Low-distortion Audio-range Oscillator

Jump to Latest
D

davada

R.I.P.
Joined 2004
1audio said:
I did not have the Shibasoku when I designed that product. All the measurements then were limited by my equipment. The amp has 20 dB of gain and I measured .0006% THD, the limits of my hardware. I'll see if I can get a unit up and measure it again soon.

I don't see a second VAS in the 725 circuit (attached). I also attached the phono version of the amp I did for NuForce and a quick test of a FET buffer for an opamp. I can post the ltspice files and my modified parts files if there is interest.
Click to expand...

Yes Demian, please send/post simulation files.

I would need the model for the depletion fet at least. The easiest way to do this is to paste the model right on the ASC file.
 
D

davada

R.I.P.
Joined 2004
Richard you asked in some post what was wrong with the IC buffer. I find with just about all of National audio range parts a lot a very high frequency noise or perhaps high frequency instability. I've found National audio parts to be very susceptible to RF. This can cause what looks like very small rectified pulse. This often looks like low frequency noise. The kind noise that Vicnic sighted with the LME stuff. Sometimes it's there and sometimes it not.

Because of this I haven't been a big fan of National audio range parts. I will gladly accept a higher voltage noise part if it end with a much cleaner output.

Mostly the problems are as Scott said, " move a wire just one inch...and...".

And "Sounds like a case of 20 all being different".

Whatever we do has to be repeatable or at least trim able.

HP hand selected op amps going into the 339A. What happened to the op amp that didn't measure up. Back to the dealer and sold to someone else?

TI send me a handful of the LME buffers after raising an issue with tech support. I was told you can't just test one.
 
Last edited:
scott wurcer

scott wurcer

R.I.P
Joined 2004
davada said:
I read somewhere it could take up to an entire year to get approval then.
Click to expand...

More like a year to get silicon on a 40 transistor circuit. Each layer drawn by a draftsman and all simulation on pencil and paper while breadboards were real breadboards made out of vector board and pins. The AD524/624/625 was my last breadboard, unfortunately it got tossed.
 
F

fortitudine

Member
Joined 2004
C0G performs well in Scott and Samuel's recent LinearAudio article.
Presumably that is why they were recommended.

I bought 15 .001uf and 10 10nf xicon polystyrene from Mouser. i have not thrown them into an oscillator but the DA is comparable to silver mica and other polystyrene that I compared them too. Biggest problem for me is the restricted range of values.

Try them and report back!
 
anatech

anatech

Administrator
Joined 2004
Paid Member
You have to measure each one as the parameter spread is wide. As for DA, they are as good as any other polystyrene I have used before. Some polypropylene capacitors are very good as well. C0G / NP0 capacitors are also good, measuring similar DA values.

Note that dielectric absorption isn't anything like ESR. Dimitry is right though, there are quite good ceramic capacitors around with tighter tolerances than polystyrene capacitors these days. I really wish someone would bring back production like the old 630 V, standard values that I used to buy decades ago.

-Chris
 
You must log in or register to reply here.
Top Bottom