[djwkjp]

i have been lurking around this forum for quite some time now and i have a few questions about amplifier design. i have built a few different amps and now i would like to design and build my own. so far i have the schematic close to done. it is based on the cmoy and has a right and left channel as well as a ground channel. i dont know what opamps yet, but right now i have some opa134s and 627s and i would also like to try ad8610s. it will also have a buf634 on each channel and be biased either by a crd or a jfet cascode. based on what i have read from the chips' data sheets, the buffer should be placed inside the feedback loop, but i noticed on the schematic of the ppa and pimeta that there is another feedback loop before the buffer that has a higher resistance. what advantage does that have? i am also wondering about the ground channel. should that be buffered as well, and should the gain on the ground be equal to the gain on the other 2 channels? also, why is it neccessary to have the bandwidth limiting cap on the ground channel and should i have the bandwidth resistor hooked up on the buffers? if it helps, i am planning to run the amp off of 3 or 4 9V batteries and probably a wall wart for home use. i dont know what gain i am going to use yet, but it will probably be 6-10 since i plan on getting some er-4s. sorry for the long post, but i will appreciate any feedback.
thanks
dan

 

[djwkjp]

after reading about the jung multiloop in the ppa, i think i have answered a couple of my questions. i ma still wondering about the ground channel. for a primarily battery operated amp, would there be any advantage to buffering the ground channel and biasing it to class a or would it be ok to skip it?
thanks

 

[tangent]

Quote:

it is based on the cmoy

Sounds more like a PIMETA clone to me.

Quote:

nother feedback loop before the buffer that has a higher resistance. what advantage does that have?

http://www.elecdesign.com/Articles/A...7207/7207.html
http://www.elecdesign.com/Articles/A...7206/7206.html

Quote:

i am also wondering about the ground channel. should that be buffered as well

Absolutely. It has to sink all currents sourced by the two output channels.

Quote:

should the gain on the ground be equal to the gain on the other 2 channels?

No. That just adds noise to the system. The gain of the ground channel should be 1.

Quote:

why is it neccessary to have the bandwidth limiting cap on the ground channel

Because it can often become unstable without it. Since you only need the ground channel to respond in the audio frequency range, limiting ground channel bandwidth is acceptable.

Quote:

should i have the bandwidth resistor hooked up on the buffers?

Only you can answer that. It's like asking whether you should upgrade your CD player. Yes? No? I don't know...

 

[djwkjp]

what exactly does the bandwidth resistor do? does it improve frequency response? now that you mention it, it does seem an awful lot like the pimeta. i guess i just hadnt realized that.

btw thanks for the links those look like interesting articles. i'll have to read them when i get home from work

 

[ppl]

In his 1982 paper Fuzzy Distortion in analog Amplifiers http://www.essex.ac.uk/ese/research/...distortion.pdf

We find on page 8 of the 10 page pdf Fig 7 entitled basic Multiple-Loop feedback Amplifier Topology. Yes folks this paper predates the Jung article. However the entire 10 pages are totally worth your time.