Opamp as a shunt voltage amplifier and supply of it.

[firewalker]

A common practice for using a small value shunt resistor (in power supply current control) is to use an opamp as a differential amplifier. If the gain is around 10 the out of the opamp is commonly 0 to 5 volts (in order to directly read it with micro).

The opamp used must be quite good in characteristics (for the given) job. In almost every design the opamp supply will be the same as the voltage input. If we have a supply of 30 volts, we must use an opamp that can stand 30 volts. Like LT1013. Using an opamp with good characteristics but with lower supply voltage is cheaper.

Since the max output voltage is close to 5 volts, can a lower secondary voltage (e.g. 7 volts) be used to supply the opamp. Does it cause problems?

Alexander.

[Siwastaja]

There are many hi-side current sense opamps that allow the common mode voltage on it's input pins outside it's own supply range. Using those is easiest. You just supply them with 5V (usually), and the shunt can sit at voltages higher than that, within some limits of course. Personally, I have used INA21x and LT1999.

These current sense opamps are typically easier and of better performance than just any random general purpose opamp. Outside of rails common mode range is one of the reasons. Fixed gain also helps PCB design by reducing part count.

[firewalker]

To answer my self... It will destroy the opamp. I just had a brainfart. Input common-mode voltage. 

Alexander.

[ajb]

With a standard opamp and four resistor diff amp, the resistors effectively divide down the input signal, so you can, potentially, use this to permit an input common mode range that extends outside the supply.  However this is not really compatible with the high gain typically required by a shunt amplifier.  It can be useful in other applications, though, where the inputs won't go far outside of the rails and little gain is required.

[Jeroen3]

You use INA199 or equivalent amplifiers.

[firewalker]

Yes I know about the application specific amplifiers like INA199 or MAX4080 and is what I would go for. The design I was referring is for a school lab. Non through hole components complicate things.

Alexander.

[macboy]

Look up "over the top" op-amps. They allow common mode voltage higher than the rails, often to several tens of volts or more. Dedicated shunt amplifiers also exist, and they probably contain a front end similar to the OTT op-amps.

[JS]

  Why not to sense the low side of the PS? Then you have a current shunt referred to GND (internal at one side, external on the other). If using a regulator IC you just need to tie the reference pin from the other side of the shunt to minimize voltage error. If using custom feedback this shouldn't be a problem at all.

JS

[JPortici]

depending on the application, it may not be wise to elevate the application "0V" potential from system ground

[Siwastaja]

Non through hole components complicate things.

They only do so when people explicitly decide, without any sensible reason, that they do so. Every time I point this out, there always are people who just keep repeating this mantra, without giving a chance to evaluate their choices and make a change. The life is totally doomed for such people anyway, and it's getting more and more difficult to do any circuits at all.

Just learn to solder. It's a must anyway.

Non-through hole components do not complicate anything; the utterly random decision not to use them does.

[JPortici]

if is it complicate in the sense that you can't put them in a breadboard... use adapters. if you have to make tons of them they are exstremely cheap
almost all the better stuff is smd anyway

[Siwastaja]

Simply don't use a breadboard to develop a power supply.

[JPortici]

.. or perfboard

[firewalker]

I have designed for the students (as part of them having a first impression of KiCAD) a pcb that can accept a plain MAX4080 or one on a adapter. I was looking for a full through hole solution for various other reasons not worth mentioning.

Thanks you guys!

Alexander.