 4k peak, maybe the problem is not the tweeter but the peak of the mid bass.
Also, its absolutely impossible that the audyn plus cap is what cause the harshness. Its a really good cap. good enough to make sure it dont adds harshness. Dont waste your time with the cap, the poblem is somewhere else.
What amp do you use, maybe your amp isnt well suited for the speakers?
Maybe you will not be able to make it sound the way you want.
Looking at the high xo for the midbass and the 4k peak, maybe the problem is not the tweeter but the peak of the mid bass.
Also, its absolutely impossible that the audyn plus cap is what cause the harshness. Its a really good cap. good enough to make sure it dont adds harshness. Dont waste your time with the cap, the poblem is somewhere else.
What amp do you use, maybe your amp isnt well suited for the speakers?
Last edited:
Agreed
Go back to the original design of troel. then pad down considerably the tweeter, like pad it too much, too really make sure the tweeter isnt too loud.
listen
if its still harsh, harshness comes from the mid, not the tweeter.
Well I added the LCR as mentioned above, today, and even my wife commented on how they "don't hurt your ears now", and that was at low level. (She even asked me to turn it up.) My wife complained they hurt her ears at much lower levels than what bothered me. ( I always attributed it to code for.."turn it down")
I think the LCR is the ticket! Hopefully I'll be able to listen at some decent levels tomorrow!
I already padded it down too much, as suggested...it's definitely the tweeter! But the LCR seems to be more progress than anything else I have done...I'm actually really pleased and excited! I didn't really think it could be fixed with the xover, in part because I din't think a crossover could do this, and of course because I don't fully understand what can be done with the xover...but this has been a huge learning experience for me. All I can say is...WOW!!!
I think the LCR is the ticket! Hopefully I'll be able to listen at some decent levels tomorrow!
I already padded it down too much, as suggested...it's definitely the tweeter! But the LCR seems to be more progress than anything else I have done...I'm actually really pleased and excited! I didn't really think it could be fixed with the xover, in part because I din't think a crossover could do this, and of course because I don't fully understand what can be done with the xover...but this has been a huge learning experience for me. All I can say is...WOW!!!
The xover does pretty much everything. 🙂
Unless the drivers are extremely poor quality, or inappropriately matched, you can make any speaker sound pleasant with the xover.
Unless the drivers are extremely poor quality, or inappropriately matched, you can make any speaker sound pleasant with the xover.
Hi Clausen,
Glad to read that things are better 😎
Listen like this for few days, after you can test increasing the value of the C : 1 mF -> 1.5 mF. Just solder another cap of 0.5 mF on // on the 1 mF.
Glad to read that things are better 😎
Listen like this for few days, after you can test increasing the value of the C : 1 mF -> 1.5 mF. Just solder another cap of 0.5 mF on // on the 1 mF.
I have the LCR temporarily twisted together at the tweeter terminal, behind the tweeter that way changes are super easy. The Audyn Standard 1.5 uf caps is less than $2, so I am going to just order those, and I need some other resistors to to try also. I am thinking about going back to the original L-Pad, and seeing if I can use the 6 or 6.5 ohm resistor on the LCR then.
I think I could live with them like they are, but if I can tweak them some more why the heck not!!???
Yeah I'm thinking about that, plus Parts Express has one for $60. (I have a wholesale account with them so I get a little discount on some things.)
I'll probably do that soon.
Glad everything is sounding better.
What wire are you using for the tweeter section? From binding posts to tweeter.
What wire are you using for the tweeter section? From binding posts to tweeter.
Like I said before, tilting the top end up/down and getting the tweeter level right. Not that the design couldn't be further improved by a lower xover point, but there's nothing that should stop this design from being made to sound decent.
The LCR in this case is working more as a notch rather than a tilt, but it accomplishes pretty much the same thing, especially as you probably can't hear much above where it's operating anyway.
The LCR in this case is working more as a notch rather than a tilt, but it accomplishes pretty much the same thing, especially as you probably can't hear much above where it's operating anyway.
What im trying to say is, dont drive blind 🙂
When you get the mic, measure as a whole speaker, then measure woofer with xover section and tweeter with xover section.
Then compare to what Troels posted on his website. (Sometimes i see there's alot of people disrespect the design and the designer without any data/proof) See if your speakers was built as it supposed to.
You can always post the graph here, im sure plenty of people here can interpret the graph for you, if you didnt understand yet 🙂
Again the scan/peerless HDS tweeters is one of the most popular tweeter to be used by DIY'ers and speaker manufacturer, they cant be all idiots 😛
Cheers
Henry
haha...true, very true...I didn't know it was that popular...funny because I have observed individuals can be convinced they are right about any given subject, in spite of what everyone else else thinks. I have often commented about such individuals, "When are they going to realize it is not everyone else who's wrong/has the problem, but them."
I think we can all trend that way, but let's hope most of us can realize when we are wrong, and not look like arrogant fools. Another reason I am glad to have this 90% (100%...possibly) ironed out.
I actually really like Troels designs, and he was very helpful. I have nothing but good things to say about him!!!! I thought it was simply the tweeter.
I am happy to report again I am amazed at how good they sound now...they sound so good it's like music in my head, instead of music I am listening too.
I still haven't had much chance to listen to them loud for any length of time, but I am pretty sure that won't be an issue anymore!
I'd still like to get a mic, just so I can measure them to see what I am listening to and do or don't like!
Hey Clausen, I read the first 3-4 pages of this thread and decided to put in my 2 cents. I'll probably go through the rest of it tomorrow.
I built the Nomex164 a couple of years ago. I used the advice of Pascal at the time to fine tune them a little, though I didn't tune the enclosure down to 35Hz. I think I settled on 42Hz. When I built them, I noticed that piano and female vocal could sometimes have too much bite, but slightly adjusting the lpad cured this.
These speakers never hurt my ears and sound technically very good, if not real special. I always thought the HDS in this application sounded not too bright, but sort of hard and not completely realistic. And not particularly shimmery.
I think some of these problems are not the tweeter and not insufficient attention paid to the woofers break up, but likely an issue with an uneven power response due to the highish crossover. I'm sure he choose 3.3K for a reason and usually drivers will naturally fall in to place when you're designing the xo.
Now, I've learned a lot since I built those and I've designed a number of speakers since, with increasing levels of success. I recently built a two way with these drivers (830875 and HDS) and it's a totally different ballgame. I used a stepped baffle to align centers and true 2nd order acoustic slopes to get as smooth a power response as possible. I had planned on a xo in the 2300Hz range, but ended up at 2600Hz, which is where everything just worked. The tweeter is shimmery and realistic and beautiful with no hint of strain. I used an elliptical notch similar to Troels' and the woofer contributes a lot to the frequency response above the stop band without any issues. These are simply completely different speakers than the Nomex and so much better. Excluding absolute dynamics and spl of course.
I'm not recommending that you build a two way out of them, your rooms too huge, but what I'm getting at is that power response may be your big issue, as others have mentioned. I simply have direct experience with this and these speakers.
For my advice to try to help, I would first of all say don't change the mid coil to 1.1 as TG suggested unless you correspondingly reduce the size of the bypass cap. I think around .75-.87uf would be your number. Otherwise they're going to ring like crazy and talk about making your ears bleed. The size of the coil and cap act as a notch for the woofers breakup and have to be just right to center on the peak, the resistor regulates how deep the notch is. Also, if you've attenuated the tweeter by going all the way from 1 to 3.3 with the series resistor and it's still too bright then the tweeter level is not your problem. As designed, the tweeter is already below the rest of the range. This should be sufficient. I would say placement, room treatment or also as others have stated, plug the ports and high pass the mains. I would also not recommend removing the .1mh coil in series with the tweeter as was suggested towards the beginning. This, bypassed by the 3.3 resistor acts like a quasi baffle step correction circuit to tame the tweeter rising response. You may get the same thing without it and simply using a lager series resistor, but maybe not. That is there specifically so they are not too bright.
I would also, not change the xo to 3rd order or whatever unless you can measure the results, and with your level of expertise, not even then. By going from 2nd to 3rd order, you are likely completely altering the phase and you'll start screwing up the imaging and coherence. No offence, and I think it's definitely worthwhile to experiment and learn, but you are unlikely to do better than Troels considering his knowledge, tools and experience. If you do want to make serious changes to the xo then I would buy some measurement gear and learn how to use it. Then shoot for a lower crossover and or increase the slope of the woofer and maybe tweeter and/or try a waveguide. Or all of these, who knows.
Anyways, good luck getting it how you want it, in the end, that's all that matters. I'm really not trying to dissuade you, it's just that some people here are suggesting that you add this and take out this without really looking into what that's going to do for you, taking all parts of the design into consideration. That means phase, power response, FR, power handling etc.
Evan
I built the Nomex164 a couple of years ago. I used the advice of Pascal at the time to fine tune them a little, though I didn't tune the enclosure down to 35Hz. I think I settled on 42Hz. When I built them, I noticed that piano and female vocal could sometimes have too much bite, but slightly adjusting the lpad cured this.
These speakers never hurt my ears and sound technically very good, if not real special. I always thought the HDS in this application sounded not too bright, but sort of hard and not completely realistic. And not particularly shimmery.
I think some of these problems are not the tweeter and not insufficient attention paid to the woofers break up, but likely an issue with an uneven power response due to the highish crossover. I'm sure he choose 3.3K for a reason and usually drivers will naturally fall in to place when you're designing the xo.
Now, I've learned a lot since I built those and I've designed a number of speakers since, with increasing levels of success. I recently built a two way with these drivers (830875 and HDS) and it's a totally different ballgame. I used a stepped baffle to align centers and true 2nd order acoustic slopes to get as smooth a power response as possible. I had planned on a xo in the 2300Hz range, but ended up at 2600Hz, which is where everything just worked. The tweeter is shimmery and realistic and beautiful with no hint of strain. I used an elliptical notch similar to Troels' and the woofer contributes a lot to the frequency response above the stop band without any issues. These are simply completely different speakers than the Nomex and so much better. Excluding absolute dynamics and spl of course.
I'm not recommending that you build a two way out of them, your rooms too huge, but what I'm getting at is that power response may be your big issue, as others have mentioned. I simply have direct experience with this and these speakers.
For my advice to try to help, I would first of all say don't change the mid coil to 1.1 as TG suggested unless you correspondingly reduce the size of the bypass cap. I think around .75-.87uf would be your number. Otherwise they're going to ring like crazy and talk about making your ears bleed. The size of the coil and cap act as a notch for the woofers breakup and have to be just right to center on the peak, the resistor regulates how deep the notch is. Also, if you've attenuated the tweeter by going all the way from 1 to 3.3 with the series resistor and it's still too bright then the tweeter level is not your problem. As designed, the tweeter is already below the rest of the range. This should be sufficient. I would say placement, room treatment or also as others have stated, plug the ports and high pass the mains. I would also not recommend removing the .1mh coil in series with the tweeter as was suggested towards the beginning. This, bypassed by the 3.3 resistor acts like a quasi baffle step correction circuit to tame the tweeter rising response. You may get the same thing without it and simply using a lager series resistor, but maybe not. That is there specifically so they are not too bright.
I would also, not change the xo to 3rd order or whatever unless you can measure the results, and with your level of expertise, not even then. By going from 2nd to 3rd order, you are likely completely altering the phase and you'll start screwing up the imaging and coherence. No offence, and I think it's definitely worthwhile to experiment and learn, but you are unlikely to do better than Troels considering his knowledge, tools and experience. If you do want to make serious changes to the xo then I would buy some measurement gear and learn how to use it. Then shoot for a lower crossover and or increase the slope of the woofer and maybe tweeter and/or try a waveguide. Or all of these, who knows.
Anyways, good luck getting it how you want it, in the end, that's all that matters. I'm really not trying to dissuade you, it's just that some people here are suggesting that you add this and take out this without really looking into what that's going to do for you, taking all parts of the design into consideration. That means phase, power response, FR, power handling etc.
Evan
Even though I wrote a novel, I forgot to mention something.... If you do change the mid coil to a larger value (plus bypass cap value) then you will probably want to try increasing the value of the tweeters cap. You can try paralleling small caps with the 4.7uf to get the value that works for you. maybe start with one and go from there. This will get you a lower crossover without creating a hole around the crossover. But, maybe a dip in the xo region would help? This should help the woofer beaming problem and maybe smooth out your power response some.
Also, if I'm not mistaken, the LCR network you tried is targeting the area around 12K? Correct me if I'm wrong. In my opinion, this area is unlikely to be causing your problem; you may not even be able to hear it. If you want to lower this region then you could try increasing the value of the resistor across the series coil on the tweeter (.1mh). This could raise the area before the top octave if you take it too far though.
And lastly, have you tried measuring the impedance of them to make sure they're wired properly and you don't have a damaged driver or crazy box resonance? You can make a jig for a couple of bucks and use REW, it's free. It's really not too difficult.
Also, if I'm not mistaken, the LCR network you tried is targeting the area around 12K? Correct me if I'm wrong. In my opinion, this area is unlikely to be causing your problem; you may not even be able to hear it. If you want to lower this region then you could try increasing the value of the resistor across the series coil on the tweeter (.1mh). This could raise the area before the top octave if you take it too far though.
And lastly, have you tried measuring the impedance of them to make sure they're wired properly and you don't have a damaged driver or crazy box resonance? You can make a jig for a couple of bucks and use REW, it's free. It's really not too difficult.