Site Links
Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Quick Links
Categories
In this Discussion
Please review the site Rules, Terms of Service, and Privacy Policy at your convenience. Rules, TOS, Privacy
Get familiar with the reaction system: Introducing the Reaction System
Is there a point where padding tweeter or midrange is too much?
I realize the theoretical answer is probably "no", but are there practical limitations to padding a tweeter or midrange a LOT?
Specifically, I have an idea for a 4-way for InDIYana 2026 using existing drivers. But the upper midrange is a 2-1/2 inch dome midrange with a sensitivity of 95dB and a tweeter with sensitivity of 91dB. But the lower sensitivity of the lower midrange and woofers probably leaves me at a system sensitivity of 82-84dB after baffle step losses. Is padding the 95dB midrange down to 82-84dB going to present me any significant problems?
Comments
No sir, should not be an issue.
Agreed. I've padded 15dB before.
InDIYana Event Website
The driver will have lower distortion, because you're asking for less output.
Agreed!
InDIYana Event Website
Yep, agreed. I've never had an issue with padding down higher sensitivity drivers to match woofers that carry most if not all the baffle step loss.
FYI the more you pad the more heat you will have.
Valid point.
This is why using sometimes 3 resistors in the circuit to accomplish the padding may be a good idea. Other things I personally tend to avoid if possible, are;
These are just things that I do, not that they are the only way or best way to do things.
InDIYana Event Website
Thanks, good point on this. I'll make sure to take this into account, maybe I'll get some of those 25w Dayton resistors. In this case, the dome midrange is probably going to handle something like 800 to 3200Hz, so I'll check the power dissipation in VCad when I'm at that stage.
I've found the more attenuation required makes it more sensitive to small resistance changes. It was annoying enough I just slapped an lpad pot on the tweeter so I could measure and make small changes to match L/R.
I think I have a 2 ohm/10W shunting the tweeters in my TCP/XT design. But I also have a 7.5 ohm in series right before it. I'll have to double check, but I doubt it'll ever see that much power. I may have to parallel two 4 ohms, just in case.
X-Sim can track power dissipation in crossover components. Argon is probably the most heavily padded design I have ever done (single 7" Anarchy woofer, a small mid and the very sensitive Peerless BC25SC0604 tweeter). The diagram below shows the simulation of the design with power dissipation of all the resistors with 10 watts of input power applied.
The big series resistors R2 and R4 are taking a pretty good beating, although none are failing yet.
Sehlin Sound Solutions
I've been pondering this. They say Pink Noise emulates the spectral power density seen in most music. Knowing that, I doubt I'll ever send more than a few watts above 2K to any home speaker I build.
But I can't vouch for other people who might want to play Barbie Girl at concert levels.
"They" say pink noise resembles power spectrum of real audio, but it still doesn't, but it's a lot better than white spectrum. There's more to it than slope of FR though, crest factor / duty cycle is also quite important, in which pink noise still contains a lot more high frequency energy than real audio.
But you don't have to believe "they" or "me", you can evaluate this yourself using REW. Start by opening the generator, select pink noise and save to file. Then go to File -> Import -> Audio Data, select the pink noise file. Repeat audio data import for a few of your favourite audio tracks, it will open only 100 seconds of audio at a time so keep that in mind. SPL chart shows power spectrum of the audio, I recommend smoothing at least 1/48 oct otherwise it will be a hairy mess. For detailed analysis at any time, go to the Spectrogram tab - very useful tool for audio analysis.
Xsim power dissipation chart is absurd for real audio consideration, it assumes constant power flat spectrum sine wave at all frequencies for whatever voltage you select. It will lead you to purchase 25W parts where 2W will do perfectly well. I'd recommend looking at VituixCAD power dissipation instead. It's default selection of "pink above 2kHz" is a good "worst case" real audio scenario for component sizing, and some other options for evaluating power delivery. Very useful.
Handy to Know ^.
Thanks!
I think both analyses have merit. If you are confident your speakers will only be used for music, then the VituixCAD analysis is appropriate. If your speakers might be used for other purposes ranging from home theater, gaming, production of content including special effects, emergency or distress signals, etc .. then you can have concentrated output at frequencies that the X-Sim analysis highlights as potential issues in my example.
Sehlin Sound Solutions
OK, I have my flame proof suit on and I'm ready...
An L pad is just a voltage divider. Capacitors are more accurate and have less variance in lower voltage situations.An L pad out in front might make more heat but should yeild better sound quality.
Please, prove me wrong with data. Thanks
I prefer to have the series R out front because it affects spectral tilt less, and reduces impedance phase angle. The shunt R should be after xover to prevent worse heat issues.
More or less, I agree with you.
InDIYana Event Website
Ben, how you have stated is the same way that almost everyone does it. My point is that reducing voltage before capacitors (and inductors) should result in better sound quality.
I always try the series R in a few places, but just to see how it shapes the response. I hadn't considered changes to the phase angle. I'll keep that in mind on my next design.
I just put padding resistors where they work best. Sometime I have only one somewhere. Sometimes I have 3; one in series before the filter, one in series right after the filter, and one right across the driver's terminals. There is no one cookie cutter way to do it. Unless, of course, every project you build has a Zobel on the woofer with a 2nd order LP, and a 3rd order HP followed by an L-pad lol!