Non-Inverting GC BETTER than Invertin GC+Buffer?!

[GregGC]

Just registered as a member of the forum.
You guys are great. I got the bug too and I'm going to build monoblocks based on LM3875. I was deciding on the topology (Inverted vs. Non-Inverted). I did quite a bit of reading and found that people prefer the inverted topology but than (for obvious reasons) they put a buffer at the input. Some claim that that way GC sounds even better. Well the buffer is an OPAMP in a non-inverting mode with a gain of 1. If the Inverting topology has an advantage, why implement it together with the non-inverted (buffer). You might as well stick to the non-inverting GC on the first place. You'll end up with less components on the signal path, which is the reason on the first place to build the GC. Please comment. I'd appreciated.

Great forum. A lot of inspiration. Thanks!

Greg

[hitsware]

Fully agreed. A resistor in series with, and a cap to ground of the + input, gets rid of the funnies that can happen with the non-inverting configuration. ~ 2.2kOhm and 0.001uF

[GregGC]

Hi hitsware,
You are referring to a LP filter at the front of the NIGC (non-inverting GC), right?

I think KIS would be my preference. Non-inverting GC with a 100k pot at the front and an optional RC LP filter as you mentioned would be the best topology IMHO. And if i want to get rid of the NF cap have to make shure I balance the input curents of LM3857 so that it minimizes the output ofset as much as possible. Also I was thinking that one of the reason for difernt sound coud be the difference in value of the NF resistor (220k vs 20k) used in the IGC. So I'll try building NIGC with NF of 220k and 22k to gnd and 22k to gnd from the + of the 3875. It should eliminate the output ouset. At the same time with input impedance of 22k in paralel to 100k pot will turn the linear 100k pot into semi Log. I tried it and it works quite well. This way, only one input cap in the whole circuit. Hope it'll work as I think.

Greg

[millwood]

Originally posted by GregGC
If the Inverting topology has an advantage, why implement it together with the non-inverted (buffer).
Greg

Greg: the same topic has been extensively discussed and debated in this forum. As you will find, there is no conclusive and / or emperical evidence that inventing topology sounds better.

The rumored "gains" in THD figures due to common mode rejection are in the sub-micro percentages, at best.

To give you any comfort, most SS amps use non-inverting topology. If inverting topology had that much of an advantage, you would have expected just the opposite.

Just because people believe in things doesn't mean those believes are right and fact-based.

And I am sure you will find a lot of similar things in audio in general and on this forum in particular.

[purplepeople]

I'm guessing that the buffer is used to drive the lower than normal input impedance of the IGC. Many people are using input impedance of about 10K for their IGC. The buffer could just equalize the system with the 47K impedance of many sources. What I don't understand is if they use an IGC, why not also use an inverting buffer?

:)ensen.

[GregGC]

If you use opamp as a buffer it'll have the same disadvantage of relatively low input impedance as the IGC, but at least the gain of the amp will not depend on the vol control. Still more transistors on the way of the signal thogh.

Greg

[janneman]

Originally posted by GregGC
Just registered as a member of the forum.
You guys are great. I got the bug too and I'm going to build monoblocks based on LM3875. I was deciding on the topology (Inverted vs. Non-Inverted). I did quite a bit of reading and found that people prefer the inverted topology but than (for obvious reasons) they put a buffer at the input. Some claim that that way GC sounds even better. Well the buffer is an OPAMP in a non-inverting mode with a gain of 1. If the Inverting topology has an advantage, why implement it together with the non-inverted (buffer). You might as well stick to the non-inverting GC on the first place. You'll end up with less components on the signal path, which is the reason on the first place to build the GC. Please comment. I'd appreciated.

Great forum. A lot of inspiration. Thanks!

Greg

Hi Greg, welcome!

What you say is logical and puts the finger on the right spot. However, you will find out that this is not necessarily an advantage or appreciated at this forum. Still, it's a great place to learn and teach. Have fun!

Jan Didden

[GregGC]

Dear Jan Didden,

Thanks for the welcome.
I think that part of the learning process is questioning what we/people take for granted. I may be wrong. I'm new at this forum and I don't mind being advised on what an appropriate behavior for this forum is (mind you I red the rules when I signed-up). I thought, though that we are all open minded individuals that exchange thoughts on how to improve on a great idea. I don't get offended if I find I'm wrong in something. Also I'm not sure I'm right at assuming (read the headline). Otherwise I wouldn't have open the topic for discussion.

Life is a learning process.


Greg

[GregGC]

Originally posted by millwood



Greg: the same topic has been extensively discussed and debated in this forum. As you will find, there is no conclusive and / or emperical evidence that inventing topology sounds better.

The rumored "gains" in THD figures due to common mode rejection are in the sub-micro percentages, at best.

To give you any comfort, most SS amps use non-inverting topology. If inverting topology had that much of an advantage, you would have expected just the opposite.

Just because people believe in things doesn't mean those believes are right and fact-based.

And I am sure you will find a lot of similar things in audio in general and on this forum in particular.

Thanks for your input. What do you think about my prvious message descibing the circuit I'd like to use (No NFB Cap)? I still don't know how to post pictures/diagrams.

Greg

[moamps]

From NS LM4780 datasheet;