Pls help me find extremely small 1:1 voltage-tracking regulator

Maybe you could use an LM723 in SO-14 plus 2R and 1C ? ;-)

Best regards, Spehro Pefhany

A surplus error amplifier could be used to modify your presently-used circuit, by driving the summing point through an impedance much larger than that of the voltage divider that is already in place. This would only 'trim' the voltage.

Perhaps this might not use much more real estate.

RL

How about using a rail-to-rail opamp, in unity-gain buffer configuration? I'm sure you can find plenty with that kind of current capability.

Thanks. That's an excellent idea !

The only potential pitfall I see is that my power could be turned off (hence turning off the opamp) but the reference voltage (the circuit I am interfacing to) is still on. Then I will be applying a voltage much higher than Vdd to the opamp.

You can put series resistance in there to limit the current. One thing you should be careful of if you try this is that most op-amps will happily oscillate away when connected directly in a unity-gain configuration to typical power supply bypass capacitors, so you need to decouple the op-amp output from any serious capacitive loading (and make sure that your compensation doesn't allow it to overshoot too much).

Best regards, Spehro Pefhany

Yes, thanks for the advice...

I am rethinking this a bit due to the advice of you and Walter. The reason why I wanted my power supply to track the other power supply is that I am afraid of the I/O pin protection clamp diode conducting (on either side), blowing things up. But as I think about it, using a high-current rail-to-rail opamp leads me back to exactly the same original problem that I am trying to solve. I am just dealing with a protection diode in a different place. And on top of that I need to worry about output capacitance like you said.

As you can see I have about 0.3V voltage difference between the 2 power supplies. I guess I can just add 10k resistors to each pin of the interface. If the power supply differences turns out to be great enough to trip the protection diodes (in the normal case this should not happen), the protection resistors should be able to limit the current and saving me from blowing things up.

Is there going to be any problem if I interface two CMOS logic systems together with a constant Vcc difference of 0.3V ? I wish the design to be both durable and reliable. I think the durable part has been addressed by protection resistors, but I am not so sure about the reliable part. I had the pleasure (sarcastic) to deal with an alarm in my car where the control signals would get mixed up, such as the door lock control crossing over to the siren, when the not-so-well-designed zener diode protection circuit to the ULN2003 driver IC kicks in. And because of this lesson I don't want my own product to act crazy under any situation.

Thanks again to the advice.

Let me preface this by saying that I don't have a clear idea of what you are trying to do. Is this correct?

i) You have system A with a power supply that's "about" 5V.

ii) You are designing system B that has logic level connections to system A.

iii) You wish it to be reliable.

If that's correct, maybe you can just use level translation methods to deal with any mismatch.

Simplest way is to put some series resistance in the signal lines. I won't go into that, there are plusses (cost, mostly) and minuses (may violate the abs max input voltage rating, may cause unforseen effects, slows down signals etc.).

You could also use 74HC4049/50s (with dual Vcc connections- one Vcc for inputs- tied to the internal Vcc, and one Vcc for outputs- tied to the external Vcc) or other level translation chips to inexpensively do this. That part will tolerate inputs as high as 16V and has a reasonable input threshold for CMOS. There are also other specialized level-translation devices:

The 74LVC245 is one jellybean choice (accepts inputs as high as 5.5V)

Best regards, Spehro Pefhany

IIRC, even CMOS doesn't swing rail-to-rail, so you shouldn't have any trouble. The only time you would have a problem would be when one power supply goes away, but why not run your circuit off the Vcc from the other circuit, CMOS is pretty low power...

How about an op-amp with a transistor added, emitter follower style?

----------------------------------------------- Jim Adney snipped-for-privacy@vwtype3.org Madison, WI 53711 USA

-----------------------------------------------