2 quick questions

[JPeitzman]

1) I am making a IC based preamp, can I do buffer > pot > buffer then maybe a gain stage? Is it better to do pot > buffer > gain? This would be mostly to be used while testing, I am designing a more permanent preamp, this would be used for testing amps as well as testing the sound of different Op Amps in a preamp circuit. I can also make some small boards using discrete parts with 8 pin adapters to test them as well.

2) I just finished my first stereo pair of LM3875 based amps, I love them! They sound great, I am still partial to my JLH though. I sampled the LM3875's along with 4 MJ15003G's (for the JLH to replace the 2N3055's and I must say wow), and 24 MUR1560's for power supplies. I must say that both National and ON where awesome companies to deal with, top rate! Plus ON drop shipped from Digikey in Minnesota so I got my ON parts in a day!! Anyways I am planing on refining my 2 LM3875 amps and then using the other two chips to make a second pair that is inverted. I might make some PCB's and add in buffers and DC Servo's...what not, see how stuff sounds.

My question is, has anyone ever tried to bias a LM3875 or the like to operate in class A? I did it with a test circuit I made for a CMOY using a LM358 (remember I said test) and I didn't notice anything. I did it too with a LM358 buffer test circuit, which I don't know if it does anything to a buffer. Again nothing audible, but I am sure there was something there. I would be curious to see if you can or if it would make a difference on a LM3875.

BTW: For being as about as cheap as op amps come (not quite 2N2222 cheap, love those for testing) LM358's really don't sound to bad. My dad uses them so I get them free for testing.

[JPeitzman]

Very very sorry for the double post, I know some forums I am a part of frown upon it.

Anyways, is there a point to straddle the pot with a pair of buffers? I just breadboarded a circuit that used ½ of a LM358 > 100k pot > ½ LM358. It doesn't sound much different than with pot > buffer. I haven't hooked the O-Scope up yet or meter or anything, but should there be a difference? I also stacked a pair of LM358, again not much difference, but I did that just for fun.

Would this circuit benefit from having the gain stage? If I do buffer > pot > buffer > gain stage, say a small gain, I assume this would be nice for driving longer interconnects.

[BWRX]

If the ratio of your source's output impedance to the pot's impedance is roughly 1/10 or lower then you probably won't notice much difference with that buffer between the source and the pot. If your source has a high output impedance then it would be a good idea to use the buffer between the source and the pot.

If you'd like more gain in front of your amp then by all means change the gain of one of the stages to your liking. Some people say op amps sound better when used at higher gains but that's something you should experiment with and decide for yourself. One thing is for sure, don't overload the op amp or the sound quality will go downhill.

[JPeitzman]

Thanks for the reply. So it really shouldn't hurt the sound quality though if I have two buffer stages should it, as long as everything is within it's operating range? This pre along with some of my amp may be traveling to friends' houses for testing and listening, so I don't always know what the source will be like.

With the gain stage too I figured I could add it in and if I don't like it I could just jumper it. This is kind of a test circuit for the more advanced pre, of which I plan on having a chip and a BJT pre so I can listen and compare...but that is a different thread.

[BWRX]

Originally posted by JPeitzman
This pre along with some of my amp may be traveling to friends' houses for testing and listening, so I don't always know what the source will be like.

Not always knowing what the source will be like is certainly a good reason to use a buffer on both sides of the potentiometer Plus you get to use both halves of that dual op amp.

Some other dual op amps worth trying are the NE5532, TL072, OPA2134, and LM4562.

[Nordic]

A buffer on either sides of a pot wil provide a more constant impendance to both the source and driven stage... however on a practical level, it is not always as easily demonstrated as theory would make you think....

Stacking buffers in series is quite popular, and many believe it leads to better performance...

[JPeitzman]

Thanks for the info guys! I think I will pick up some TL072's soon or the like, I have to put in some orders soon anyways. I have a very simple sample proto board mostly done, work has draw me away from finishing. I will finish on Sunday (today) night and see how it goes. I am still playing with resistor values on the gain section, and still debating about biasing class A or not.

I have 1 quick question, when I build the inverted GC with buffer do I have to have anything after the buffer? IE: on my current GC there is the 22k resistor that ties the input to ground, you would have that on the buffer at the input, so do I need one on the LM3875 as well? Or do I need to stick a resistor between the buffer and the LM, or is it just something I have to play with? I am a very strong believer in keeping passives out of the signal path as much as possible.

The transistor pre is soon to follow, then hopefully a transistor power amp. It is a slow process for me due to work and school, but it has been a very rewarding hobby. I have a ton of tubes too, and I love designing with them and the sound they make...just can't afford all the tranies, lol.

[BWRX]

Originally posted by JPeitzman
when I build the inverted GC with buffer do I have to have anything after the buffer?

No, you will not need any resistance in series with the output of the buffer when it is driving the input resistor of an inverted GC.