Hmm. then additional output buffer amp is not required.
Do you recommend isolation resistor for bufferamp output?
i mean 33E-47E resistors to isolate the opamp output from track/load capacitance(D.self recommends for tl072 opamps). i am asking because as you have said, minimum input impedance is required for good performance.
The filter's op amp does act as a buffer.
I would not use an output resistor for the input buffer, because the filter circuit works best with
a low source impedance. A pair of 100 ohm resistors on the filter circuit's op amp outputs is fine.
Add decoupling capacitor, 100nF between +15V and -15V close to each opamp, the NE5532 definitely needs this, some opamps are less critical.
The Sallen+Key filter circuits are unity gain in the passband normally.
Place between preamp output and individual power amp inputs, so whatever level that is.
You can make the resistors smaller and capacitors larger to reduce impedance which lowers noise floor. Something more like 2--5k for resistors will be good for this (it may be an issue with sensitive tweeters, should the crossover happen to be the limiting factor for low-noise)
Thnks. I know, douglas self recommends that & he says NO NO to tl072 for sallenkey filters! 🙄
5.5k & 15nf is close, please see 🙂
Attachments
I fully understand that the enquiry is from a novice so what I say next may be confusing but worth saying. Firstly, the tweeter usually needs very little power. A 1 watt amplifier might be enough. This makes class A very possible. I dare say a LM1875 could be configured as a SE class A device using a resistor to the negative power rail. The LM1875 can even be configured as an active filter. The output can be AC coupled as a protection against DC and filter pole. Unfortunately this requires test gear and knowledge to do.
With tweeters physical interaction with the baffle is the sound quality factor that really matters. In view of this the quality of the amplifier and the value of the output capacitor might be prime. Then the gain of the amplifier. It was always said phase shift is most important where tweeter crosses over. This coupled with the chosen frequency is the sound of the arrangement. Some tweeters like Swan ribbon type have no obvious low frequency resonance. Used with care the crossover can be a single capacitor. This makes an audiophile solution possible for a novice
Often the more complex filters are to allow the tweeter to go loud at a lower frequency. In pro audio work this is important as tweeters will see significant power. A friend recommended I use his pro audio digital filter as it can do any filter you can imagine without phase shift even. To his surprise it even has single pole types. His observation was that an analogue filter can not match this. He was an engineer at SSL mixing desks so no amateur. SSL were not the best sounding which they knew. That was due to the signal passing through nearly 100 op amps in series sometimes. These were the ubiquitous NE5534/32 types. They probably made possible many treasured recordings despite easy of use being dominant at SSL. They admired Neve as a competitor. John knows I wouldn't use digital for this. He thought I should try it as he felt ideal filter curves a revelation. I would imagine this technology is possible for the novice.
With tweeters physical interaction with the baffle is the sound quality factor that really matters. In view of this the quality of the amplifier and the value of the output capacitor might be prime. Then the gain of the amplifier. It was always said phase shift is most important where tweeter crosses over. This coupled with the chosen frequency is the sound of the arrangement. Some tweeters like Swan ribbon type have no obvious low frequency resonance. Used with care the crossover can be a single capacitor. This makes an audiophile solution possible for a novice
Often the more complex filters are to allow the tweeter to go loud at a lower frequency. In pro audio work this is important as tweeters will see significant power. A friend recommended I use his pro audio digital filter as it can do any filter you can imagine without phase shift even. To his surprise it even has single pole types. His observation was that an analogue filter can not match this. He was an engineer at SSL mixing desks so no amateur. SSL were not the best sounding which they knew. That was due to the signal passing through nearly 100 op amps in series sometimes. These were the ubiquitous NE5534/32 types. They probably made possible many treasured recordings despite easy of use being dominant at SSL. They admired Neve as a competitor. John knows I wouldn't use digital for this. He thought I should try it as he felt ideal filter curves a revelation. I would imagine this technology is possible for the novice.
Hello nigel 🙂 Yes, you can say that. i know little bit of electronics(DIY) but i'm aware of complexities related to audio. Currently i've more than hundred of books, from dostal jiri to bob cordell & i'm going through them.
i have a wish to meet mr.pass, one of my dreams
BTW i've couple of good Class-A design, from 1w to 10w.
But i hate complexity & costs 😛 🙂
If you want to say anything understandable to me, you're very much welcome 🙄
Last edited:
I am using a PAM8406 Class D module 1.6 watts. 5V USB powered. It's quite nice sounding and being class D a little like class A in distortion graphs. Worth a thought. It's a very ingenious design. I have it with Teac LXS-8 mini monitors which were sold by others. My TV speakers. The tweeters are supurb considering how reasonably priced they were.
Hi rayma. The filter(good, as much as i understand) you have recommend me is a 2nd order (12dB/octave) linkwitz-riley filter & has a Q of 0.5 (sub-Bessel), and the crossover frequencies align at the -6dB frequency rather than the
-3dB frequency found with Butterworth filters(Source: ESP).
My question is while the Q of my enclosure is near 1.0,
why did you choose sub-bessel type filter? On the other
side; interestingly you choose 2nd order filter for sealed box(same rolloff slope). Can you explain in little detail?
My next question is, can i play with other values
(like 5.5k & 15nf) instead of 3.3nf & 23.7k using this equation? 1/(2piRC) (the value you recommend is very good, close to crossover point but 3.3nf is bit of odd value, maybe it is available in mouser but i don't know if it is available here, locally). My last question is, though i trust you & i aware of this configuration but i heard that changing the polarity of one of the driver can effect sound quality, is it true? how much? & why not invert the signal of high pass filter??😕
True, Technology is much better now 😎
Very affordable price with good performance.
But i was saying as a diy enthusiast. lil diy only 🙂
Learning is better than copying.
hi, have you seen my post?
i'm waiting 🙂
hi, have you seen my post?i'm waiting 🙂
Last edited:
I bought mine ready built for £1. Not bad and works on a 5V battery pack. Thats a very pure power supply. My pack usually would work a phone. That's stereo for that price
hmm. Class-d is now very mature, efficient & good sounding.
Not old sinclair one(first class-d module)🙂
The curves of that chip are extensive. They look like class A circa 1969. As Douglas Self says the crossover distortion is at the switching frequency of 240 kHz in this case. Bridge output. Music out of phase and switching in phase. Reduces the components used. Very nice engineering.
Sinclair after the not so great class D went on to make the.Z30 class AB that inspired many including A&R A60.
Sinclair after the not so great class D went on to make the.Z30 class AB that inspired many including A&R A60.
If you can still get them MC33078/79 are very nice. If no one said a few 5532 could drive a tweeter.
😀 Yes, 5532 can. D.Self have made one prototype, maybe for Elektor magazine. I don't want to spend too much, 5532 is enough for now.
BTW 'small signal audio design' by self is a nice book.