Greeting everyone,
iam trying to use conventional buck converter (lm5117) as constant current source for variable resistive load (8-14 ohms). I managed to simulate constant voltage, but resistance of load can almost double so my current drops a lot. When I tried typical application scheme from datasheet (https://www.ti.com/lit/ds/symlink/l...l=https%3A%2F%2Fwww.ti.com%2Fproduct%2FLM5117) figure 35 analysis shows that nothing happens. It is very probable that i made novice mistake somewhere.

Attached:
Image 1 - Scheme
Image 2 - Transient analysis

Thanks in advance for any useful answers.
Mike

 

I might have reiterate little bit. I can do quite easily constant current source by using sense resistor after load and measuring voltage across, but I was trying to use metod mentioned in TI paper on more effective ways to do CC source.
Here is the link : https://www.ti.com/lit/an/snva829/s...55206&ref_url=https%3A%2F%2Fwww.google.com%2F

Is it possible that CM pin is not modelled ?

 

I think it would be easier to just use a Hall-Effect-Current-Sensor if
You are so concerned with Efficiency.
But, with the relatively light-Loads You mentioned I can't see it as being
worthwhile to eliminate a standard Current-Sense-Resistor.
.
.
.

 

Thank you! Pcb will have to be very small (probably less than 2 cm squared and 4*6mm is size of converter) so every unnecessary part had to be eliminated. Less parts also increases reliability and reduces cost. At last but not least circuit will be operated in between 70 and 90 deg Celsius, so any additional unnecessary heating is unwanted.

edit: yes I can use op-amp and smaller resistor for better efficency but ..... more parts :/

 

Make your life easy ................
.

 

Why don’t you find a buck regulator that is designed to be a constant current source? There are hundreds of them available for driving LEDs?

 

This one is also designed with interal analog current monitor. Iam just not sure if transient model accurately describes behavior of CM pin. In datasheet (link above) is equation which describes the relationship between R_sense current and output of CM pin but in analysis behaviour doesn't follow equation (img4,5).

 

This one is also designed with interal analog current monitor. Iam just not sure if transient model accurately describes behavior of CM pin. In datasheet (link above) is equation which describes the relationship between R_sense current and output of CM pin but in analysis behaviour doesn't follow equation (img4,5).

The correct circuit is shown in figure 35 of the datasheet. That’s not the circuit you have. You have the FB and CM connections to the voltage divider reversed.

 

Yes, thanks But that divider only reduces voltage and does not change rising tendency of voltage, it would also rise if divided only it would be smaller value, my problem is that CM is rising while voltage across R_sense is gettig lower. I am working on stability, set up wanted current should be quide easy afterwards.

At fist I tried excatly follow datasheet (img1) but it didnt work so i moved R_sense (img3) because I am not sure how I should measure current through energy dissipation from coil (through R_sesnse to ground) because it is on low side.

ps: img6 Reaction iam aiming for but this is with use of big R_sense like in old times.