I did my homework shopping for a mid priced tweeter and ended up with a pair of the Vifa D27TG-35-06 on my workbench. I come here looking for suggestion and support. At its price point, I do not expect it to be perfect, but it meets my budget objective and seems to be a well respected performer.
I am in the process of trying to come up with a suitable crossover.
My reference point is the Larry Hughes bookshelf DIY using the Audax HD17B25H, Audax HD12x9D25 and Falcon 21B crossover. (HiFi Answers, July 1978) I am aware of the claims of super tweeter such as the Scanspeak Illuminator D3004/6640, but that's not the subject here.
keilau
And what you have there (D27TG-35-06), will do what you want. LR2.
May I suggest looking at Linkwitz Lab on driver distortion test.
And what you have there (D27TG-35-06), will do what you want. LR2.
May I suggest looking at Linkwitz Lab on driver distortion test.
When I had the D27TG-35, very long time ago, I suspected that something wrong with the waveguide. I was about to modify the default waveguide of the faceplate, but then decided to sell it.
Now that I have seen the ATC SCM20 with modified waveguide, I think my intuition could be correct. And it is not difficult to copy the waveguide of the tweeter in ATC. Your homework is of course to find out if these tweeters only differs on their waveguides. If so, you could have a very good tweeter there.
The waveguide only affect the frequency response slightly, but the perceived audible difference can be huge!
Which design method are you using:
(1) Full measurement method
(2) FRD consortium method (FRD/ZMA can be either measured or predicted)
(3) Design by ears method (you need to have plenty of coils and caps)
Series crossover? It will work for the D27TG, how about the woofer? It must have smooth roll-off response, which is rare with budget woofer.
If the design method is the FRD consortium, with no access to measuring tools, the best imo is to search for proven design using any of the drivers then work it out in a software.
FRD/ZMA method is not accurate, or uncertain soundwise (you cannot predict the sound from just a FR chart). By copying proven design most of this uncertainty will be removed.
The HF filter on the ATC SCM20 can be used for the base filter for the D27.
After I found out that ATC used this relatively cheap tweeter and Quested an equally cheap but now deleted Morel I wondered why these respected manufacturers would use them over more expensive units which would easily fit within their budgets.
What I found is that neither this Vifa nor the Morel were the last word in low distortion.
However both had an exceptionally clean CSD compared to all but the most ludicrously priced tweeters and both can take a lot of 'abuse': Morel uses Dynaudios tag line of 1kW peaks (they also seem to use a lot of Dyns technology) while the Vifa is rated 100Wrms via a 700Hz 12dB high pass.
IME both features help these to outshine much of the competition when it comes to reproducing accurate transients.
I can't help but thinking that the Vifa could do really well in a proper waveguide.
Think I'll try that some time since it should bolt straight onto the Visaton 148mm WG and they cost only €9ea.
What I found is that neither this Vifa nor the Morel were the last word in low distortion.
However both had an exceptionally clean CSD compared to all but the most ludicrously priced tweeters and both can take a lot of 'abuse': Morel uses Dynaudios tag line of 1kW peaks (they also seem to use a lot of Dyns technology) while the Vifa is rated 100Wrms via a 700Hz 12dB high pass.
IME both features help these to outshine much of the competition when it comes to reproducing accurate transients.
I can't help but thinking that the Vifa could do really well in a proper waveguide.
Think I'll try that some time since it should bolt straight onto the Visaton 148mm WG and they cost only €9ea.
I started that thread on Audax. When it led to a dead end. I started this one on the Vifa.
So your FRD/ZMA have:
(1) already had phase information
(2) been produced based on a known box
Do you have your FRD/ZMA files somewhere?
What is baffle width or baffle step frequency? Easier is not to include the BSC effect on the FRD, so you can adjust based on your needs and situation.
So you know the limitation of this method. I mean, if you don't measure your own driver's FRD/ZMA, and if it is OFF spec, then it means that the diffraction effect inclusion for example, has no meaning.
And then there are more to it than just doing this design method. You can make many prototype that seems acceptable but not so when you have built and listened to it. And copying "proven" design will save you from the uncertainty. But of course, the design you copy must not be a design based on the same non-measurement method 😀
So now you have ATC SCM20. Then you search for other designs (exclude designs from unknown designers), I think many has used these drivers.
Then with the FRD/ZMA availability, and an access to design examples using the drivers, you can design your own speaker (or anyone can easily help you by proposing one) with much better chance for success (soundwise). If you know why, this copying idea removes many uncertainty in your design variables.
I saw that you use 4th order filter. It is very dangerous to use such high order filter (especially if not the text-book acoustical filters) when you cannot hear the result.
So it give you a lot of satisfaction to put down others? You did not even read this thread before making your comments. It is pathetic.
The difference between you and others is that the others made suggestions and I tried to follow up. You contribute nothing to this thread.
This is a DIY forum, not a professional audio society. If we are not up to your standard, please, go elsewhere. BTW, you have no credential by no showing a single example of your work.
It's a suggestion on how IMO people can help you better, and for you yourself to be able to make a better speaker. You post your FRD/ZMA files, find the projects using the drivers (professional ones because you will rely on their measurements and judgements), use those as reference.
I am far ahead of your suggestions. (And I am almost certain you have not read the posts I made on this forum.) If you want to have any credibility, please, show the result of your "make a better speaker". An example of your process and the measured result of your project will be very helpful. Otherwise, you gave me the impression of someone who likes to talk about things you know nothing about.
If making a hobbyist speaker is too difficult for you to handle, that does not mean no one else should try it. I and others migrate to this stage after building a few of the commercial kits and magazine published designs. It is the natural next step. We are here to help each other out. If you have not noticed, this is the diyaudio.com forum.
You still have not answer my first question on what was "so awful" about the charts?
Thank you for your comment. The Vifa is my mid priced choice to replace the older Audax HD12x9D25. It is a budget option.
Any comment, critique you may have on the crossover approach I posted in #28 will be welcome and greatly appreciated. I believe the crossover frequency (~ 2KHz) falls into the range you recommended (in your sticky posts). What other tweaks that you can suggest for me to try?
Good then. I thought I knew something you didn't but it seems you know a lot.
Do you mean ALL your posts? Then the answer is NO. If you were 10 years longer in the forum, you bet I'll recognize your reputation hehe 😀
No, I don't need credibility at all. But I was trying to show you... but if it was not good enough for you, it's okay.
No, I have no problem with that. I was trying to reply on System7 post.
Nah... forget it. You want to know the answer but you don't want to listen to the answer. Why not build the crossover and listen for yourself. There's nothing you can learn from me. You know to much already.
Keilau, I went through a stage back then of a couple of things.. low crossover points, and minimalistic crossovers. On the first point I still believe this is a good thing but too far won't cut it with most domes, especially with a simple filter. I could cross these first order at 1500Hz but thermal issues get in the way.
On the second point when I said first order filters I would always use a notch filter on the tweeter peak (obsession is one thing but it's not always realistic). Still I think I'd use at least second order If I were doing it today.
If I remember well enough 2kHz second order was more reasonable but higher might offer a minor improvement in thermal issues (ie the sound at all levels, as well as overheating with a change of tone after some louder listening.)
The down side to crossing higher though is twofold. A crossover's issues become more critical as we are more sensitive at 3-4kHz as well as the wavelength being shorter.. and it also means you'll require a smaller mid to cross it to.
It's all a compromise.
On the second point when I said first order filters I would always use a notch filter on the tweeter peak (obsession is one thing but it's not always realistic). Still I think I'd use at least second order If I were doing it today.
If I remember well enough 2kHz second order was more reasonable but higher might offer a minor improvement in thermal issues (ie the sound at all levels, as well as overheating with a change of tone after some louder listening.)
The down side to crossing higher though is twofold. A crossover's issues become more critical as we are more sensitive at 3-4kHz as well as the wavelength being shorter.. and it also means you'll require a smaller mid to cross it to.
It's all a compromise.