WEBENCH® Tools/LMR14006: Error in webench/model

Additionally: for the new design, this would work with LMR14006X provided that ton_min is still 95ns.
However the datasheet only specifies ton_min for LMR14006Y: can you confirm that LMR14006X has the same ton_min = 95ns?

BR, GPEDomotica

1. For unstable issue, the webench design the inductor according to equation 1 on page 12, so if your Iout target is only 0.1A, the L will be very large, too large inductor will harm the phase margin, which may raise unstable risk. So the best way is to change the Iout to 600mA, they the design should be fine.
2. For Ton_min, LMR14006X have the same to_min=95ns. I'm not sure what is the problem you raised here. For 40Vin, 5Vout @ 2.1MHz, it certainly triggers ton_min.

Hi Luke,

thanks for the prompt reply.

1. True, but the strength of this product seems it has good light load efficiency. This means if I design for 600mA I can still use it reliably at 100mA for long time periods. I guess due to the low inductor value the current limit may activate the pulse skip mode but this should work without risks, correct? Then WEBENCH is probably not really optimized.

2a. For 40Vin, 5Vout @ 2.1MHz, it certainly triggers ton_min: correct, which means why WEBENCH does not recognize this in the first screenshot, where Vmax is 40V?

2b. LMR14006X have the same to_min=95ns: this means for 40Vin and 5.5Vout @1.29MHz ton_min > 110ns, which means LMR1400X is able to work fine. But WEBENCH does not propose this part!

BR, Vincenzo

vinvenzo,
1. Yes, you can design it for 600mA and use it for 100mA only. There is no risk of operating in pulse skip mode.
2a. The max value in first screenshot is 15V, so there is no problem..
2b. What's your input/output requirement? Maybe there is something wrong with the tool, we'll check with the team. You can find the device, open the design and input your requiremnet again to recreate your circuit.

Hi Luke!

1. Perfect, thanks!

2a. In the second screenshot Webench proposes a design with LMR14006Y and max Vin 40V. This won't work because of the min on time, as discussed before.

2b. My requirements are in the third (last) screenshot.

I will design pen and paper with LMR14006X because it is the best part to fit with my specs. I cannot use Webench because it does not find this part.

Vincenzo,
I'm pretty sure LMR14006X can meet such design requirement, not sure why the tool warns error. I create a design for 12V~36V-->5.5V/600mA with LMR14006X. The circuit should also apply to your requirement.
webench.ti.com/.../SDP.cgi

Hi GPEDomotica,
We are currently looking into this. We will get back to you.

Regards,
Gerold

Dear Luke,

yet another question about the pulse skip mode:

I would like to design input and output filters to comply with CISPR 22 B limits, e.g. following the classic approach described in http://www.ti.com/lit/an/snva489c/snva489c.pdf

How to consider conducted emissions due to pulse skip mode?

For example, if we design for 600mA with 30% ripple we will be in DCM for currents lower than 90mA.

For 30mA we will be in pulse-skip mode, and subharmonics of the main tone will appear in the input current spectrum for example.

My questions are:

1. Is there a guideline or app note to evaluate these subharmonics?
2. Can we say that if the converter complies with CISPR 22 B at max output power with lowest input voltage it will also be compliant with it when working in deep pulse skip mode?
3. Would it be better to use a part like LM5166 in COT mode to ensure EMC compliance?

Thank you for your help,

BR Vincenzo

Hi GPEDomotica,

Thanks for using WEBENCH and bringing this to our attention. Yes, you are right, for LMR16006Y with Vin_Max = 40V the WEBENCH design shouldn't have been created since it's violating the IC_Min_Ton as you had pointed out in your comment. We have fixed this issue and released the latest changes to production. Also, please note that LMR16006X and LMR16006Y are two different devics in terms of switching frequency and loop compensation. However, their minimum Ton is the same (as highlighted in their datasheet as well).

Please go ahead and try your above test conditions again and let us know if you have any question.

Thanks,
Hadi

Vincenzo, sorry for late reply. As the current pulse is small in pulse skip mode, the energy of the subharmnics is small. To my understanding, it's easier to pass the test in pulse skip mode even though the attenuation by the input filter is smaller at low frequency. Also in deep pulse skip mode, the frequency will decrease to lower than 150kHz, CISPR22 has no limitation for frequency lower than 150khz.