TPS746: How to Estimate Soft-Start Time

Hi Itoh-san,

I checked the validation data and can not find more details on the soft-start function. I will have to get the design team involved.

HI John-san,

OK. Please check it with design team.

Regards,

Itoh

Hi Itoh-san,

The following steps should help to predict the rise time as a function of COUT for any LDO with internal Soft-start control :

1. Find the minimum of the following slew rates
   1. 0.03uA / 27pF à soft-start slew rate – internal charging current of 0.03uA into 27pF internal cap
   2. I_LIM / C_OUT à Fastest slew rate ever possible on VOUT
2. If the answer in step 1 is soft-start slew rate, rise time on VOUT = 0.65V / soft-start slew rate, where 0.65V is the error amp’s reference voltage.
3. If the answer in step 1 is (b), rise time on VOUT = (VOUT_OPTION * C_OUT ) / I_LIM

Where VOUT_OPTION is whichever output voltage you selected. 

I hope this answers your question.

Hi John-san,

Thanks for reply.

1. In the datasheet. reference voltage of error amplifier is not described. However, VFB is defined as 0.55V. Please double check.

2. Comparison between fastest output voltage slew rate and internal SS cap slew rate makes sense. However, I wonder why voltage divider value is not taken into account in your formula. Could you please elaborate on that?

Best regards,

Itoh

Hi Itoh-san,

Sorry, typo above, the VFB is the reference voltage and is 0.55. With this correction, you can use VOUT as 0.65V is steps 2 and 3.

Does this help to clarify your question? 

Hi John-san,

I got reference voltage is 0.55V. But my VOUT is not 0.65V. I don't really get what you mean. VFB=0.55V but you can use 0.55V in step2 and 3?

Also, please answer my second question. 

I wonder why voltage divider value is not taken into account in your formula. I don't understand why slew rate of internal SS cap and output cap can be compared directly. Could you please elaborate on that?

Regards,

Itoh

Hi John-san,

It makes sense, thanks!

Regards

Itoh

Hi John-san,

Sorry for bothering you again.

From your formula, if soft start cap slew rate is dominant, the SS time will be 0.55V / (0.03uA/27pF) = 495us.

However, the SS time on the datasheet scopeshot looks much shorter than the calculation (200-300us).

Could you please clarify?

Regards,

Itoh

Attaching image again

HI Itoh-san,

It does not appear that the internal SS capacitor is dominant here. This means the ramp is limited mostly dominated by both the current limit and the foldback current limit. This makes step 3 a little more complicated as you would need to break up the slew rate into two linear parts:

1. ILIM_FOLDBACK/COUT during the phase where 0<VOUT<VFOLDDBACK (VFOLDBACK is not provided in the EC Table, but can be estimated as ~1.5V below VOUT as shown below):

And 2. ILIM_BRICKWALL for VFOLDBACK<VOUT<VOUT_TARGET Where VOUT is the actual VOUT during the soft-start period and VOUT_TARGET is the target output voltage. 

This is difficult to calculate precisely and it might be easier to get an EVM and measure directly with the desired output capacitance.

I hope this helps to answer your question. 

Hi John-san,

Let's go back to step1.

(a) Soft start cap based slew rate is
0.03uA / 27pF = 1.1kV/s

(b) Let's say I_foldback = 0.7A.
In the datasheet scopeshot, COUT is 1uF.
I_foldback / COUT = 0.7A /1uF = 700 kV/s

The minimum one is (a).
Based on (a), SS time should be 0.55V / 1.1kV/s = 500us.
I don't understand why you think foldback current limit is dominant here.
I can see 200-300 us SS time from the datasheet scopeshot.

Could you please double check?

Regards,

Itoh