EEVblog Electronics Community Forum

Electronics => Projects, Designs, and Technical Stuff => Topic started by: MBrahim on November 23, 2024, 07:52:20 pm

Title: Need Lil help in my power supply project (Current sensing)
Post by: MBrahim on November 23, 2024, 07:52:20 pm
I'm designing a Bench power supply and can't get a say in current sensing trouble.
I'm confused between choosing big ohm shunt + low gain amp, or small value shunt with high gain amp.
it's maximum current is 10A
so I put a 6mohm and gain 50
Is it better to increase the shunt resistance to get more accurate measurements or it doesn't matter?
like 15mohm with 20G amplifier, if it's more accurate I would sacrifice the loss it doesn't matter.

All measured values would be the same Theoretically but I'm scared that the lower one would catch noise Easly (I mean before amplifying which will cause terrible misreading), or it doesn't matter?
I didn't really get that experience in such a high sensitivity staff so I'm asking.

Thanks in advance.
Title: Re: Need Lil help in my power supply project (Current sensing)
Post by: Harry_22 on November 24, 2024, 01:36:52 am
Hi!
What measurement accuracy do you need?
Title: Re: Need Lil help in my power supply project (Current sensing)
Post by: Doctorandus_P on November 24, 2024, 05:03:57 am
I'm confused between choosing big ohm shunt + low gain amp, or small value shunt with high gain amp.

It's a combination of the two.
Heating of the shunt, and resistance change because of that is a big error source. Lower resistance is less heating, but bigger resistor also results in less heating.

And a hot resistor combined with wiring and a cold amplifier will also result in thermocouple effects.

And the amplifier has it's own errors, from input offset and noise to gain error.

I assume that designing your own power supply is for educational reasons. Parts are pretty cheap (unless you buy really high accuracy / stability parts). I suggest you build some variants, and then make some measurements. Hands-on experience is a good way to learn. Theory and simulations are useful tools, but they are no substitute for real experience.

Title: Re: Need Lil help in my power supply project (Current sensing)
Post by: MBrahim on November 24, 2024, 05:32:57 pm
100mA is enough.
Title: Re: Need Lil help in my power supply project (Current sensing)
Post by: MBrahim on November 24, 2024, 05:34:33 pm
Oh, good Idea, I will definitely do many Variations seems smart.
Thanks so much.
Title: Re: Need Lil help in my power supply project (Current sensing)
Post by: Terry Bites on November 25, 2024, 04:21:49 pm
"Voltage burden!" Can cause an error in your voltage regulation. The make sure the voltage dropped across Rsense is not going to have an adverse effect on your specs.
You'd want to keep this to a minimum especially when low-side sensing is used.
Say you only dropped 1mV at full current. A sense amplfier needs to provide a gain in the 1000 range to get ino the 1v of control voltage range. An LM358 has a typical Vos of 3mV, 3mV*1000=3V of error!! Even a precision opamp with 0.1mV (OP07) is going to give you grief

The amp would need to have pretty good offset/ offset drift spec to maintain accuracy. A PSU with high currents can be a noisy enevoinment with switching hash and ripple in the power rials and ground. That can create havoc for a ground referenced opamp amplifier circuit. A single ended solution would be a very bad idea.

It also depends on where your regulator senses the output voltage and the output current.
Ideally the current sensing should happen upstream from the voltage sense node.