LMR51606: TI Recommendation for +/- 12V converter

Matt Calvo

Part Number: LMR51606
Other Parts Discussed in Thread: LM5168-Q1, LM5168, LMR51610, LMR51450

Tool/software:

Hey DC/DC Team,

I have a customer working on a new power supply design and we were hoping to get guidance from your team to see if you had any implementation or device recommendations. Please see a summary of our requirements below!

We need to generate +12V and -12V at about 40mA from a nominal 24V input (14V to 32V range) (10V to 32V input range would be even better). This is for some analog circuitry, so it needs to be low noise inherently or by filtering.

In the past we have used isolated, dual output DC-DC converters with additional filtering. These tend to be expensive, and we don’t need the isolation.

We are considering using the LMR51606 to generate 12V and then an inverting charge pump to generate -12V.

Please let us know if your team has any advice or guidance on TI solutions or recommended TI parts we should be considering for this power supply.

Let us know if anything needs further clarification or if you have any follow-up questions!

-Matt

6 months ago

0 Richard Fu22 6 months ago

TI__Mastermind 31315 points

Hi Matt,

The LMR51606 should be suitable for your application for your buck. However, I'm not so sure about the inverting charge pump to generate the -12V. The reason for this is that I took a look in the charge pump portfolio and could not find any devices that support a 12V input voltage.

However, what you could also do with the LMR51606 is to utilize it in an inverting buck boost application,

https://www.ti.com/lit/an/snva856b/snva856b.pdf?ts=1742581568006&ref_url=https%253A%252F%252Fwww.google.com%252F

One thing we want to consider is that IBB's have higher power dissipation than the typical buck , but since you are running at 40mA, I think you will be ok with the thermals, though. Please follow the recommendations in the application note.

Hope this helps,

Richard

0 Matt Calvo 6 months ago in reply to Richard Fu22

TI__Genius 14755 points

Thanks for the reply and advice Richard!

I'll pass this along to my customer and will have them comment back on this thread with any follow-up questions they may have!

-Matt

0 Richard Fu22 6 months ago in reply to Matt Calvo

TI__Mastermind 31315 points

Hi Matt,

Sounds good. Thanks for promoting our device.

Cheers,

Richard

0 David Shroyer 6 months ago in reply to Richard Fu22

Prodigy 70 points

Hi Richard,

I saw some converters with Fly-buck capability on the TI website (LM5168-Q1). Do you have any info that compares using buck plus IBB vs fly-buck to generate dual supplies?

Thanks, David.

0 Richard Fu22 6 months ago in reply to David Shroyer

TI__Mastermind 31315 points

Hi David,

One thing that you have to consider with a flybuck converter is that there is a constriction of the duty cycle to make sure you do not hit the peak current limit or negative current limit.

The LM5168 does have adjustable frequency so you may have some leeway in the coupled inductor selection, but I would typically expect the IBB to have a bit more freedom in the design with the components you can choose. I would suspect you may need a large coupled inductor if implementing a fly-buck topology with the LM5168. It should also be noted that a fly-buck also has poorer load regulation since the isolated rail is generated off of the primary rail as a reference.

The app note should describe a procedure and you can decide afterwards between the trade-offs.

https://www.ti.com/lit/an/snva790a/snva790a.pdf?ts=1742845868003&ref_url=https%253A%252F%252Fwww.google.com%252F

Thanks

Richard

0 David Shroyer 5 months ago

Prodigy 70 points

Hi Richard,

I am attempting to simulate the LMR51606 in PSpice for TI. I am encountering some issues:

1) LMR51606 does not exist in the library. I can only find LMR51610.

2) When I add LMR51610 from the library to the schematic, it adds LMR51450_TRANS instead.

3) When I run the simulation I get this message: ERROR(ORPSIM-15461): Incorrect number of interface nodes for X_U1. (U1 is the LMR51450).

I installed PSpice for TI yesterday, so I assume it is latest version and the libraries are latest as well.

Any suggestions? Thanks,

David Shroyer

0 Richard Fu22 5 months ago in reply to David Shroyer

TI__Mastermind 31315 points

Hi David,

Please check the attached file model with the encrypted LMR51606 with the steady-state and start-up waveforms.

Let me know if that helps.

Cheers,

Richard

LMR51606YF_PSPICE_TRANS.zip

0 David Shroyer 5 months ago in reply to Richard Fu22

Prodigy 70 points

Hi Richard,

I got this working. Thanks.

Could you send me a model for the LMR51606X?

Is there a way to speed up the simulation? My fairly simple simulation takes many minutes to complete.

Thank you,

David Shroyer

0 Richard Fu22 5 months ago in reply to David Shroyer

TI__Mastermind 31315 points

Hi David,

Unfortunately we do not have a PSPICE model for the LMR51606X. Our closest is the LMR51610X, which I've attached below.

LMR51610_PSPICE_TRANS.zip

How long are your simulations? When I run the test profile from our model, it takes about 3-5 minutes.

Thanks,

Richard

0 David Shroyer 5 months ago in reply to Richard Fu22

Prodigy 70 points

Hi Richard,

This is what I was simulating. 4ms duration. Took maybe 20 min. Maybe more. I didn't time it.

I am trying to see if I can get +13 and -13V. Results look promising, but it is difficult to make adjustments to the design when it takes that long to simulate.

Thanks,

David

0 Richard Fu22 5 months ago in reply to David Shroyer

TI__Mastermind 31315 points

Hi David,

I see, were you simulating start-up or steady-state? In the model we provided, there are different simulation profiles for both steady-state and start-up, is it possible that your simulation profile combines the two?

Thanks,

Richard

Power management

Power management forum

LMR51606: TI Recommendation for +/- 12V converter