DIYHiFi.org

This thread is a follow-up of the original one, which I have started at the other place almost two years ago:

http://www.diyaudio.com/forums/showthre ... adid=40783

As diyhifi.org shows more respect for the principles that I defend and respect, I will stick around, so you'll get assistance and answers to your questions from me here.

Basically my goal was to use the AD815 current-feedback (video) op-amp for an audio preamp, alone.
That means not using an input buffer stage (which would make it easier to deal with the input and output DC offset of these chips) or servos.
I also didn't want to muck it all up with input and (it had to be both) output coupling caps, the easy way to kill DC and the music too.
You can see caps on the output of my pre, but also a direct output, which recommend as long as you have input caps on your power amp.
Please don't use output caps on the pre if you have input caps on the power amp (or the other way around), because you'll immediately notice (if you care to listen) that you've done something wrong, the pre is very transparent and doesn't deserve that.
With the DC null circuit you will be able to null DC (aha!) to 0.0mv but only at a volume pot setting that you choose for that.
I recommend setting the volume pot at ~10 o'clock position and then trim DC.
Yes, as there is no input buffer (and hence no constant input impedance for the chip) the output DC offset is dependent from the position of the volume pot, but trimming it you'll have around +/-20~30mv DC on the outputs, worst case, 0mv at 10 o'clock volume setting and very low at the volume levels you usually listen to.
This is not critical because there are DC coupling caps somewhere, as explained above, but this circuit also minimizes the input DC offset of the chip, which otherwise would not be good for the health of your volume pot, wouldn't last long.

Build it for your personal use and enjoy.
I can guarantee that you are not loosing your time.
Enough talk, here's the schematic.

_________________
Carlos Filipe

"Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius - and a lot of courage - to move in the opposite direction." Albert Einstein

This an excellent pre Carlos, it blows my modified Nac pre into the weeds thanks for sharing it!
Now I've got an AD815AY spare I must have a go at building one point to point
Is the AD815 any good for headphones Carlos?

Curiously I didn't try it with headphones, but some did and say yes.
It has enough output current to make your headphones sweat.

_________________
Carlos Filipe

"Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius - and a lot of courage - to move in the opposite direction." Albert Einstein

Thanks Carlos
You may know I've already etched one using Rudi's pcb layout and servo which sounds very nice, infact it works great with all my amps
Now I want to build another point to point and keep it faithful to your circuit apart from the decoupling caps which would be 22uf, its the nearest I have to 33uf

Anyway for anybody else thinking about building Carlos AD815 pre I can highly recommend it

Carlos,

I don't get it. What's the novelty with your AD815 linestage schematic? It looks like a regular non-inverting stage. Something in any opamp data sheet. What makes this one special?

Didn't Jung and others promote using the AD811 many years ago?

jh

EDIT: Oops, my bad. Doesn't belong in censorship.

MODS' EDIT: AD815 related posts moved into the appropriate thread.

_________________
Jim Hagerman

I only see the power supply and offset nulling.
Please post the schematic too Carlos.

As I see it Jim it's more about an unusual application of the AD815.
Of course Carlos is free to post about his findings.
I am inclined to put a pair of FETs at the inputs......
Then make it folded cascode....only using the AD815 as a buffer.....
Then replace it by the double diamand buffer.......
Hey then I need the AD815 no more!
I have built a discrete opamp.

Jokes aside, go ahead Carlos!

Thats strange, I can see it at the top of the schematic

Yes now I see it too!

Yes Jim, if you look at it, it doesn't look different from a bog standard non-inverting op-amp aplication.
It is when you build it that you find yourself with some issues, respect those component values and everything is ok. Read my post and you'll understand why.
Btw I really mean respect the component values because someone used bigger caps on the output of the regs (33uF recommended) and it changed the sound completely, for the worst. When he put 33uF caps, as recommended, he was amazed.
There's more on that schematic that it is apparent.

Jung promoted the use of the AD811 in the feedback loop of a voltage-feedback op-amp. I didn't want to do that.
As you can see in my post, the goal was to use the AD815 alone.

Elso, you can put whatever you want before and/or after the AD815, but IME more is less.
Removing the AD815 is not an option.

_________________
Carlos Filipe

"Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius - and a lot of courage - to move in the opposite direction." Albert Einstein

Hi Carlos,

Have you compared AD815 against other CFB opamps? I?ve preferred LM6181 but since I?ve tried OPA627 it became my part of choice. This experience is limited to the I/V conversion though... anyhow, if I understand you well, in such an application you?ve found AD815 superior to the OPA627? Any idea about its behavior in the I/V application?

Pedja

Ok Carlos, I'll try and find some 33uf, its a shame really because I have some nice 22uf Rubycon ZL's

I've built a few op-amp based pre-amps before including OPA627 and none of them came close to the AD815 imo:)

BTW Pedja have you seen the footprint of the AD815AY, it looks like a LM3875

Hi Pedja,

I've also once tried the LM6181 for I/V but put back the AD627, I prefered it.
I didn't try the AD815 for I/V, but as a preamp it's in a different league to the OPA627. It is superior, and not by a small margin.
It has a dynamics to cry for, but with very tight bass, gorgeous midband, and an exceptionally clean and extended treble.
I really mean clean, precise treble.

This treble precision has a reason, and as a follow-up to my response to Jim, the AD811 didn't appeal me, it isn't much better on paper than some others.
Look at the PSRR graph on the AD815 datasheet.
This graph goes up to 300Mhz , and PSRR never hits 0
It's 'typical' PSRR figure of 66db doesn't impress, but it doesn't say it all. As always, this spec is for the low frequencies.

_________________
Carlos Filipe

"Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius - and a lot of courage - to move in the opposite direction." Albert Einstein

Hey, that's ok!
Smaller yes, bigger no.

_________________
Carlos Filipe

"Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius - and a lot of courage - to move in the opposite direction." Albert Einstein

Thats great I really like Rubycon ZL's but I also want to try and keep my next AD815 as faithful to your circuit as possible

How about posting some piccies of your pre