TPS62A02A: Shutdown at maximum load of 2A

Hi Rasathi,

Thanks for posting your query. I will get back to you as soon as possible.

Best Regards

Hi Rasathi,

Can you tell us more if the device is still working or it is not working anymore and is the above statement for both failing Vouts (1.8/1.1V)?

Also, when you try to use the device at 25'C and max load (2A); how long do you operate and see that it is still working fine?

In your next step, do you check the device at -40'C or at 85'C?

If you operate the device with this flow:

1. 1.8Vout
   1. 25'C
   2. -40'C
   3. 85'C
2. 1.1Vout
   1. 25'C
   2. -40'C
   3. 85'C

What do you see? Can you share this data with us?

Best Regards,

Zakir

Also,

If you tell us when does this shutdown happens, how quick is it? Does the device gives Vout (1.8V/1.1V) for sometime? or turns off quickly?

Best Regards

Please find the Vin voltage sagging for 1V8 buck (TPS62A02A) at 85degC, 2A. there is sw frequency switching to discontinuous conduction mode.

Hello Zakir,

Device is working. At 25degC, i used more than 5 mins to monitor, so that is no issue.

I go in this order only 25 --> -40 --> 85. only at 85 I am seeing this issue. 

I found out why buck is turning off at 85degC and 2A. There are 2 reasons.

1) Voltage drop in bode 100 lead:

During stability test, When connecting B-WIT-100 transformer output to the 10ohms injection resistor, the output is switching off. how much ever short lead i used, this issue persists. Maybe too much voltage drop is happening in the lead, but interesting point is, this happens only in 85degC. In other degrees (25 and -40) stability is passing.

2) SW_node shifting below 0 line:

after removing the complete bode100 setup, i tried to measure the Vin, SW_node, Iout. I have attached the image below this thread. In the image you can see that the input voltage is sagging to 2.9V from 3.3V. Also the SW_node is moved down a little below 0 line. I noticed this only for 1V8 buck. Please note that in 1V1 and 1V25 IC, sw_node is correct.

3) In 1V1 why it is turning off even though the sw_node is correct, is because of b-wit 100 transformer connection. (please check 1st point). In this input voltage sagging didnt happen. it is because of bode 100 lead, buck output turning off for 1V1 IC at 85degC and at 2A

It turns off quickly, as soon as the load is switched on.

One more thing. 

The input of 1V8, 1V1 and 1V25 buck is from another 3.3V buck LMR604403SRAKRQ1. In layout two output caps were placed on either side of the inductor as shown below. Could this be the reason? Because my colleague was parallely testing 3V3 buck, he found out the output was not properly regulating, so he brought all the output caps one side. Then it was properly regulating.

Hi Rasathi,

The waveform you shared above, is it for 3.3Vin/1.8Vout at 85 degree with 2A load? or is it for the 1.25Vout case?

The issue would not be due to bode 100 measurements until and unless the device works fine without bode setup and at 85'C with these Vin and Vout.

The turning off of the device could be due to UVLO triggering because you also mentioned the Vin is sagging. Can you clarify at what points you are probing in your circuit and what type of supply you are using, is it the output of any buck right now or any other supply? Also, from PCB layout it seems like the traces are not wide enough and the UVLO could most probably happen due to drop in the line. 

Can you share the waveforms (the transients), for the instant when you apply the load and probe at Vin, Vout, SW nodes. Please also mention the nodes on PCB layout where you are probing, to understand the problem better.

Regarding the passive placements, we recommend to place them as close as possible near the device to avoid parasitics. 

Best Regards,

Zakir

Hello Zakir, Please find my answers inline.

The waveform you shared above, is it for 3.3Vin/1.8Vout at 85 degree with 2A load? --> Yes, the waveform shared is for 3.3Vin/1.8Vout at 85deg and 2A load.

The issue would not be due to bode 100 measurements until and unless the device works fine without bode setup and at 85'C with these Vin and Vout        --> Yes, you are correct. The issue in 1V8 buck was because of Vin sagging, and not because of BODE 100 setup.

Can you clarify at what points you are probing in your circuit --> you are asking about probing points for stability? B-WIT 100 transformer connected across 10ohms resistor, CH1 from bode 100 to junction of R0237 and R0238, CH2 connected to 1.8V(Vout). 

what type of supply you are using, is it the output of any buck right now or any other supply? --> Yes output from 3V3 buck taken as Vin for 1.8V buck. For powering the 3V3 buck is, a Linear DC supply - GPR3510HD (35Vmax, 10Amax) 

Also, from PCB layout it seems like the traces are not wide enough and the UVLO could most probably happen due to drop in the line. we recommend to place them as close as possible near the device to avoid parasitics--> Is there a simulation file available to verify the drop in line and to simulate the parasitics before going for design update?

Can you share the waveforms (the transients), for the instant when you apply the load and probe at Vin, Vout, SW nodes. --> I will measure and share the waveform.

mention the nodes on PCB layout where you are probing, --> please find below the 1.8V buck layout,

Vin - taken at 3.3V buck output testpoint (3V3 block is in TOP), 1.8V Vout - taken at TP0245 testpoint (1V8 block is in BOTTOM), both ICs are in extreme end of the PCB. Length of the board is 99.34mm.

Hi Rasathi,

It would be nice and clear if you share the scope plots for the device when it turns on with 2A/85'C/1.8Vout. You mentioned before that the device turns off quickly as soon as you apply 2A load but from the waveforms you shared above, it is evident that the device runs for sometime and then turns off. Can you show us the waveforms for Vin (probing at the TP0037 and also at immediate Vin terminals of 1.8VIC) with current waveforms if possible for load and inductor currents and Vout too, for the cases when you turn on the device without load and then when the device turns off when load is applied.

The parasitics of your PCB layout could be checked from ANSYS software. Once you know these parasitics, you can feed them in the PSpice/Simples models to see the behaviour at different nodes

You can also confirm this line resistance if you try to operate the 1.8VIC without the first buck and with direct external supply to see if it is not UVLO which is turning off the device. If the input voltage is reaching the UVLO limits somehow then it would probably the reason of shutting down of the device.

Lastly, would appreciate to have zoomed view of the PCB layout to get a better view of the device and its connections (much better with PNG files) :) 

Best Regards,

Zakir