Hello,

I made a device where I am using Texas Instruments LM2700 boost converter (2.7-4.2V to fixed 5V output, or in other words 18650 battery to 5V.) I have already made the PCB (according to the datasheet example) and assembled all the components. However, I face a really strange output voltage behavior. When there is no load, I get my expected 5V output, but when I load the LM2700 with the intended 1W load, my output voltage drops to 2.2V something volts... This should not happen, because this IC is advertised to supply up to 2.5A of current (headline of the datasheet), and my load is consuming just a bit more than 200 mA (at 5V). I had a feeling that the suggested 2.2 uH inductor might be a bit to small, changed it for 10 uH, 50 uH, 100uH and finally even 1 mH. But with 1mH inductor I was still getting only 4.05 volts at the output (inductor was orders of magnitude larger than the datasheet recommendation.) I dont have much experience in power electronics, and it seems I have reached my edge of knowledge here. Thank you in advance for any help and ideas provided.

 

Have you measured the input voltage when the output is under load?

 

Have you measured the input voltage when the output is under load?

Yes, the input voltage is raw 18650 batterie voltage (I measured 3.8V) under the load or not

 

The diode symbol you are using does not have a part number. What kind of diode is it? It should be a Schottky Diode for a minimal voltage drop from the switch to the output

 

I am using https://eu.mouser.com/ProductDetail/Panjit/GS1510FL_R1_00001?qs=sPbYRqrBIVniWZEAvLQ9DA== diodes
But like I said, under no load I am getting nice 5V output, so is there any chance that these diodes are causing such a huge voltage drop under load?

 

Does the inductor meet the requirements in the " INDUCTOR AND DIODE SELECTION " part of the LM2700 datasheet in respect to the saturation current rating ? The choice of inductor in switch mode power supplies is quite important. Just because it has the correct inductance does not mean that it will work.

Les.

 

Does the inductor meet the requirements in the " INDUCTOR AND DIODE SELECTION " part of the LM2700 datasheet in respect to the saturation current rating ? The choice of inductor in switch mode power supplies is quite important. Just because it has the correct inductance does not mean that it will work.

Les.

At first I was using this inductor: https://eu.mouser.com/ProductDetail/Coilcraft/1812FS-222JLC?qs=ZYnrCdKdyeefUssecCBPrQ== which has a saturation current of 1.8A (way above maximum switching current) but like I said, I was getting 2.2-2.3V output

 

I am using https://eu.mouser.com/ProductDetail/Panjit/GS1510FL_R1_00001?qs=sPbYRqrBIVniWZEAvLQ9DA== diodes
But like I said, under no load I am getting nice 5V output, so is there any chance that these diodes are causing such a huge voltage drop under load?

It has a forward voltage of 1.1 V and a maximum forward current of 1.5 A. I would definitely NOT choose this diode for a boost Converter. Also the more current you pull the more power you will waste in the diode. Wasting power is not something you wan to do in a boost converter

 

It has a forward voltage of 1.1 V and a maximum forward current of 1.5 A. I would definitely NOT choose this diode for a Boost Converter.

Could you suggest what should I be looking for in that case?

 

Could you suggest what should I be looking for in that case?

I already did. You should be looking for a Schottky Diode with a forward voltage drop of 0.2 - 0.3V and a current handling capability 2-3 amperes.
N.B. 0.3V * 2A = 600 mW. You want to keep the dissipation in the diode below that number.

ETA: I did an open loop simulation of your power stage and found the current ripple to be excessive (80% of average current) for loads drawing 500 mA and 1A. The voltage ripple is quite respectable at 0.2%

 

The diode was designed for 60hz power line applications. It will work in the audio range but not at 600khz. It takes more time for the diode to turn off than one cycle.

Never use a 1000V diode in a fast application. Look for a 50V diode. (or 35V) 1kv diodes have high voltage loss compared to a low voltage diode or a Schottky. (forward voltage)
Schottky Diode or a 25 to 35nS fist switching diode. I think much of the trouble is Reverse Recovery Time too slow.

 

Here is the simulation I did with revised requirements for current ripple and voltage ripple.
I(out)av = 200 mA with ripple less than 20%
V(out)av = 5V with ripple less than 1%


I can show the calculations for the inductor and the capacitor if you are interested. As is typical for boost converters you will notice that the inductor current is about 50% higher than the load current. this oversight can often lead to core saturation.