TPS650250 efficiency low.

Perry Frogge

using a 100 ohm resistor, I loaded the VDCDC3 and VLDO1 outputs ( 1.0v and 1.2V respectively). The VLDO measured 42% efficient, the VDCDC3 was only 58% efficient. Vin was 5V for VDCDC3, Vin=1.8v for LDO.

I have MODE pin pulled low with 4.7K. 58% is quite far from the +80% the data sheet curves show (at 1.3v). Inductor is a 2.2uH, 180mOHM, MLZ1608N2R2L. I reviewed your app note on measuring efficiency. Any ideas on what is going on? How can I improve the efficiency?

the other ones DCDC2 @ 1.8V was 65% eff, DCDC1 @3.3v was 90% eff.

Thanks

Perry

over 12 years ago

0 Michael Green over 12 years ago

TI__Genius 16870 points

Hi Perry,

For the LDO, the best case efficiency for Vout = 1.2V and Vin = 1.8V is about 66%. So, the 42% isn't too far off from expectation with the FET loses taken into account.

For the VDCDC3, 58% is a bit lower but, we should remember that you are comparing the Vout of 1.0V to 1.5V. For buck converters, as difference between Vout and Vin gets greater the expected efficiency lowers. This could explain the difference that you are seeing. Otherwise, how about you short MODE to GND and re-take the data. You can send me a scope shot of the SW node and MODE pin too so, I can verify the IC operation.

Also switching loses dominate at low currents so, the DCR of the inductor should not affect the efficiency too much and you should mind the core losses.

0 Perry Frogge over 12 years ago in reply to Michael Green

Intellectual 370 points

Thanks for feedback, will be next week before I get back in the lab. What do you mean by SW node? pin L3?

0 Michael Green over 12 years ago in reply to Perry Frogge

TI__Genius 16870 points

Hi Perry,

I am sorry for my terminology. Yes, you're right, the L3 pin. By looking at this node you can see how the converter is operating. It is typically the most informative node when debugging a switching power supply.

I will watch for your response.

0 Perry Frogge over 12 years ago in reply to Michael Green

Intellectual 370 points

I grounded the mode pin with no change, attempting to include scope shot of L3, 10x probe, AC coupled

Post

0 Perry Frogge over 12 years ago in reply to Perry Frogge

Intellectual 370 points

oops,

that was with no load on 1.0V output, here it is with 42mA load on 1.0v output. 2v per division, 1usec div

0 Perry Frogge over 12 years ago in reply to Perry Frogge

Intellectual 370 points

Bump to top, any feedback on waveforms? hardwire of pin did not change efficiency.

0 Michael Green over 12 years ago in reply to Perry Frogge

TI__Genius 16870 points

Perry,

As I mentioned before this is because you changed the Vout voltage. In general the more delta between Vin and Vout for buck converters the lower the efficiency. The graphs in the datasheet are not applicable since, the Vout voltage is different.

I measured the efficiency of the DCDC3 Vout = 1V and record similar efficiency performance. This is expected.

Power management

Power management forum

TPS650250 efficiency low.