Hi Folks,

Slightly related to my other thread about the pull-down resistor for an opamp… (All suggestions will be in v2)

I have a problem, which to me, is weird. See diagram attached. I realised I made a slight mistake of the diagram though, I’ll explain later.
When the shunt is connected (low side , common ground) and the ADC (Netduino) and OP powered by a separate power source (LiPo battery or USB) then all is well – The netduino is happy and current readings are as expected.

When I then hookup the same battery which supplies the load (12V Pb battery) to the voltage divider (up top on the diagram, sorry, no labels :|) and connect that to the netduino, I lose the current readings.
I am guessing I have some sort of common ground / feedback etc etc which is causing a problem. Having looked at the diagram I cannot see why so fresh out of ideas.

In short, I cannot measure voltage (12V) and low-side shunt voltage at the same time. The shunt values vanish.

Is there something else I should be doing? I’ve read about DC-DC “isolation convertors” (have one winging its way to me)

Thanks
Crispin

 

The feedback resistor connection on the op-amp looks wrong. The end connected to ground should feed back to the negative input.

 

The feedback resistor connection on the op-amp looks wrong. The end connected to ground should feed back to the negative input.

Doh - drawn wrong.
It is indeed the other way round, 38k from output to inverting input, 1k from inverting input to gnd.
The circuit is that way round, I drew the diagram after the fact.

 

...I drew the diagram after the fact.

DOH !!!! How can you wire a circuit that doesn't exist?

Note: you also show your MCU running off the "12" point. If that point is 12 volts, then you should have about 1.4V at the sense point to the MCU.

Your general scheme looks OK.

 

I never write things out first Always in my head and then pen them after. My day job is the same

Yes, the MCU is running of the same 12V, the 7805 is an assumed regulator which is on the MCU. Did not know how to draw that. I did comment it in my first post but it crashed.

So, MCU powered by 12 (same 12 as load) and 5V for OP is pulled from the netduino's 5V regulator

 

So this MCU must have an internal 5V regulator. Try disconnecting the signals from the MCU and see what a plain old voltmeter says they are. A scope would be better if you have one, it will show any oscillations.

 

Found the problem

It's the common ground. When I have only the shunt connected, things are ok. When I connect the voltage, the current drops. The reason is that the voltage is also supplied via the shunt sense cable. This causes a voltage drop across it which, well, kills my 5mV I am seeing on the shunt.

if I remove everything, the OP, the LED (forgot to put that in schematic ) then everything behaves as it should. The draw from the LED (simply a power-on LED across 12-0) is enough to cause a few mV drop.

So I think again.

Thanks for the nudge Ernie