TPS61094: Supercap not being charged due to capacitive load?

Bob van Haren

Part Number: TPS61094
Other Parts Discussed in Thread: LMR66420

Tool/software:

Hi,

We've got a problem with the TPS61094 in combination with a supercap and a capacitive load.
The TPS61094 is fed by a DC/DC buck converter (LMR66420) at 3.83 Vin.
The output voltage is selected as 3.6 V.
On the output of the TPS61094 are various things: an LDO linear regulator for an additional voltage rail (which in turn has an MCU and other peripherals) and a modem.

Upon powering up the board, the supercap does not charge and it seems as if the TPS61094 is just in force bypass mode:
Yellow: VIN, Blue: VOUT, Purple: VCAP

However, when I remove the two large capacitors (alu/poly 180uF each) from the modem, the supercap starts charging upon powerup (can also confirm by looking at the current drawn from the bench PSU):

I cannot toggle the MODE or EN pins on this board, as they are permanently tied to VIN without 0R resistor unfortunately.
We would prefer not to add components if not required, but we were thinking perhaps we need to toggle MODE/EN pins after a short delay, waiting for the output to stabilize first?

Any help is much appreciated!

6 months ago

0 Travis Wang 6 months ago

TI__Expert 8615 points

Hi Bob,

Please share schematic. I would like to know where you placed the 180uf.

Best Regards

Travis

0 Bob van Haren 6 months ago in reply to Travis Wang

Prodigy 10 points

Hi Travis,
Below are some important snippets:
Buck converter:

TPS61094:

Modem (caps C4 and C5 in the top left area are 180uF):

Below is part of the layout with some distances:

Thanks for having a look!

0 Travis Wang 6 months ago in reply to Bob van Haren

TI__Expert 8615 points

Hi Bob,

Thank you for your input. I'll try repeat this condition on bench.

Best Regards,

Travis

0 Travis Wang 6 months ago in reply to Travis Wang

TI__Expert 8615 points

Hi Bob,

I've repeated the phenomenon on bench. 3.83V is not high enough to let the IC enter charger mode. Please increase LMR66420 3V8 Vout to 4V, or reduce TPS61094 Vout to 3.4V

Best Regards,

Travis

0 Bob van Haren 6 months ago in reply to Travis Wang

Prodigy 10 points

Hi Travis,

Many thanks for the reply and for confirming this!

Essentially, you would recommend a minimum delta of 0.4 V then?
Could this be solved in another way, say by triggering MODE or EN pins slightly later?

With kind regards,
Bob

0 Travis Wang 6 months ago in reply to Bob van Haren

TI__Expert 8615 points

Hi Bob,

Yes, but changing Vout would be a more simple solution without changing the board.

Best Regards,

Travis

0 Bob van Haren 6 months ago in reply to Travis Wang

Prodigy 10 points

Hey Travis,

Thanks for that.

Luckily, this is a prototype, so I still have some time to make changes.

What I will do is the following:

On the Prototype boards:
- Change the Vout to 3.4 V (no changes to board required).
  Disadvantage here is the low margin to the lower input voltage limit of the modem (3.3 - 4.3 V) but for the proto this is acceptable.
For the new version of the boards, I will:
- Use Vin 3.9V, Vout 3.7V to have more margin. Use the PG output of the LMR66420 to toggle MODE+EN. I will add diodes to the MODE+EN pins (like on the EVM board) to use either the PG pin or the SUP+ pin to determine the state (meaning a temporary loss of input voltage will not change the mode of the TPS61094):

Would you agree that these changes have the desired effect?

Many thanks for the help so far,
Bob

0 Travis Wang 6 months ago in reply to Bob van Haren

TI__Expert 8615 points

Hi Bob,

I'm not sure about your system design, but I think TPS61094 part will work. I'll verify your new design on bench for further check.

Best Regards,

Travis

Power management

Power management forum

TPS61094: Supercap not being charged due to capacitive load?