TINA/Spice/LM5166: Unexpected high Input-current in simulation

Martin Hansel

Part Number: LM5166
Other Parts Discussed in Thread: TINA-TI, , TPS54260, TPS92515

Tool/software: TINA-TI or Spice Models

Hi,

when I try to simulate the "LM5166 Startup simulation" everything works like expected, but the measurement of the input-current is showing unrealistic high values !?

Regards

Martin

over 8 years ago

0 Nikhil Gupta (Product Folder Models) over 8 years ago

TI__Mastermind 39210 points

Martin,

That is the artifact of the behavioral model. Because of the ideal nature of blocks used in the modeling, spikes are observed at switching transitions. If you ignore the spikes at the switching transitions, the results should be reasonable.

0 Martin Hansel over 8 years ago in reply to Nikhil Gupta (Product Folder Models)

Intellectual 390 points

Hi Nikhil,

my problem are not spikes, but the average input-current drawn by the device.

For example:

Vout=1.5V, Iout=150mA gives me on the input-side:

Vin=9.8V, Iin=141mA !!!

Regards

MartinLM5166.TSC

0 Nikhil Gupta (Product Folder Models) over 8 years ago in reply to Martin Hansel

TI__Mastermind 39210 points

Martin,

You are looking at the current in between 2 input caps Cin and C2. You need to look at the current flowing into the Vin pin. I have attached the modified schematic with a new probe and the results. As mentioned before, please ignore the spikes if calculating the Average input current manually.

LM5166_NG.TSC

0 Martin Hansel over 8 years ago in reply to Nikhil Gupta (Product Folder Models)

Intellectual 390 points

Nikhil,

I think, C2 will average the current! Even if I roughly check the estimated input-current by "your" signal:

InputCur_Act is about 200mA, duty-cycle about 1/5, so I expect an average current of about 40mA, not 140mA...

btw: the same circuit is behaving well with other SMPS. Currently I'm working with a design based on a TPS54260 -- everything works as expected.

Regards

Martin

0 Martin Hansel over 8 years ago in reply to Martin Hansel

Intellectual 390 points

Hi Nikhil,
your "current-spikes" are the source of the problem!
As you can see, C2 is integrating your spikes, which are the main-contributors to the overall current.
I could reduce the effect, by putting a series-resistance of 1Ohm in the Vin-Line of the switcher.
But anyway: either the current-amplitude of your spikes (about 15A!) or the duration seems not to be realistic.
Can you change the model ?
Regards
Martin

0 Nikhil Gupta (Product Folder Models) over 8 years ago in reply to Martin Hansel

TI__Mastermind 39210 points

Hi Martin,

We will look into this, but it might take a while to release the model with the fix. What exactly are you trying to get out of the Iin measurement?

0 Martin Hansel over 8 years ago in reply to Nikhil Gupta (Product Folder Models)

Intellectual 390 points

Hi Nikhil,
Energy-Harvesting.
I'm trying to keep my Power-Source in his "MPP". For that I must know the input-current.
Regards
Martin

0 Nikhil Gupta (Product Folder Models) over 8 years ago in reply to Martin Hansel

TI__Mastermind 39210 points

Martin,

We have identified the root cause and are working on fixing it.

0 Martin Hansel over 8 years ago in reply to Nikhil Gupta (Product Folder Models)

Intellectual 390 points

Hi Nikhil,

Great!--What do you think - can I wait for it?

Regards

Martin

0 Nikhil Gupta (Product Folder Models) over 8 years ago in reply to Martin Hansel

TI__Mastermind 39210 points

Martin: We expect to have this released by next week.

0 Martin Hansel over 8 years ago in reply to Nikhil Gupta (Product Folder Models)

Intellectual 390 points

Hi Nikhil,

good to hear, so I will wait for the fix!

But: Yesterday I ran into the (same?) issue on a TPS92515 (spikes of 90A/50ns in DRN-Pin when switch turns on), with your sim "SLUSBZ6–APRIL 2016"

Will you have an eye on that, too ?

regards

Martin

0 Nikhil Gupta (Product Folder Models) over 8 years ago in reply to Martin Hansel

TI__Mastermind 39210 points

Martin,

We will look at TPS92515 and get back.

0 Nikhil Gupta (Product Folder Models) over 8 years ago in reply to Nikhil Gupta (Product Folder Models)

TI__Mastermind 39210 points

Martin,

We have released the LM5166 (ADJ) revision to the web. It should show up in 1-2 days in the product folder.

0 Martin Hansel over 8 years ago in reply to Nikhil Gupta (Product Folder Models)

Intellectual 390 points

Great!

Who will I see, that it is the "new" model (something like xxx_Rev.A) or will it be the same name?

Martin

0 Nikhil Gupta (Product Folder Models) over 8 years ago in reply to Martin Hansel

TI__Mastermind 39210 points

Martin,

Since there is already a Rev A, you will see Rev B. Also the date will change.

0 Somen Das over 8 years ago in reply to Martin Hansel

TI__Intellectual 1040 points

Hi Martin,

The latest PSpice model is now available on web after the Bug Fix.

You can download it from the below link.

www.ti.com/product/LM5166/toolssoftware

Thanks,

Somen

0 Martin Hansel over 8 years ago in reply to Somen Das

Intellectual 390 points

Hi Somen,

I was working with the circuit exported by the Web-Bench. Is the Export already using the new model ?

regards

Martin

0 Nikhil Gupta (Product Folder Models) over 8 years ago in reply to Martin Hansel

TI__Mastermind 39210 points

Martin,

We have not propagated this change to the WEBENCH to TINA-TI Export feature yet. This does require us to test out a few more things which might take some time. In the mean time, you can try to see if you can import the unencrypted PSpice model available in the product folder into TINA-TI instead.

Simulation, hardware & system design tools

Simulation, hardware & system design tools forum

TINA/Spice/LM5166: Unexpected high Input-current in simulation