TIDA-01480: Output capacitance and expected lifetime of ICs

Hello Paul, 

Please next time assign a responsible organization so we get to you sooner. 

1) output capacitor recommendation for this device is listed on the output cap selection section in the datasheet. where do you see the COUT in TIDA to be more than 500uF effective?

2) the variants should not affect the life time of the product. A in of 100khr info are provided on the site below 

https://www.ti.com/quality/docs/estimator.tsp?opn=TPS56C215RNNR&cpn=&partNumber=tps56c215#resultstable

https://www.ti.com/orm/home?actionId=2801

Thanks! 

Tahar

Hi Tahar,

What do you mean - 'assign a responsible organisation'?

I think you have misunderstood my questions

1) Looking at table 8.2 in the TPS56C215 datasheet - it states a max Cout of 500uF. The TIDA-01480 schematic shows a Cout of approx 200uF. As well as this, the load will be very capacitive due to the bypass/ decoupling capacitors on the Zynq MPSoC. In total this will be approx 1mF - I wanted to confirm that this is not a problem? Note: I am planning on combining rails 1A and 1B and powering both from the TPS56C215.

2)  I actually meant how long are these components expected to be in an active life cycle status ? Or in other words, are there any plans for either the TPS56C215, TPS65023B or TPS512000 to go end of life in the near future?

Look forward to hearing back on this.

Many Thanks,

Paul

Hi Paul,

1. TPS56C215 is compensated internally. There is a range of output capacitance it can support. The max output capacitance is 500uF. If customer has 1mF capacitance, the system will be unstable. You can run PSpice simulation to confirm it.

2. Regarding product longevity, our guidelines state that TI standard products are not eligible for obsolescence if any of these three conditions exist:

a. The product has any sales within the last five years (seven years for automotive or high-reliability products)

b. The product can be manufactured and has been in production for less than 10 years

c. The product has any current customer demand

Hope these helps,

Thanks,

Nancy

Hi Nancy,

I don't understand - this is an approved reference design for powering the Zynq MPSoC - so must have been tested with a Zynq MPSoC (along with it's decoupling capacitors) ? If what you say is true then the reference design does not work

Hi Paul,

The reference design schematic only shows 188uF for VCC rail. Did VCC rail need 1mF output capacitance? I am not familiar with Zynq MPSoc.

Thanks,

Nancy

Hi Nancy, Yes the Zynq has about 1000uF  of decouplers on it.

Hi Paul,

This reference design was designed in 2017. I am not sure about the history since I joined the team in April, 2023. TPS56C215 is D-CAP3 control with fast load transient response. The 500uF total output capacitance may be enough for the system to save cost. Just for your reference.

Thanks,

Nancy

Hi Nancy,

I am a bit worried - this is supposed to be an approved tested design - but it now looks like it is not capable of driving the Zynq MPSoC after all. Please could you confirm if the design has actually been tested with the MPSoC?

A spice simulation will not show loop stability - you would need to do a SIMPLIS simulation. 

Thanks,

Paul

Hi Paul,

From this reference design user's guide, it tested all the rails through a well-defined Samtec socket-terminal strip connection in Figure 13. 

For the PSpice simulation, I run load transient with 0.1A-6A with 1A/us and did see the double pulse on switching node to indicate loop possible unstable as below. It is risky to have 1000uF output cap for TPS56C215.

Hope this helps,

Nancy

Hi Nancy,

I appreciate your help, however am disappointed as the solution that is being promoted as an off-the-shelf tested solution for powering the Zynq MPSoC does not work. I think the app note should be updated or removed to avoid people designing in a solution that does not work.

Please could you recommend an alternative to the TPS56C215 that IS capable of powering the highly capacitive load?

Please could you confirm the TPS65023 is capable of driving its load? The highest capacitance is going to be about 300uF

Hi Paul,

For TPS56C215 alternative, I would like to commend TPS546B24A which can power the high capacitive load. Please check the datasheet.

For TPS65023, our team doesn't support this part, you may post a separate question with TPS65023. The expert will help you.

Thanks,

Nancy

Hi Paul,

For TPS56C215 alternative, please check TPS543A22/TPS548C26 which can support high capacitive load.

Hope this helps,

Thanks,

Nancy

Hi Paul,

Brian Berner is the author of TIDA-01480. Loop him here. 

Hi Brian,

Your application note used TPS56C215 for VCC rail of Zynq MPSoc. Zynq VCC has about 1000uF of decouplers on it according to Paul. However, TPS56C215 only can support max 500uF output capacitance. Could you please confirm if the design has actually been tested with the MPSoC?

Thanks,

Nancy

Hi Nancy,

Thanks for your help.

I would like to simulate the regulator to have confidence that it meets our requirements. Both the TPS543A22 & TPS548C26 only have SIMPLIS models available and will not simulate in Webench. Unfortunately, I do not have SIMPLIS so am unable to run the models. Does PSPICE-FOR-TI support SIMPLIS models ?

Thanks,

Paul

Or do you have a SPICE model ?

Hi Paul,

We don't have PSpice model for those two devices. But we do have PSpice model for TPS546B24A in ti.com. Is TPS546B24A overkilled?

Thanks,

Nancy

It is a bit overkill. I am currently designing a solution with the TPS548C26. The design calculator suggests it should be stable as long as I keep the output inductor small and implement a feed-forward capacitor - I will have to rely on this to show the design is stable

Hi Paul,

The calculation tool is pretty accurate. If you give me your application condition, I can run Simplis simulation for you.

Thanks,

Nancy

Hi Nancy,

That's really kind - Thank you.

Are you able to e-mail me your e-mail address, I would prefer not to put the design on a public forum ?

Thanks,

Paul