RIAA design

[canielsen]

Here is a design for a low noise RIAA amplifier, that i made some times ago. The power supply has a capacitor multiplier to reduce noise from the supply.
I have build several of these amplifiers to my friends.



( The box to the power supply is not on the picture, while i dont have the box when the picture was taken )

The schematic and PCB layout can be downloaded at: http://netload.in/dateiJVCn08HZEh/PolyphoniaRIAA.rar.htm

[optikon]

Thanks for sharing.. the input stage bias ~ 10mA is kind of high, you might see better noise reduction if you reduce that to 1 - 2mA on those 2sk170's

[sarah90]

Nice build. I had to google to find out what a riaa amplifier was (silly me). If others need to know, I found this: http://www.beoworld.org/article_view.asp?id=67

[metal]

May be you have to read more Audio, it is not a good idea to use LM337. Even to make things worse, you better use shunt regulators instead! Something like that made by a member called Salas in diyaudio.com

good work any way, thanks for sharing.

[sarah90]

I'm just wondering, what makes the lm337 bad for audio?

[metal]

this paper shows in way or another that LM337 is a bad idea: http://waltjung.org/PDFs/Sources_101_P2.pdf

[optikon]

Most typical pass transistor style regulators like LM317, 337 and the 78xx or 79xx style have poor high frequency rejection. This is mostly due to a single non cascoded pass stage.

It doesnt necessarily make them bad for audio use. Why? Because the audio circuits themselves still have rejection of unwanted components that make it through the regulators. As long as it's not bad enough
to be heard then it's fine.

Have a look at these for better than usual rejection

Positive
http://www.linear.com/product/LT3080

Negative
http://www.ti.com/product/tps7a3301

[sarah90]

Nice to know the good ones. Although according to the jung article the lm317 was not as bad as the lm337 and could even be recommended for some designs.

I hear generally nice things about micrel regulators. Have they better ones for audio?

[optikon]

Nice to know the good ones. Although according to the jung article the lm317 was not as bad as the lm337 and could even be recommended for some designs.

I hear generally nice things about micrel regulators. Have they better ones for audio?

Micrel as far as I can see has nothing advantageous for audio. One of the issues with audio is that the power supply rails tend to be > +/-12V and Micrel has most of its devices in a lower voltage process (5V) so that greatly limits the choices. I did find a few of their higher voltage devices but the one I looked at didn't even show a plot of rejection vs frequency, which tells me its BAD. Looking at it's schematic, it also has a single pass transistor output so I expect it to be bad in this regard. They do have something called a ripple blocker power filter but it looks like it's for low voltage digital devices (like powering FPGA's, uP, uC etc..) Maybe Micrel has good cost going for them, I dont see it.

The ripple rejection issue is an analog problem and Micrel isn't known for being good at analog. Stick with TI, Linear, ADI etc.. or roll your own for these types of designs.

[Sideshow Bob]

I once used the suggested design in the LT1115 datasheet. http://cds.linear.com/docs/Datasheet/lt1115fa.pdf
 With a proper power supply the sound was actually very good. To bad the LT1115 is somewhat pricy