HIGH Gain non-inverting feedback loop question
Mar 31, 2006 at 8:28 AM Thread Starter Post #1 of 3


100+ Head-Fier
Oct 31, 2004
Hi guys, I have a uni project atm. I'm not directly asking for spoiler help but.

Me and my partner are building a wireless IR speaker/headphone audio transmitter.


We have to change the design extensivly, and the first thing to go is the ca3237e chip. From datasheets I have calculated the the expected gain from that chip the way its configured is about 101.2dB, yes decibels, not Gain as a ratio.

Basically I know I could do a 2 stage feedback loop similar to theirs using 2 chips off an opamp, just say OPA2227U, which is what we are going to do.

But then it comes down as using 2 feedback loops with R2=1M ohm, and R1 = 2.5kohms.

I know there is an alternative way to do the feedback loop, I still remember from first year it looked something like this

I know the formula is going to be relatively simple. ie (r1 + r2 / r3)+1 but that is only doubling the gain and we need a gain value around 800 to keep this constant with the original data.

Does anyone else think that 101.2dB sounds quite rediculous?

Any help or comments or alternative feedback loops would be greatly appreciated.
Mar 31, 2006 at 7:29 PM Post #2 of 3


Headphoneus Supremus
Jul 24, 2002
the T network you show doesn't change the V gain

simple Resistive ratio dividers for 2x cascaded Av 300 amplifiers are not a problem

if you are going to keep the 40KHz carrier frequency you need much faster op amps, - look at op amp open loop gain at 40 KHz, not the DC value

for the gain of a negative feedback amplifier to be defined by the feedbasck, the open loop gain must be much higher, > 10x at the frequency where you need the gain

you also need AC coupling/highpass or even bandpass filtering - room lighting will totally saturate the amplifier with large DC and low frequency signal

you may aslo need clamping diode limiters - signal will satrurate the ampifiers at short range and many op amps may be too slow to recover

there are good reasons dedicated chips are made for this use- look them up and read their data sheets/appnotes carefully for design info - even if you can't use the specific chip much practical information is in the manufacturers' literature, not textbooks

you probably need more time than you presently can imagine to gather the info, reread/relearn basic material, make and find all of the mistakes you're bound to make - dig in !

Users who are viewing this thread