TPS65123 incorrect Output Voltages on VGH and VGL

Hi Mike,

Can you tell me the resistor- values of VMAIN, VGH and VGL?

Thanks,

Ilona

Ilona

VMAIN = 147K TOP, 100K BOT

VGH = 1.54M TOP, 100K BOT

VGL = 232K TOP, 100K BOT

As mentioned previously, I re-configured to match one of TI's reference designs to see if it was a resistor issue.  I set VMAIN to 5V, VGL to -12V, and VGH to 12V, I changed all TOPs to 887K, VGL bottom to 365K, Vmain bottom to 287K, and VGH bottom stayed at 100k.

See below circuit

Thanks,

Mike

Hi Mike,

Try this: Pull R146 not to VMAIN but to VGH. Does that help?

Regards,
Ilona

Hi Ilona,

No go.  I got the same thing on VGL and VGH when the bottom side of R146 was tied to VGH instead.  Also tried that tied to ground instead of Vmain or VGH, but that didn't show any results either. 

Thanks

Mike

Hi Mike,

I am working on this.

Regards,
ilona

Ilona,

Any update with this? I am having a similar issue with the same part.

Thanks,
David

Hi Mike,

Did you record the start-up behavior of all outputs? Can you share a scope of VBOOT+VMAIN+VGH+VGL at the start-up ?
Maybe the precharge phase (precharge BOOT cap to VIN) does not power up the device because of voltage drops on VIN. You have to consider the start-up sequence, first comes VMAIN then VGL and then VGH.

Thanks,
Best Regards,
Ilona

In the first image above waveform A is VBOOT, B is VMAIN, C is VGH, D is VGL, and 1 is VIN.  The images are broken up so I apologize that the data has to be looked at in pieces.

It looks like the power up sequence is going on before Vin is actually stable.  I am assuming I need to control the RUN and EN pins and toggle them once the power supply is stable?

Ilona could you please comment on the screen shots posted above?

Thanks,
Dave

Hi David, Hi Mike,

I assume you both are working on the same schematic?!
I am a bit confused how VGL (VGH) behaves. How is it possible that before even VIN starts VGL has already 500mV? Additionally there is no way that VGL can jump that fast up to ~VMAIN except the path via Resistances (R146+R147~340kOhm) and the output cap (100nF), which would charge VGL at a time constant of ~34ms. But in this case it jumps as fast as VMAIN <100us.
Is VMAIN/VGL/VGH terminated with some load? If yes, does that start-up behavior appear when the load is not connected?

The sequence is actually correct, VMAIN starts to ramp if EN=~1.21V, and in your case EN=VIN, which is ok. Same applies to VBOOT and the "Enable"-Pin RUN. Is it possible to zoom in at the ramp-area of VMAIN-VGL-VGH? How does the Switching- Node SWN behave, it might be helpful if you could at this plot as well.

Thank you,
Best Regards,
Ilona

Ilona,

Yes we are working off the same schematic.  I took some more images below.  Just as a note the previous images were taken with a TFT LCD attached while the ones below have no load at all.  There is a large operational difference in regards to VGH with and without the LCD attached.

A = VMAIN

B = VGL

1 = VGH

No load attached and all signals are triggered at power up of system

A = SWN

1 = SWP

No load attached please note that both images were taken at different amplitude scales.

Hey guys,

Ok let me check this in the lab.
Regards

Ilona,

Very helpful! That did the trick. The decal for the diode was incorrect as I thought a different part was installed that had a reversed polarity. Swapped the part around and the whole circuit worked perfectly. Thanks again for the help!
Dave

Great to hear!
Ilona