TIM reduction for tweeters

Status
Not open for further replies.
That will depend on how much TIM your power amp generates.

... And as I stated I dont know how much influence it has on the sound. Just thought it is an interesting subject.

I did use the filter in som serial 6dB crossover designs, but I never made listening tests before and after applying the filter.
 
The sim is interesting. It certainly reduces supersonic impedance. Which amps may like. Possibly to 8 ohms.

And gives about 2dB reduction at HF.

The LED must be dubious, but the rest is OK. Works a bit like a Zobel or even a midpass filter.
 

Attachments

  • HF Filter.PNG
    HF Filter.PNG
    20 KB · Views: 132
IM generates series of signals above and below the frequencies being inter-modulated.

So nothing happens to lower frequencies, therefore no TIM will be reduced.
Most of the combination frequencies may be in the mid and tweeter covered frequencies, so no reduction of TIM here either.
You do not hear what you do not hear.
 

Attachments

  • IM.jpg
    IM.jpg
    30.9 KB · Views: 118
Actually it is fair to assume that TIM will be distributed around a mid frequency band, so that most of the TIM will be below 20 KHz.

That is the kind of friendly and factual response that we can learn from, not just that what we assume is nonsense. Nobody can learn anything from that.
 
Hi adason,
thank you for your post.

There is of course no way to dampen the unwanted signals below 20 KHz.
Arthur,

Actually there is.
The Jensen Hypex D-30 driver used in the old Leslies has no phase plug in front of the diaphragm, so the enclosed space between it and the thread on 5/8" exit is an acoustical band pass filter, resulting in a roll off starting at 2500 Hz, with very little output capability (regardless of electrical input) above 12.5kHz. Because of the acoustical band pass, it also had some of the lowest HF distortion of all the drivers I have measured.

Left you a PM, I have one of these drivers for sale if interested.

Art
 
It probably doesn't matter what the exact effect is called. But there IS a very distinct effect with HF impedance on Class B amplifiers. You can hear it.

At HF, amplifiers run out of open loop gain and current very dramatically. It can fall from 500,000 to 50 or so. Hence the whole feedback game doesn't work very predictably.

535870d1457287689-tim-reduction-tweeters-tim-reduction.jpg


Your circuit is not dissimilar to the sort of thing you fit at the output of a Class B amplifier for stability reasons. See below.

An amp is generally tested into an 8 ohm dummy load, and probably optimised for that too, since THD figures affect sales. But lighter loads, and remember a tweeter is nearer open circuit or inductive at HF has a different response. The 0.1uF/10R Zobel is fitted at the output of the amplifier here. You can see how no load tends towards resonance or instability, but the Zobel saves it from oscillation with no speaker attached.

520384d1450681833-crossover-sounds-better-closed-loop-response.png


520385d1450681833-crossover-sounds-better-shunt-68r-wirewound.jpg


Here's a couple of other ways of getting a lower supersonic impedance with a Zobel or a 22R. I use both. They sound better IMO.
 

Attachments

  • Class B Audio Amp.PNG
    Class B Audio Amp.PNG
    76.5 KB · Views: 60
  • Supersonic Impedance.PNG
    Supersonic Impedance.PNG
    12.9 KB · Views: 64
Last edited:
IMG]http://www.diyaudio.com/forums/attachments/multi-way/520385d1450681833-crossover-sounds-better-shunt-68r-wirewound.jpg[/IMG]

Here's a couple of other ways of getting a lower supersonic impedance with a Zobel or a 22R. I use both. They sound better IMO.

Of couse it sounds better because with the shunt resistor the driver got less power because it's split up between the driver and the resistance. It even lowers the higher frequencies because the impedance of the driver is rising with the frequency because of the inductance of the coil and the higher the frequency the more power goes through the resistor instead of the voicecoil. That's simple equalizing but doesn't reduce the TIM itself.
 
System7,
that is probably correct, but I am sure the critics would state that this is not a TIM reduction issue, and they would be correct.

Adason shortly, friendly and precisely presented an argument that convinced me that this circuit does not reduce TIM to a large degree.

I though still think it will help me in my Leslie application, but not because of TIM issues.

Here we say: "It was a duck" 🙂

But still I find some of the posts comptetely unacceptable. Insulting statements with nothing that remotely reminded me of valid arguments anybody could learn something from.

Think: It is a real human being that is sitting in the other end, who has feelings and who does not need to be pissed upon.
 
It probably doesn't matter what the exact effect is called. But there IS a very distinct effect with HF impedance on Class B amplifiers. You can hear it.

At HF, amplifiers run out of open loop gain and current very dramatically. It can fall from 500,000 to 50 or so. Hence the whole feedback game doesn't work very predictably.

Any class of amplifier can have negative feedback. Not sure why you are singling out class B to discuss the fact that, if you go high enough in frequency gain drops to unity. All very predictable and part of standard university teaching on control systems. And argued to death here as to whether an amplifier needs to be flat to 100kHz or not.
 
snip...Think: It is a real human being that is sitting in the other end, who has feelings and who does not need to be pissed upon.
Arthur, there actually are some real human beings here, but TBH, because this is teh interwebs, there are also some who could learn a thing or two. If I named them , I would get into trouble. But a lot in your thread. 🙂

I build speakers. Some win a cigar, some just don't sound nice. Supersonic impedance matters. You can hear it. I use Class B as a useful example, because it is what a lot of us listen too, for power efficiency reasons.

In my dreams I might have Class A, but it's a much more expensive and hot-running approach. My own Rotel front-end setup below. It's sensitive to load. But works well with careful design of speakers and crossovers.
 

Attachments

  • Rotel RA931 Preamplifier.JPG
    Rotel RA931 Preamplifier.JPG
    22.8 KB · Views: 103
  • Rotel RA931 Output Stage.JPG
    Rotel RA931 Output Stage.JPG
    31.6 KB · Views: 103
  • Rotel RA-931 and Rotel RCD-965-BX Amp and CD.JPG
    Rotel RA-931 and Rotel RCD-965-BX Amp and CD.JPG
    14.1 KB · Views: 96
  • Modified_MA-R300MD_HT22-8.jpg
    Modified_MA-R300MD_HT22-8.jpg
    79 KB · Views: 152
Hi Steve,
my main system is: NOS 1794 DDDAC -> Pass BA-3 preamp -> Pass F-5 ClassA -> OBLs and 2 active Hypex/ScanSpeak subs,. It sounds very open and plain fantastic.

The OBL is inspired by Gravensen OBL10, but with a Tang Band W8-1808 midwoofer and my own designed crossover.

My PC system is: Twisted Pears 1794 DAC -> Rotel -> Dynaudio 2-way with serial filters, which also sound very good, but may need a sub.

The Dynaudio speakers are from their automotive series, and the Rotel will at some point in time be substituted by a high biased ClassAB with MOSfets.

As you can see almost all of it is homegrown, and much of it comes from this site.
And you have to follow your dreams for a ClassA 🙂

You could start with the PSU, that always is the same in Pass´s First Watt amps, then buy a 4 unit cab with cooling fins from Italy, and then you can decide wich amp to put in.

The newest model is the F6 which is reported to sound very good, but we are all waiting for the F7 🙂

The Pass threads are very long, but very helpfull, and the member 6L6 use to make visual building guides for the amps.

This is all a little OT - Sorry, but we probably came to an end anyway.

Best regards
Arthur.
 
I dont think it came from this clip 🙂

I tried to google it, some think it comes from a German saying mentioning a blue duck, and others that it is older and comes from France,

The saying has been used mostly in connection with newspaper articles.
 
Ducks aside, and getting back on topic, because we are, after all, in the business of building great sounding speakers which work well with cheapish ancillaries...😀

I like Nelson Pass's first watt ideas. Crossover distortion is a real issue with Class B amplifiers. I have always been a fan of those classic valve amps too.

You often wonder why some speaker designs win a cigar, and others sound horrible. Nobody likes the idea of a £6 cone tweeter, but they work for me. 😎

536753d1457667757-tim-reduction-tweeters-modified_ma-r300md_ht22-8.jpg


I found a kindred soul in Joe Rasmussen, who has tried equalising impedance to pure resistance totally. Hard to do with the bass peak, because of the values involved. But otherwise quite easy with some circuits.

Some ideas below. The red components are just impedance correction. The Sonus Faber Grand Piano Impedance looks VERY amplifier dependant. Sort of thing I avoid. But a moderate 1kHz impedance hump is not a terrible sounding thing. Most amps ought to cope with that.
 

Attachments

  • Sonus Faber Grand Piano Impedance.JPG
    Sonus Faber Grand Piano Impedance.JPG
    26.9 KB · Views: 501
  • BW3 Impedance Corrected Circuit.PNG
    BW3 Impedance Corrected Circuit.PNG
    9.3 KB · Views: 74
  • BW3 Impedance Corrected.PNG
    BW3 Impedance Corrected.PNG
    12.8 KB · Views: 41
  • BW3 Frequency Response.PNG
    BW3 Frequency Response.PNG
    19.7 KB · Views: 38
Last edited:
Status
Not open for further replies.