TPS63802: Power for Multi Wireless Protocols

Sid .

Part Number: TPS63802
Other Parts Discussed in Thread: , TPS631010

Hello TI Support Forum,

Some facts about the wearable product we are working on:

1) Product is powered by a standard 3V7 1000mAH Li-po / LiFePo rechargable battery. This represents an input voltage range of 3V to 4V2.

2) 3V3 power is required for 3 wireless modules, a 2.4GHz module, a 865MHz – 867MHz module and a 1.2GHz - 1.7GHz module, some LEDs, a 17ohm magnetic buzzer, an accelerometer, etc.

3) Worst case current required is below 600mA peak, the average current is much lesser.

4) When the product is in sleep mode, the current required is around 200-350uA average. In deep-sleep the current is less than 10uA.

5) The battery is charged with a the cheap and common TP4056 battery charger IC.

6) As this is a wearable, we are also space constrained on our 2 layer PCB. Also, our PCB is a 0.8mm thick 2 layer board only!

We have tested a prototype by powering it with the above mentioned battery and the standard AP2114 LDO. It works fine but we cannot utilize the full battery capacity.

We have been looking for a good buck-boost DC-DC converter that will take in 3 to 4V2 and output 3V3@ at least 600mA. We require the converter itself to consume as less current as possible on no-load while still switching since it will always be powered and never off throughout the products lifetime. After spending time on-line we have decided to use the TPS63802DLA. We have looked at the datasheet, ordered (it's still on it's way) the TPS63802EVM. It's relatively a new chip. We also had a look at the Webench design.

Please let us know of any important points that we should consider and bear in mind while designing / working with the TPS63802 as this will be a new design and we can still change it if need be.

We intend for the TPS63802 to directly power the application in place of the LDOs. Is this ok?

Based on the info given above, do you think the TPS63802 would be the right choice for an application such as ours. If not, why?!

We are open to alternate, better suited converter chip suggestions that you may recommend / suggest for this application.

Thanks in advance,

Sid

over 2 years ago

0 Tao Zhang over 2 years ago

TI__Mastermind 27440 points

Hi Sid,

thanks for considering TI'device.

about in place of the LDOs directly, please be very carefull with the switching noise and make sure they are acceptable for the load in next stage. as you know. output of LDO always much clreaner than converter.

for the device suggestions, may be TPS631010 is a more suitable one for your case. the datasheet attached and it will be released soon.

RegardsTPS631010.pdf

Tao

0 Sid . over 2 years ago in reply to Tao Zhang

Prodigy 10 points

Dear Tao,

Thank you for your reply and suggestions.

Yes we are aware of switching noise. We are also concerned about “transmit mask” at the time of certification.

Anyway, we have seen some posts, a few are here:

https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/918076/tps63802-need-assistance-to-identify-causes-for-tps63802dlar-ic-burn-out

https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/886827/tps63802-output-ripple-voltage-is-almost-2-vpp

https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/898922/tps63802-can-t-make-this-part-work

We do intend to apply learnings from:

https://www.ti.com/lit/an/slva773/slva773.pdf

and

https://www.ti.com/lit/an/slyt614/slyt614.pdf

to our PCB. Do you think our 2 layer, 0.8mm thick PCB would be ok for TPS63802 operation or must we absolutely have a 4 layer board (our design works fine on a 2 layer, cost is also very critical !!!)

Thanks,

Sid

0 Tao Zhang over 2 years ago in reply to Sid .

TI__Mastermind 27440 points

Hi Sid,

yes, these application notes is really good and will help to make a good PCB layout.

except a worse EMI performance , 2 layers is acceptable. but i am sorry that i did not have dates to share about the EMI difference between 2 and 4 layers.

i would like to take the 2layers but not 4 because i think the difference in EMI performance can be acceptabe in your case. low current and voltage.

Regards

Tao

0 Sid . over 1 year ago in reply to Tao Zhang

Prodigy 10 points

Dear Tao,

Please check both 2 layer PCB layouts:

Layout 1: https://we.tl/t-ZvZ4xeKjCD

Layout 2: https://we.tl/t-FRe62kS7hb

We've put vias around the inductor pads in "Layout 1" just like an RF PCB transmission line. Will this help with regards to EMI? Also, we are concerned that these vias this may increase the stray capacitance between each inductor pad and ground, will this matter towards the TPS63802 performance? Is this ok or would you suggest it's better with no vias like "Layout 2"?

Thanks and regards,

Sid

0 Tao Zhang over 1 year ago in reply to Sid .

TI__Mastermind 27440 points

Hi Sid,

seems links share blocked by TI. could you share some pictures about your layout?

Regards

Tao

0 Sid . over 1 year ago in reply to Tao Zhang

Prodigy 10 points

Layout 1:

Layout 2:

0 Tao Zhang over 1 year ago in reply to Sid .

TI__Mastermind 27440 points

Hi Sid,

i think both are acceptable.

Honestly, i am not sure which one will be better in EMI but TPS63802 will be ok with both layout.

one more thing let me confused is that why did not you connect the GND of cap directly to GND net, it is not a good idea to separate them, right?

Regards

Tao

0 Sid . over 1 year ago in reply to Tao Zhang

Prodigy 10 points

Dear Tao,

Top:

Bottom:

Yes you are correct it's not good to separate them. We have not separated them!. In both the layouts we sent earlier, they are connected. Both input and output caps are connected directly to ground on both layers. Here's a slightly modified version. We've shifted the Vout trace connecting to FB so you can clearly see the GND plane on both layers.

Please share your thoughts?

Thanks and regards,

Sid

0 Tao Zhang over 1 year ago in reply to Sid .

TI__Mastermind 27440 points

Hi Sid,

From my side, i think both layout are acceptable.

Regards

Tao

0 Tao Zhang over 1 year ago in reply to Sid .

TI__Mastermind 27440 points

Hi Sid,

Do you have any concern about your layout? i saw you clicked "This did NOT resolve my issue" button.

Regards

Tao

0 Sid . over 1 year ago in reply to Tao Zhang

Prodigy 10 points

Dear Tao,

Sorry about that, no concern yet. I just wanted to keep the thread from auto-closing.

Our PCBs are ready and both components and PCBs are current in shipping. Once they arrive, we intend to immediately get our EMS assembly vendor to assemble the boards for us. After we check that everything's functioning well we shall close this thread.

Thanks and regards,

Sid

0 Tao Zhang over 1 year ago in reply to Sid .

TI__Mastermind 27440 points

Hi Sid,

That is ok.

i thought you still have concern about your layout and need more discussion here.

my fault that try to close this thread without ley you know. ok, let us keep this thread open till everhting is fixed.

Regards

Tao

0 Tao Zhang over 1 year ago in reply to Sid .

TI__Mastermind 27440 points

Hi Sid,

i forgot we should close the E2E threads in time, so i am sorry i have to close this thread.

if you need more discussion, please open another one.

Regards

Tao

Because of the holidays, TI E2E™ design support forum responses will be delayed from Dec. 25 through Jan. 2. Thank you for your patience.

Power management

Power management forum

TPS63802: Power for Multi Wireless Protocols