regarding eliminating noise in current measurement ciruit.

Hi friends,

I am working on a project solar charge controller using MPT612.

PWM frequency is 20khz.

I am facing problem with battery current measurement.
A resistor is connected in series with battery negative and drop across the resistor is taken for measuring current. This differential signal is connected to OP-AMP for amplification.
i have observed the NOISE on the OP_AMP input, same is amplified and hence i am not able to get correct measurement.
voltage spikes with pk-pk 230mv @ 30%PWM @ 20khz are observed.These spikes are increasing with increasing the Pulse width(i.e. 800mv @ 40%).

can somebody tell me how can i eliminate these voltage spikes, noise.
do i need to tune snubber circuit of the MOSFET?
I have attached the schematic of the BUCK-BOOST Converter.

Thanks and Regards,
Viswanath

If you're seeing transient spikes (which isn't "noise") during switching, then it may just be a layout issue. Is this circuit on a PCB, or a breadboard? Can you post the op amp schematic?

Also, have you verified with a high bandwidth current probe that those current spikes aren't actually real? They might be, because of those snubber networks.

Is your current sense resistor a low inductance type? wire wounds have large L in series that cause large spikes.
besides you have to filter the amplified current by a simple RC filter and proper values just near the ADC input and AGND.

Hi mtwieg and Mehrshad74,

Thanks for your replies.

The Buck-Boost converter is on the HEATSINK (MOSFETS and Shottkey input, output capacitors), The control circuit is on the PCB. I have attached the schematic of the current sense circuit.

Yes i am measuring with 100mhz oscilloscope.

The current sense resistor is wire wound type (10mohm). Yes i do have 2pole active filter at ADC input and i have tried with the input filter at the opamp input.

How do i calculate snubber values?

Thanks and Regards,
Viswanath

sorry attachment is missed in the previous reply.

please find the current measurement circuit in the attachment.

Regards,
Viswanath

yviswanathbe said:

The Buck-Boost converter is on the HEATSINK (MOSFETS and Shottkey input, output capacitors), The control circuit is on the PCB. I have attached the schematic of the current sense circuit.

Click to expand...

If the two circuits aren't on the same PCB then it's likely a layout issue, at least partially. You're probably creating ground loops, etc. Try using a diff amp circuit to connect to the current sense resistor, instead of a single ended amplifier. That should help reduce the effect of ground loops. Also on the power circuitry, try to lay it out to keep high di/dt loops (like the one made of Q1, D23, Q3, and C7) as small as possible. Add a smaller, high frequency bypass capacitors to the input.

Yes i am measuring with 100mhz oscilloscope.

Click to expand...

But are you using an actual current probe to test what the current in the sense resistor actually is?

The current sense resistor is wire wound type (10mohm). Yes i do have 2pole active filter at ADC input and i have tried with the input filter at the opamp input.

Click to expand...

You should make sure the resistor is noninductive (some wirewounds are, most aren't). If it's inductive then it will give spikes no matter what.

How do i calculate snubber values?

Click to expand...

For RC snubbers it's generally a trial and error process, since the values depend on the parasitic inductance in your layout. For now I would try to minimize the parasitic inductance and remove the snubbers, unless you absolutely need them.

Hi,

Sorry for the late reply.

Now i have almost eliminated the noise causes in the system.

But now i am facing another problem in ADC reading.

In no load condition the read values are accurate.when comes to load condition the PV voltage is not accurate.
In my circuit PV voltage sensing is with respective to Battery Ground. I have observed there is 3V difference to the actual PV Voltage when the load is ON. I am suspecting the 3V is dropped across the MOSFET Q2(Q2 will switch on after power on.i.e, Boost mode). Is my guess is correct? How do i rectify this problem.

Thanks and Regards,
Viswanath.