WEBENCH® Tools/LM5175-Q1: The time it takes for the output to go down from 12V to 5.5V

Hi Anna,

You can export the WEBENCH design into PSpice / TINA-TI tool. Please make use of Voltage-controlled switch in the available in library and modify the circuit accordingly. The Voltage/resistor parameters of the switch can be defined as per your requirement.

Please refer the page no.218 in the below Pspice reference guide:

/cfs-file/__key/communityserver-discussions-components-files/234/PSpice_5F00_ReferenceguideOrCAD.pdf

Thanks & Regards,

Praveen

Hi Praveen, 

Thanks for your reply. I already exported the design into TINA-TI and the switch itself isn't the problem. When I try to change the load to a power dependent load, the simulation doesn't compile (Tina file attached). Can you tell why I get a compilation problem?

Thank you for your assistance, 

Annawebench_design_6501547_3.tsc

Hi Praveen, 

Thank you very much for this simulation. 

I will indeed change the input as you suggested - but the problem still remains the same.

The load resistance in the WEBENCH simulation model is a simple resistor. I need to simulate the time it takes the output to go from 12V to 5.5v, when the power remains pretty much the same after shutdown (65w).

This is because I need the total time constant in my circuit to hold the power for a few more mseconds after shutdown.

So, I need a constant power resistor (that will decrease with the decrease of the voltage). In your simulation you used a simple resistor as well. The problem is that when I try to replace the Rload with a constant Power resistor, the simulation doesn't converge or it provides a negative output voltage. 

By the way, I made a mistake with the part number, and the one that we purchased is LM5176-Q1.

I would really appreciate it if you could send me the simulation files (with LM5176-Q1) that works with a constant power load as we already explored all the options we know of. 

Anna

Hi Anna,

I have forwarded the query to product experts for further assistance. As this is a holiday season, the replies will be delayed.

Best Regards,

Bhushan 

Hi Praveen, 

I'm trying to implement your solution on a simulation file that I build according to our design. 

First I tried to simulate the output voltage with a constant resistor and I get a sudden voltage drop as you can see below. 

Can you identify the issue and why is happens? (Simulation file attached).

webench_design_03_01.TSC

I would be happy to explain to you about our application using your DC to DC via email, please let me know on which email address I can reach you guys. 

Thank you very much for your assistance. 

Anna

Hi Anna,

Could you please share the WEBENCH design for the TINA exported file using the 'Share Design' option?

Thanks & Regards,

Praveen

Hi Praveen, 

I modified the design after I exported it and added new macros for the transistors and the inductors and some of the capacitors in the circuit so it'll match in the best way our design. 

This is what I exported: 

https://webench.ti.com/appinfo/webench/scripts/SDP.cgi?ID=359359B5B685B435

Please note that this is only the template. I made afterwards all the modifications in TINA because I wasn't able to choose some of our design parts in the WEBENCH. 

 Best regards, 

Anna

Hi Anna,

Thanks for sharing the design. We will verify the design and get back to you on this.

Thanks & Regards,

Praveen

Hi Praveen, 

I found the mistake in the design, thanks a lot for your help, no need to verify it. 

We have a long slow start with capacitance of CSS=100n. 

Do you know how can I speed up the process and reach steady state faster in a transient simulation, without modifying the capacitors value? 

In this conditions I unfortunately can't reach steady state and then simulate the shutdown, it takes too much time. 

Thanks, 

Anna 

webench_design_good_v.tscHi Praveen, 

The simulation provides me some unrealistic results for the power consumption of the FETs and the inductor (Tina simulation file attached). 

I did a transient simulation for 35 msec. Can you indicate why the power consumption results are that large? 

In addition, do you have an evaluation board for the LM5176 DC to DC that we can order? 

Thanks, 

Anna

Hi Praveen,

I'm sorry for the multiple and confusing questions - we just found out that the large power consumption is due to a high Ciss of the MOSFETS that we chose. 

The WEBENCH suggests this model: CSD17304Q3, but it's not suitable for us because it doesn't have automotive qualification. 

Can you please suggest a compatible automotive MOSFET for the LM5176- Q1 DC to DC, similar to the WEBENCH suggestion, but in an automotive version? 

Does TI have an automotive option? I couldn't find one. 

My question on the evaluation board is still relevant, but the simulation assistance is not needed at this point, thank you.

We would appreciate your assistance as soon as possible, this is quiet urgent for us. 

Thank you very much, 

Anna

Hi Anna,

Currently, WEBENCH doesn't have an option to select automotive MOSFET. The queries related to the automotive MOSFET and EVM board will be answered by our product expert team. We will forward your queries to the expert team.

Thanks & Regards,
Praveen

Hello Anna,

Please find the link to the evaluation module in the product folder of LM5176 under "Design and Development": 

I understood the communication the way that you spoke about the LM5176-Q1, even if the headline shows LM5175-Q1.

Hi Anna,

Thanks for the inquiry. TI does not have any automotive qualified FETs and we have no plans to qualify any of our FETs to automotive standards. Sorry I cannot be of more help.

Hi Praveen, 

I saw that WEBENCH does suggests MOSFETS with similar capacitance as ours - so I don't understand why my simulation power output, when I import our automotive FETS, is very high and unrealistic (I Got 40W through one of the FETS) . When I tried the same simulation with WEBENCH suggested  MOSFETS I got a reasonable power dissipation through all 4 of them (2W maximum). Is it problematic to import a macro to a WEBENCH exported design?

The TINA simulation file is attached, could you please assist in identifying the problem? we don't understand if the MOSFETS we chose should be replaced or if the Tina simulation doesn't work with this imported models. 

Thank you! 

Anna6320.webench_design_good_v.tsc

Hi Anna,

The difference in power loss could be due to the below reasons. 

1. MOSFET characteristics: MOSFET power loss is dependent on many parameters like Ciss, Coss, Crss, Qg, Qrr, Rdson. The cumulative difference between the two MOSFET parameters could lead to high power loss. Could you please let us know how did you measure the FET power loss? Kindly calculate and compare the power loss using Excel FET Loss Estimator.

2. MOSFET TINA model: WEBENCH uses a generic template-based SPICE model for MOSFETs. While the key low-frequency characteristics like dc bias derating for Cap,  Rdson for FET etc are included in the model, the high-frequency behavior like ESL value is not very accurate in WEBENCH models. We recommend you download the vendor MOSFET SPICE models from their respective websites and perform simulations. Kindly let us know what is the source of the TINA model used for automotive MOSFET?

Could you please confirm if you are comparing the below MOSFETs?

Thanks & Regards,

Praveen B

Hi Praveen, 

1. I did a transient simulation of 35 msec. I measured the drain voltage and source voltage and the Id current with a probe, and then used this function: abs ((Vd-Vs)*Id). After that I did an average over 10 msec when the system reached its steady state, which produced the RMS value . This is what I did in both cases (the suggested MOSFETs and the ones we chose). 

 2. I downloaded the SPICE model directly from https://www.infineon.com/. They have some levels of the models there, I used the basic one: 

.SUBCKT IAUC100N08S5N043_L0  drain  gate  source

3. This is the MOSFETs WEBENCH suggested for us:

M1- CSD18514Q5A

M2- CSD18533Q5A

M3- CSD16415Q5

M4- CSD17573Q5B

Thank you, 

Anna

Hi Anna,

Thanks for the confirmation. We will get back to you on this.

Thanks & Regards,

Praveen

Hi Anna,

We do not think this is the correct way to measure the Mosfet power loss. But for further confirmation, our product expert will get back to you on this query.

Thanks & Regards,
Praveen