Improving current limiting with LT3780,season 2


Now that' i've been using my LT3780 power supply for some times, some limitations are showing.
In particular :
* Cannot set max current to less than ~ 100 mA
* Resets if actual current is way above limit , when limit is low
* Linear pot, so painful to set low values

As far as requirements are concerned, i aim at max ~4A

First things first, let's have a closer look at the first stage of our current sensing measure :

It is a non inverting  opAmp amplifier , that we'll simplify here that way.


It is done on a breadboard using cheap LM358 from ebay (~1$ for 20)

Let's drive it from 0 to 14 mV, representing a current sensing from 0 to 2 A
The result is as follows, output voltage vs input voltage in mv:



We can see the  input voltage offset, it should be ~  2mV according to the datasheet.
It is only 0.74 mV here.

That graph  can be approximated as

VOut = (Iin*7mOhm-0.74mv)*153  (value <0 are grounded)

That can also be expressed  as :

VOut=IIn-113mV  (value <0 are grounded)

Gotcha, that's the ~ 100 mA minimum we were looking for.

So, if we add  an output offset of ~ 113mV, we are back in business.

Let's have a look at TI nice application note

We already have the multiplier part

* Rf= 1.5 MOhm
* Rg = 100kOhm

Let's assume R1 and R2 are very small compared to those
The offset becomes

B=5v*150*R2/R1=113.10-3

So 

R2=6500*R1

And we end up with a rough computation :

* R1 = 10 Ohm
* R2=  65 kOhm  (still small compared to Rf)

R2 will be an adjustable multiturn pot of 200k, to account for rounding errors and component accuracy.





Comments

Popular posts from this blog

Component tester with STM32 : Part 1 ADC, Resistor

Fixing the INA3221

INA3221, weird wiring