It appears to me that Peter's real world experiment and analysis in #51 clearly shows that the effect is so small as to not warrant the effort or cost. Can you please detail why you think this is not the case?
Certainly in my case my tweeter and SS amp has similar charactistics to thoese he used.
Many suggest that an actively driven tweeter/treble be fed THROUGH a capacitor. This gives long term DC offset protection.
That capacitor can be very large and then ignored in all the filtering equations. Or you can make it part of the crossover using the 1pole roll-off as part of your overall Filtering/EQ.
This works well with a 2way, where the cap is can be quite small for an F-3dB ~3kHz.
from another reply:
You can apply the same method to protect the mid range driver in a 3way.
All that's different is that the capacitor will be ~10times higher value, if the F-3dB ~300Hz or 20times larger for ~150Hz.
Then you ONLY need DC detection and isolation for the bass only driver.
You could add a flashing LED (+Zener + resistor) across the DC protecting caps in the mid and treble to indicate excess DC voltage. Poke these through the front panel.
Last edited:
True! It's a requirement for safe operation of tweeters in active systems.
The Cap should be big enough for the amplifier to dampen the reverse electro-motoric energy from the speaker as well as it's resonant frequency. Phase shift is also a problem using a Cap, so the bandwidth ought to be at least extended by 1 full decade beneath the HP frequency. Otherwise you're in for some compensation like in passive XO's.
Apart from that aspect, sometimes the characteristics of the tweeter require some compensation from perspectives of measurement or taste.
Last edited:
I used cap for -3dB point at least one octave below the tweeter Fs actually, so between three and four octaves below crossover frequency. In my case, it is about 200Hz for the combination I use.
The cap will form a peaking filter with the inductive impedance of the tweeter coil at Fs. This is a neat trick used in high Q systems to increase woofer output at low frequencies, but would be a problem when used with a tweeter. Hence I go for caps that are quite a bit larger than a one-octave gap from Fc.
I noticed the "1 decade"-rule (big word...) years ago experimenting on the lower cut-off frequency of amps. If the cut-off frequency was higher than the "-1 decade" beneath the desired minimum frequency, bass response (until lower-mid!) proved to be audibly compromised.
Or if your system is only mid budget such as mine, with mid price Vifa tweeters, just do what I did....I just picked up a couple of spares on special a couple of years ago. Still sitting in a cupboard. I know it sounds like a waste, but this saved the purchace and installation of protective caps. So I don't need to worry about protection until I have a problem. So far....so good....after about 8 years of daily use. ( I know....famous last words!)
Last edited:
No details, you'll have to figure out the why of the statement.
Scientifically and logically .. with respect to observation, hypothesis and expected results and outcome.. it is all as clear as can be.
That is as far as I will go. I'm not trying to convince anyone of anything.
have a good one.
I strongly agree, because turn on/off thumps can destroy a tweeter. IMO, deploying a cap as the first pole of an otherwise active tweeter highpass is prudent design.
I am a follower of empirical reality.
I agree, let's.
However, if you actually understood the reality-based position you'd know that hand waving assertions about measurements is just that. Provide a cite to a controlled listening test published in a peer reviewed scholarly journal that backs up your position, and it wouldn't be so lightly dismissed as mere delusion in the service of selling your audio parts.
Isn't it funny how two people with pecuniary interest in voodoo mythology think so...
Whatever happened to "Let's skip the obvious vendors of snake oil," supra?
Dr. Rich's "reality" is somewhat different than the one I live in.
The trouble with this sort of position is that "objective" measurement gets one just so far. But this discussion has been had, baked, flipped, and basted. No point in rehashing. I am 100% FOR objective measurement, but the empirical reality is that it may be necessary, but it is in and of itself not sufficient. Gets you quite far, but not all the way.
Ur entitled to believe what he says.
I addressed your "controlled listening test" canard earlier.
Once more for completeness. Show me the "controlled listening test" where A) the test has been replicated by other researchers B) confounding factors have been properly controlled C) the "statistically valid results" actually can be shown to have been testing the thing that is nominally being tested (think about this a while) and finally D) the "statistically valid results" were able to be generalized and they were valid for more than the specific test that took place.
Get back to me on this?
Which is funny.
Except for the fact that this is a direct insult.
Untrue as well.
(trying to be nice about that).
Bring something solid to the discussion, back up your positions, and please stop the insults.
If you think that every amp and cable sounds the same, assuming it has enough power and low enough distortion, perhaps you ought to go to a different forum for discussion - or start a thread about this idea and Dr. Rich. Have at it.
The only problem with the idea that the Zobel on the tweeter makes no difference based on the calculation(s) presented earlier is that in some cases a Zobel on the tweeter clearly makes it "better" to listen to, and in some cases it makes the tweeter clearly sound "less better". In which case (after fiddling with values, to be certain) it gets yanked.
Of course your experiences may differ from mine.
_-_-bear
Of course your experiences may differ from mine.
_-_-bear
It should be noted that if these tests are ABX they are unable to prove that 2 devices are the same...
dave
Yes, all literally ANY test could find is that there is a null result; that is, no difference found, but not "proven" to be the same.
Last edited:
One octave below Fs and 1 decade below nominal pass band amount to much the same thing. In practice this means about 150uF, which is a bit large for a non-electrolytic
The problem here is the mistaken impression that a Zobel confers some goodness to the circuit around it.
It does not.
A Zobel is an impedance altering device that may or may not create a beneficial response shift by flatening some region of impedance. It alters the load seen by the previous series arm of the network and from that, may or may not improve response. Anyone that thinks "Zobels are always a good thing" has lost sight of the primary job of the crossover network: to shape system frequency response to give the sound we want.
For the most part, inductance flattening Zobels are unnecessary. Desired passive network shapes can generally be had without them by optimizing other available network elements. If they are a necessary part of the topology for fine tuning high frequency tilt, then the best result is from the values that give the target response, not the values that give some arbitrary section impedance.
Of course the above assumes a passive network where multiple dB shifts can be had when a Zobel is used. When directly amplifier driven then we are changing the shunt load hanging across a very low source impedance. In such cases we would expect to see no response change.
In case the amp, as a source, isn't so low in impedance, then minor response effects be evident. What makes anyone think that response will be (generally? always?) more desirable with a Zobel than without? It is pot luck as to whether response is better or worse. Plus, don't we have to measure the driver as driven via the amp to design the crossover? If there is a mild interaction between amp and driver it gets rolled into the raw measured response. It will be fixed by proper network design.
For those that think that there is some distortion impact from the inductance rise, why are you ignoring the motional impedance peak? This is significantly higher with a considerably wider phase angle swing.
David
Dave, I don't think anything I said implies or says explicitly that there is anything intrinsically of "goodness" to the Zobel on a driver. Did I?
What I said is that sometimes they seem to make the tweeter sound "better" and sometimes "worse". Usually I have not personally had the experience of them doing nothing at all, although I suppose that is possible.
_-_-
What I said is that sometimes they seem to make the tweeter sound "better" and sometimes "worse". Usually I have not personally had the experience of them doing nothing at all, although I suppose that is possible.
_-_-
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.