TPS63070: USB reset when PCB Board connected to USB port.

Hi Manimaran,

thank you for using our device!

Could you please share your schematic.

regards,

Moritz

Hi Moritz,

Thank you for your response, the device is good, I am not looking to switch to other device, looking forward to solve these issues. Please see the circuit below and the layout.

Hi Moritz any update on the above request.

Hi Manimaran,

can you measure the VIN, VOUT, input current, L1 and L2 of TPS63070 during start up of the RFPA.

regards,

Moritz

Thanks for your reply, will capture the requested waveform and share it.

Hi Moritz,

Sorry  I came back late.

Few correction on the initial statement, actual the board is drawing approximately 750mA. So a USB port 500mA will not support this, but if bypassed the TP63070 and supply directly to with 500mA current from USB no issue found. Not sure why the USB not resetting when bypassed.

Answering  your question, 

Vin = 4.85V at the input of TP63070.

Vout = 5.1V at the output of TP63070.

Current consumption is 750mA for the board.

Below is the current plot when Wifi is transmitting, around 750mA.

Two plots, plot (1) is when using TP63070 to power board including RFPA (USB supply to TP63070). and

plot (2) is when TP63070 is bypassed and supply directly from USB.

Plot 1, WiFi On. There is interval on the transmission when using TP63070, not sure is this caused by TP63070, but no interval when TP63070 is bypassed. See plot 2.

Plot 2, WiFi On, Powered from USB, TP63070 bypassed. No interval. 

Thank you.

Hi Manimaran,

how does the schematic look, if you bypass the TPS? My first assumption is, that you gain the additional cap C29,C30 to buffer those 750mA load transients.

Please calculate your DC current that actually loads your USB.

regards,

Moritz

Hi Moritz,

I am able to send you the full schematic. Can I send to you privately?, or can I make this discussion private.

Brgds 

Hi Manimaran,

Please send your schematic to bcs_request_fs@list.ti.com, and also if possible the PCB layout around the TPS63070. The layout you showed above only shows the top layer, and it seems that the GND connection is not optimal. Please also see the recommended PCB layout in the datasheet. It is very important to have a good GND connection from the device to the input/output capacitors.

Best regards,
Milos

Hi Milos,

Thanks, I will send you the files by today. I am using Altium, can I send you Altium project files or you prefer PDFs.

BRgds

Hi Manimaran,

You can send the Altium files.

Best regards,
Milos

HI Milos,

I have just send the design file (schematic and PCB files) to the email address. My email address that I send my email is from manimaran.kanesan@nuheara.com

Thank you.

Hi Manimaran,

I checked the PCB layout and I have a few remarks:

1. The output capacitors C31 and C32 should be as close as possible to the TPS63070, and the GND return from these capacitors to PGND of the TPS63070 should be as short as possible. Please take a look at the recommended layout in the datasheet, and also check these application notes on PCB layout:
   http://www.ti.com/lit/an/slva773/slva773.pdf
   http://www.ti.com/lit/an/slyt614/slyt614.pdf
   They are for buck/boost converters, but the same is valid for the buck-boost converter.
2. GND side of R20 should be as close as possible to the GND pin of the TPS63070
3. If you don't use the PG pin, you can remove R18 and leave PG pin floating or connect it to GND.

The PCB layout is critical for switch-mode converters, otherwise you might have oscillations or excessive currents. You can try running your board with a TPS63070 EVM to bypass the TPS63070 on your board. If you still trigger the USB protection, than your load is too high for USB.

Additionally, when you run your loads with the TPS63070, you will have somewhat higher input current because the converter is not ideal, the efficiency is 0.9-0.95. This difference might be also a reason why the USB overcurrent protection is triggered when the TPS63070 is used, versus when you bypass it.

Best regards,
Milos

Hi Milos,

Thank you for the comments, I will implement the changes accordingly. Can I send to you the new design for review?

Like to understand, if the eff is higher 0.95, the input current is high? why this is?, isn't that the input current is based on the load?

BRgds

Hi Manimaran,

Yes, please send the new design files when you finish them.

I was referring to the case when you have Vin = Vout. When using a converter, the input current will always be a bit higher then when not using a converter (when you bypass the converter), since you don't have converter losses. The difference is small, for example 5%, but this difference might be enough to trigger the USB protection.

Best regards,
Milos 

Hi Milos,

Thanks for the explanation, I understood the details. I am also purchasing the dev board.

I like close this ticket now, and should I send to the same email for the review, 

Thanks