Improving amplifier input stage

[stevensoosai2527]

Most commercial home audio amplifier are inferior in high freq response due to overcompensation or very basic design of input stage. For instance, they use single resistor ccs instead of zener stabilised. They also not using current mirror at the collectors of diffrential to aid current balance. In computer simulation, implementing these additions improves amplifier response. In reality, how much improvement will we get if we put all these into an ready made commercial amplifier? Is it worth for the effort? Can anyone pls answer?

[jkuetemann]

I don't think it is a simple a just 'bolting on' a few new building blocks to improve performance in any design, commercial or DIY. There will always be trade-offs.

For instance, take a design with a resistively loaded LTP input. It will have a certain amount of gain (transconductance). If one were to now introduce a current mirror load to the LTP, the gain (transconductance) increases and the gain structure of the amplifier changes. This can have detrimental effects on stability ranging from increased THD to outright oscillation. Same can be said for any such 'improvement'.

The whole amplifier has to be evaluated and possibly re-engineered to accommodate significant changes. Not all commercial offerings are solely designed 'on the cheap' either. I've seen many that use good paper designs then throw out the baby with the bath water when it comes to supporting circuits such as the PSU and input sections, not to mention some of the heresy seen in PCB / chassis layout. Some can be fixed and others aren't worth the effort.

The reality is if one wants to incorporate certain elements together in a design, then design and build one. I have built a few conventional amps with various design features that have performed well. That said I am inclined to leave a commercially made amplifier alone lest it actually require a repair, if only to avoid upsetting the balance the original designer has put into place. For example, I acquired a Yamaha PD2500 power amplifier gratis that only needed a minor fix on an auxiliary PSU. While it was stripped down for a full re-soldering and cleaning I was tempted to muck about to try and improve it. After careful thought and examination of the schematic It just didn't seem worth the risk to the amplifier when it sounds decent without modifying it.

I think that there can be some improvement to be had in some commercial products by basically fixing the compromises imposed on the engineers by the accountants and the errors brought by the draught persons responsible for the physical packaging. Aside from that IMHO, unless you are prepared for a thorough redesign, let sleeping dogs lay.

[AKSA]

Are you absolutely sure that these 'improvements' improve the sound quality? It might be consider that lots of companies have spend a lot of money developing rudimentary, inferior circuits.

[stevensoosai2527]

I am absolutely sure the "improvements", i.e. utilising high stability current sources and current mirrors contribute towards lesser THD. As how im seeing in simulation, when input stage diffrential is not accurately balanced, or when current starving occurs at the ccs, it aid to the generation of third harmonic immediately at the voltage amplification stage. Most commercial manufacturers may well realize this, but not investing into it due to the minute improvement at the upperband of the opertating frequency of the amplifer is not worth for cost of the improvement. Pls not current mirrors are widely used in high performance ic op amp circuit. Any reason for that? What im not sure is, in the real world application, can we perceive the subtle difference of the improvement.?

[Vostro]

"very basic design of input stage"
Input stages can be vastly different, and also the VAS they supply.
I'm not sure what type of input stage you refering to.
Some conbinatons, have there own complications, and limitations,
Sometimes leaving out current mirrors for example, and make a mass produced amplifier more stable, they dont have the choice to manually test and fine tune each amp.

With double ltp input stage feeding push pull VAS it becomes very difficult to put current mirrors in ips, because its nearly imposible to define VAS current, without complicating things.

Trade offs them must be made

Regards

[sreten]

Hi, comprehensively covered by D.Self, online and in books, rgds, sreten.

Distortion In Power Amplifiers It is linked to compensation