TPS61230: The boost converter when enabled does not output 5V

Hi Miguel,

We have devices such as TPS61022, which has higher start-up capability with load. You may have a try if available.

One possible solution is to add a pre-bias voltage at the TPS61230 output, for example, 3.3V. Higher pre-bias Vout will have higher current limit during start-up.

Miguel Rego said:

Can I use a load switch at the output of  TPS61230, so i can wait for it to startup and later enable the power for the other device ?

This is also a good solution.

Regards

Lei

Hi thanks,

Do you have any samples of TPS61022 that you can provide ? If not, do you think the load switch can solve this and should I take in consideration anything before going down this path ?  Meanwhile I will try your solution, but using  pre-bias voltage will probably increase the power consumption in my case.

thanks again!

Hi again,

I tested the pre-bias voltage and it worked! As a quick fix do you think I could generate this pre-bias voltage from a GPIO pin ? Also, what is the correct way to do this or what is usually done for this kind of situations ?

thanks again!

Hi Miguel,

Let me check.

Regards

Lei

Hi Miguel,

The only concern for pre-bias the Vout from a GPIO pin is whether the GPIO will be pulled down to a low voltage by the load in your application?

 If it's not a problem, then yes the GPIO pin + diode can be used to pre-bias the Vout.

Regards

Lei

Hi ,

Thanks! Do you think if I put a resistor + diode I should be safe ? Also, what is the correct way to do this, is to use an LDO or a delay circuit ? 

Thanks again Lei!

Hi Miguel,

This pre-bias solution depend heavily on the load at the TPS61230.

For example, what's the load type? Is it a resistive load or will it extract a constant current? How big is it? What's the threshold to active the load? What's the max Iout of the GPIO? 

In you trying that the pre-bias works, does the pre-bias voltage come from a power supply or GPIO?

Actually, I prefer the load switch solution. The control is simpler: just turn on the load switch when the TPS61230 finish the start-up.

Regards

Lei

Thanks again Lei for the help,

So, the max Iout for the GPIO is 15 mA. This GPIO as pre biased is just a quick fix so I dont have to print another PCB with the load switch solution for now. The load draws variable current, but at the start I think it needs 150 mA. 

When I tried, I used the power supply.

Yeah that probably the approach I will take in the final PCB. Thanks! 

Hi Lei,

I did some more tests. Seems that at lower voltage, like 3 volts the TPS61230 cannot keep with the load... I do the pre bias voltage and then enable the TPS61230, but the Vout voltage is 2.5 volts. Should I use another boost converter for this ? Or can I redesign what I currently have ? Here is the Power Designer link i used: https://webench.ti.com/appinfo/webench/scripts/SDP.cgi?ID=9393FD4B0CA1A4CE

Thank you!

Hi Miguel,

Seems strange that the TPS61230 can start-up if the pre-bias is lower, such as 2.5V. Again it's related with the load condition at the TPS61230 Vout, for example, above which threshold will the load be active? And how much current will the load extract?

Only with a clear understanding of the load conditions can we work out a feasible and stable solution.

Regards

Lei

Hi again Lei,

Im going to resume all the tests I did so we can take better conclusion:

• Load initially disconnected and Vin at 2.7 volts. The TPS61230 outputs 5 volts correctly. Once I connect the load, the output goes to 2.5 volts. What could be the issue here ? 
• Load initially disconnected and Vin at 3.6 volts. The TPS61230 outputs 5 volts correctly. Once I connect the load, the output stays at 5 volts. Everything ok.
• Load initially connected and Vin at 3.6 volts. Pre bias voltage from a GPIO pin with diode and 1k resistor. Once I enable the TPS61230, it outputs 5 volts and works correctly.
• Load initially connected and Vin at 2.7 volts. Pre bias voltage from a GPIO pin with diode and 1k resistor. Once I enable the TPS61230, it outputs 2.5.

The load Im using is a Arducam Mega 3MP. In the datasheet it says the current consumption from this device is 156 mA, but we should consider peaks of 400 mA. 

Thanks again Lei, hope this info make things more clear

Hi Miguel,

Miguel Rego said:

Load initially disconnected and Vin at 2.7 volts. The TPS61230 outputs 5 volts correctly. Once I connect the load, the output goes to 2.5 volts. What could be the issue here ? 

There should be a gap between the actual load current and the load you read from the datasheet. When Vin = 2.7V, Vout = 5V, and the load is added when the 5V Vout is on, the max Iout will be around 1.9A, much larger than the 400mA. 

You may measure the actual load current by a current probe.

Miguel Rego said:

Load initially disconnected and Vin at 3.6 volts. The TPS61230 outputs 5 volts correctly. Once I connect the load, the output stays at 5 volts. Everything ok.

When Vin = 3.6V, Vout = 5V, and the load is added when the 5V Vout is on, the max Iout will be around 2.5A

I suggest to focus on solving the issues of the above two conditions first.

Regards

Lei 

Hey Lei,

Thank again for the help! I think I probably found the issue. I`m using a diode, B2100-13-F, at the power supply input and I think the voltage drop gets bigger depending of the current being demanded by the circuit. Thus, seems that the input voltage at the TPS61230 goes lower than the 2.7 volts - it goes to 2.35 ish. If I bypass the  B2100-13-F, the TPS61230 works with a input voltage of 2.7 V - it still drops to 2.35 but can keep the the  TPS61230 Vout at 5 volts.
And the funny thing is that, without the diode, the TPS61230 can start with the load connected and there is no need to pre bias voltage. 

Could be that the diode is causing some kind of current limitation ? It shouldnt be of a big issue. 

Hi,

Thanks Aurora, but can you help me understand if possible what could be the root cause here, if you can of course ? 

thanks!

Hi Miguel,

A normal diode conducts with a voltage drop. I'm not sure what the actual voltage to the VIN pin of the device is at low input voltage, you can measure it. Pls make sure it doesn't fall below the minimum voltage of the device. Thanks.

Aurora

Hi again,

I know that, but please check what I said before : 
I think I probably found the issue. I`m using a diode, B2100-13-F, at the power supply input and I think the voltage drop gets bigger depending of the current being demanded by the circuit. Thus, seems that the input voltage at the TPS61230 goes lower than the 2.7 volts - it goes to 2.35 ish. If I bypass the  B2100-13-F, the TPS61230 works with a input voltage of 2.7 V - it still drops to 2.35 but can keep the the  TPS61230 Vout at 5 volts.
And the funny thing is that, without the diode, the TPS61230 can start with the load connected and there is no need to pre bias voltage. 

Could be that the diode is causing some kind of current limitation ? It shouldnt be of a big issue. 

Can you help me understand why ? 

Thank you