Re: Piezo audio preamp works but I would like advice

From: John Woodgate (jmw_at_jmwa.demon.contraspam.yuk)
Date: 07/26/04 

Date: Mon, 26 Jul 2004 22:00:46 +0100

I read in sci.electronics.design that Jon <u035m4i02@sneakemail.com>
wrote (in <298d9774.0407261224.2531a4c4@posting.google.com>) about
'Piezo audio preamp works but I would like advice', on Mon, 26 Jul 2004:

>isn't the gain 680/1000?

Yes: my arithmetic on-screen is prone to error
>
>> With 1000 pF feedback capacitors, the lower -3 dB frequency is about 150
>> Hz. You might consider increasing the value to 1.5 to 2.2 Mohm.
>
>I found some 1500 pF caps instead.
>
>> You don't need the 10 uF output coupling caps. The outputs of the op-
>> amps are very close to 0 V. If you keep the caps, electrolytics are not
>> ideal because there is no polarizing voltage. 0.1 uF is just big enough
>> (-3 dB at about 40 Hz, with the 24.75 kohm load), but you could go a bit
>> larger.
>
>absolutely not necessary? i have this instinct to keep them there...
>and it might see lower loads than that, if i have the two connected in
>mono mode, right?

If you have room, keep them, but they are not necessary.
>
>> Add another 22 uF from + to ground, of course.
>
>hmm.. i only have one 22 uF. the only other tiny caps i can find (from
>junk boards) are 220 uF. won't hurt anything to go to 220 will it? i
>noticed the clipping level was kind of asymmetrical when i used large
>caps before, but leveled back to normal after they charged up.

That's the disadvantage of using a very large value. Put the 22 uF from
0 to - and one 220 u from + to -.
>
[snip]
>
>so either 1 rail-ground and 1 rail-rail, or 2 rail-grounds are valid
>configurations? can you explain what these caps do?

They prevent signal-frequency changes in supply current (due to the
Class AB or B output stages) working with the supply-line impedances to
produce variations in the supply voltage at the pins of the op-amps,
particularly keeping signals from that source getting to the + inputs of
the active op-amps. Such variations can produce the weird effects you
observed with large signals.

>i search "bypass
>caps virtual ground" and similar, but find very little info. why do the
>rails and ground have to be connected by low impedance at audio
>frequencies, and why did that make the amp "clip" early?

See above.
>
>http://www.analog.com/library/analogDialogue/archives/35-02/avoiding/
>
>This site shows only one bypass cap from positive rail to ground, but it
>has *two* of different values in parallel. This has to do with one of
>the caps no longer behaving like a capacitor at high frequencies, right?

Yes, but a modern 1 uF is good to 1 MHz, so it's not an audio issue. The
22 uF and 220 uF are quite OK for audio, provided they are not dried-up.
>
>why do they only have the cap on one rail? or wait... those are across
>the power terminals, not connected to the virtual ground. i guess that
>is a different type of single supply circuit altogether? i am
>confused...

Yes, it is a different technique. You have an op-amp buffer to produce
your mid-point, so you actually have a dual supply as far as the signal
circuits are concerned. 

-- 
Regards, John Woodgate, OOO - Own Opinions Only. 
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

Relevant Pages

  • Re: silk screen
    ... WAF is not a factor in my Audio Dungeon. ... The shipping cartons they sold me to ship it in for maints cost more ... than the TFTF caps I am now listening happily through ... ... Keep your ears happy. ...
    (rec.audio.tubes)
  • TFTF caps, once more into the breach, dear fiends
    ... disinterested newsgroups really improves my experience. ... I put new caps in my Audio sculpture and enjoy the sound and my post ... I am an old sick man who enjoys playing with my audio toyboxes. ... I still think Graham is a bit of a shit. ...
    (rec.audio.tubes)
  • Re: TFTF caps, once more into the breach, dear fiends
    ... stuff but the industry standard Audio Precision kit. ... It's become the recognised norm for audio measurements. ... What signal level, what frequency, what bias, and what caps did you try to ... A bit like measuring a metre rule and finding it measures a metre you know! ...
    (rec.audio.tubes)
  • Re: need help troubleshooting a pair of powered speakers for pc
    ... But I am just using spare time at work to learn a little ... and consider it an educational expense. ... I will check those caps, ... > audio very weakly through speaker. ...
    (sci.electronics.repair)
Quantcast