TPS54540-Q1: Can Not Deliver 5 A Current In Higher Swicthing Frequency (Fsw)

Armin Choopani

Part Number: TPS54540-Q1
Other Parts Discussed in Thread: TPS54540, , TPS54540B-Q1, LMR14050

Hi everyone

I am using a TPS54540 IC as the main buck converter in my system(to decrease 13 V to 5 V). Also, the system contains AM radio which its frequency band is 522 kHz- 1610 kHz. So, The TPS54540 switching frequency should be above 1.8 MHz. The TPS54540 delivers 5 V, 5 A properly when Fsw=1 MHz. But, when I increase the switching frequency (above 1.8 MHz), it can not deliver 5 A (it delivers 4A). As the current goes more than 4 A, large ripples appear on the output voltage which its frequency is under 10 kHz.

I wonder, does this IC can deliver 5 V, 5 A when the switching frequency is above 1.8 MHz without such ripples or not?

I appreciate any help in advance.

Armin

over 5 years ago

0 Katelyn Wiggenhorn over 5 years ago

TI__Genius 11040 points

Hello Armin,

I have forwarded your question to the product expert who will follow up with you by tomorrow.

Best Regards,
Katelyn

0 Armin Choopani over 5 years ago in reply to Katelyn Wiggenhorn

Prodigy 80 points

Hi

Thank You Katelyn.

BR
Armin

0 Robert Blattner over 5 years ago in reply to Armin Choopani

TI__Expert 6205 points

Hello Armin,

If you wish to proceed with the TPS54540-Q1 we will need to look at a couple of items:

1) How hot is the part getting? If it is above 65 C on your bench, then your design will likely run too hot in a sealed case with 85C ambient. More analysis would be needed.
2) Are you actually using the TPS54540B-Q1 device? The B version is more noise immune than the TPS54540-Q1.

If you are interested in a change, the LM14050-Q1 has better response both at high frequency and in dropout. The only difference in pinout is that COMP is replaced with a soft start pin.

Regards, Robert

0 Armin Choopani over 5 years ago in reply to Robert Blattner

Prodigy 80 points

Dear Robert

Thank you for your response

We prefer to use TPS54540-Q1. But other low-cost options are welcomed. I couldn't find good information about "LM14050-Q1". Can you give me its datasheet?.

First, on the IC, it has been written "54540Q, TI 7.5A, P14FG4", so I believe it is TPS54540-Q1 version.

Second, by using a thermal gun, the IC's top temperature is about 75 C @ 5 V, 4 A, 2 MHz after 10 min of operation. The total power consumed in the switching part is about 4.4 W (IC + schottky diode + inductor). By assuming Junction-to-top characterization parameter of 5.9 C/W, the junction temperature is 75+4.4*5.9=100 C which is well below maximum junction temperature (150 C). So, I think this cannot be the main problem, at least in the ambient temperature of 25 C.

For more detalied information, the schematic of the circuit is attached.

Best Regards

Armin

0 Armin Choopani over 5 years ago in reply to Robert Blattner

Prodigy 80 points

Dear Robert
I got some new information about the heat problem which I want to share with you. I set the buck converter at 2 MHz, 5 V to deliver 4.1 A (The maximum current it can deliver before instability (appearing low-frequency ripples on output node) happens). In this situation, I use a soldering system to increase the temperature to see whether instability happens or not. After heating up for 10 min, not only instability didn't happen, but also the buck converter was able to deliver 4.4 A without any problem. So I believe the thermal shutdown is not the problem.

Also, another interesting point is that when I increase the input voltage to 22 V, the buck converter was easily able to deliver 5 V, 5.2 A at 2 MHz which was weird. In this case, the top of the IC's temperature was 105 C!! When I decrease the input voltage to 8 V, it only delivers 3.5 A (5 V, 2 MHz) before going unstable.

Best Regards
Armin

0 Robert Blattner over 5 years ago in reply to Armin Choopani

TI__Expert 6205 points

Hello Armin,

The TPS54540-Q1 is more sensitive to noise than the TPS54540B-Q1. Since the part is not shutting down due to heat, noise is the most likely problem. Three approaches to reducing noise sensitivity are:

1) Use the TPS54540B-Q1.
2) Careful selection of the catch diode. The diode on the EVM is a good one.
3) Careful attention to layout. Note that this will also help with the high temperatures that you are seeing.

Since layout is typically proprietary, feel free to email me at robert.blattner@ti.com.

Regards, Robert

0 Armin Choopani over 5 years ago in reply to Robert Blattner

Prodigy 80 points

Hi
I sent you the layout.

BR
Armin

0 Robert Blattner over 5 years ago in reply to Armin Choopani

TI__Expert 6205 points

Hello Armin,

I have recieved your email and am looking at it.

Regards, Robert

0 Armin Choopani over 5 years ago in reply to Robert Blattner

Prodigy 80 points

Dear Robert

I have been waiting for your suggestions and help for roughly a week. This issue (current limitation) halts our whole project. I appreciate if you help me on this problem.
I am really confused. Why the IC can works properly when its input voltage is 22V (Output: 5V, 5.1A, fsw=1.87MHz) but not when the input is 13V (Output: 5V, 4.1A, fsw=1.87MHz)? Why by increasing the temperature the IC is able to deliver more current? and ...
If you need any further information regarding to the schematic and layout, I will be happy to provide it to you.

Best Regards
Armin

0 Robert Blattner over 5 years ago in reply to Armin Choopani

TI__Expert 6205 points

Hello Armin,

I will use email my today. Also, I found that earlier, I forgot the R in the alternate product's part number. It is the LMR14050.

Regards, Robert