Help Needed with Transistor Voltage Regulation

[HypnoToad]

I have a NAD receiver and want to switch out the old NJM2043 op amp with a newer model but the catch is it's getting + & - 17.5vdc and that's a bit close to the max of most modern ones. I want to drop the voltage down to around + & - 15vdc.

My question for someone more knowledgeable is, it has 24vdc going in and 17.5vdc coming out, does this work like a voltage regulator and can I adjust the value of the 3.3k resistors R629 / R631 up or down to get the desired voltage? If not what do I have to change to lower the output voltage?

 

[dr*audio]

I would add a 15V zener from the base of each transistor to ground. Then it will be a voltage regulator. The line side of the diode goes to the + side of the cap.

 

[Hamish119]

...does this work like a voltage regulator and can I adjust the value of the 3.3k resistors R629 / R631 up or down to get the desired voltage? If not what do I have to change to lower the output voltage?

It's a capacitance multiplier.

The output voltage is derived from the Op-Amp quiescent current and the consequent voltage drop across R628 & R630 (OHMS Law).
The Op-Amp voltage rails may NOT be stable.

If you replace the current Op-Amp with something that has a lower quiescent current, the Op-Amp voltage rails will rise!
Smoke may billow from the new Op-Amp!

Consider using Zener diodes of the appropriate value in parallel with C620 & C621.
Alternatively, be prepared to spend many hours of experimentation with OHMS Law and different values of R628 & R630 - you may lose some Op-Amps during experimentation.

Good Luck!

Example: +/-15V3 outputs.
**N.B. R628 - R631 values will require modification**

 

[ConradH]

I'm surprised the factory didn't use zeners there. Setting opamp voltages with dropping resistors only leaves too much to chance.

 

[HypnoToad]

Example: +/-15V3 outputs.
**N.B. R628 - R631 values will require modification**

Thanks for the replies, looks like adding some 16v Zeners is the best way to go, with regard to the resistors, how about using 10 ohms for R628 / R630 as isolating resistors and then lower R629 / R631 to around 500 ohms?

 

[BinaryMike]

Those values put zener current around 16mA and dissipation around 250mW, which is a bit high for 500mW zeners. I suggest using 1K at R629/R631 with 1N5246B zeners.

 

[HypnoToad]

Those values put zener current around 16mA and dissipation around 250mW, which is a bit high for 500mW zeners. I suggest using 1K at R629/R631 with 1N5246B zeners.

I was going to use 16V 1N4745 1 watt zeners that I can get locally, so do you still think I should go with 1k Mike? I appreciate yours and every ones help on this.

I'm surprised the factory didn't use zeners there. Setting opamp voltages with dropping resistors only leaves too much to chance.

The more I see inside this NAD amp the more I wonder about NAD's reputation.

 

[SaSi]

What kind of op-amp are you considering to use? The NE5532 is limited to -/+22V at the rails. Also, the higher the supply voltage, the lower the distortion (within limits, of course).

 

[HypnoToad]

What kind of op-amp are you considering to use? The NE5532 is limited to -/+22V at the rails. Also, the higher the supply voltage, the lower the distortion (within limits, of course).

LME49720 which has a recommended max of -/+ 17v, it seems easy to lower the voltage using the zeners to be on the safe side.

 

[BinaryMike]

I was going to use 16V 1N4745 1 watt zeners that I can get locally, so do you still think I should go with 1k Mike?

I couldn't lay hands on a reference to cite, but I believe you will find that lower bias current is needed in lower-power zener diodes to obtain equal zener impedance relative to higher-rated zeners. Certainly the datasheets for 1N5246B vs 1N4745A show a higher test current for similar Zz in the 1W diodes vs 500mW diodes. Would it matter? Probably not, because good op-amps have enormous power supply rejection figures. Just something to keep in mind for some future case where it might matter a lot.