TPS61070 Startup and Reliability Issue

1.  If the FB pin is at 0.5V DC and the output voltage is at 3.5V DC then there is something wrong with your voltage divider from Vout to the FB pin.  Check your soldering and board layout.

2.  As well, this seems to be an issue with your soldering and board layout.

The device works fine when Vout is > Vin.  As a boost converter, it is not intended to operate when Vin > Vout, but it will still start.

Thanks for the reply Chris. 

I've checked soldering and board layout they are both OK. In this failure mode the device just isn't switching.

• Should I still get 5V out at say 4.9V in?
• How close can Vin be to Vout?

Please see this wiki page (FAQ): http://e2e.ti.com/support/power_management/non-isolated_dcdc/w/design_notes/870.aspx

If it's not switching, then it is either damaged, not soldered correctly, or you have applied too heavy of a load (or used an electronic load).  The D/S shows the highest load these devices can start into.

Hi Chris,

Thanks for the wiki page but my Vin (4.83V) is less than my set Vout (5.06V). My load is 100mA maximum which is only applied after the regulator is enabled and has a few hundred ms to stabilise first.

I have confirmed the device is not damaged and is soldered correctly. If I gently turn the input voltage down, the regulator spring in to life when I reach 4.72V. I can then increase the input voltage to  beyond 5V and the regulator continues to output ~5V. 

This behaviour is different to that described on the Wiki page which states "If the TPS61070 input voltage goes higher than the programmed output voltage, the IC will stop switching and the output voltage will track the input voltage, minus any voltage drops across the inductor and internal MOSFET"

It does look to me like a start-up issue at 'high' input voltage. What else could be going wrong here?

Thanks.

What are your power requirements that put Vins so close to your vout?

Could you attach waveforms of what you describe?  Vin, Vout, and SW when it is outputting too low and at startup?

Normally we run from batteries (2 x alkaline in series). The unit has the option to run from a 5V mains power adaptor (after this goes through a diode the 5V is actually 4.8V). The problems occur when running from mains derived power.

Pls see the plots below as requested:

Yellow = Input (at pin 6), Blue = Output (caps at pin 5), Magenta = SW (pin 1)

Outputting too low:

Decreasing Vin slightly and device starts switching:

Switching from start-up:

Thanks for the waveforms.  Since your third waveform shows a proper startup, I'm confused as to how you get the first waveform?  What is different in these test conditions?  Also, what is your load for each waveform?

A simple solution seems to be adding another diode in series with the input source to provide more drop.  Since you probably don't care about the losses when running from mains power, this should work.

Another option might be to use the TPS61220 instead.  This device has a different startup routine without the much reduced short circuit current limit of the TPS61070.  You might be somehow triggering this under your conditions.

Thanks Chris. There's nothing different in these test conditions other than the slightly differing Vin. The load is the same ~15mA, then it switches to ~100mA after a delay of around 300ms (you can see this as a slight Vout drop in the last waveform). The load is simply 5 LEDs in parallel.

I had concluded the same (adding a diode or one with a greater Vf) as a work around. I also have some TPS61220s on order as well.