TPS65251 current Limit

If you change the value of the Ilim resistors and the current limit does not change, then there is something else going on.  What inductor part number are you using?  Have you verified that its saturation current rating is higher than your current limit value?  Have you checked the current limit on the 12V input source to make sure that it is not current limiting?  Please send a screenshot of the output voltage, LXx node voltage, 12V voltage, and inductor current when one of the rails goes into current limit.

Hi Chris,

Please find my reply:

What inductor part number are you using?  SLF7055T-4R7N3R1

Have you verified that its saturation current rating is higher than your current limit value?  It is rated for 3.6A

Have you checked the current limit on the 12V input source to make sure that it is not current limiting?  Yes verified, it is not current limiting.

I am attaching the zip folder containing the waveforms as requested by you. I was unable to get inductor current, hence, I am attaching the waveform for voltage across the inductor.

I am also attaching the TPS65251 schematic.

Another, interesting point of note:

• We have tried using values of 100K, 120K, 150K, 160K for RLIM3 without much change in the cuurent lint (aprox. 1.4A), however if we change RLIM3 to 1M the current limit drops to 600mA
• We used a value of 47K (max current limit) for RLIM2, however the current limit remained aprox. at 950mA

the zip folder seems to have disappeared, i am attaching again

Thank you for sending your schematic and these waveforms.

To verify that this is an over current condition, we really need to see the inductor current.  This can be done by lifting one end of the inductor, adding a small length of wire in series with the inductor, and putting a current probe around this wire.

If this is an over-current event, zooming out on the LX2 node in the over-current condition to around 5 ms / div should show it.  Can you take a picture of this?

I would also recommend replacing your tantalum capacitors on the VINx pins with ceramics.  A high pulsating current is drawn from these caps and the tantalums have too much ESR to be effective at delivering this current.

Could you also post your board layout?

How are you applying and measuring the load?

We do not have a current probe available with us, is there any other method to view the inductor current waveform? Will the inductor core saturate, when we are using the same inductor as recommended in your EVM, which is rated for 3.6A?

Please find the layout attached.

We are connecting a 24Ohm variable rheostat, at the +ve terminals of capacitors C25 & C27 and GND terminal, with a digital multimeter connected in series.

Viewing the inductor current can only be done with a current probe and it is the only way to know for sure that you are having a over current event.  I don't think this is over-current, but is simply noise.

On the layout, I assume that layer 2 is a ground plane that ties all the grounds together?  Do you keep separate signal and power grounds?  One thing that is not good is how pins 30 and 31 are tied.  These two pins are analog ground and power ground.  They tie together and then tie back to the power pad.  They should tie back to the power pad (your single point ground) separately.  I would also highly recommend a decoupling cap on the Vin inputs.

Are all the capacitors on the board ceramics?

Can you look at the LX2 node when it is not regulating at around 5 ms / div?

Thanks Chris and why don't  you send one TPS65251 EVM to customer for checking comparison data. I can send EVM to him. Can I get shipping address?.

Regards YG Kim.

Please find my address:

Jasneet Dua

Director - Tech

Terron Microsystems Pvt Ltd

D-6/3, Okhla Industrial Area, Phase-1,

New Delhi - 110020

India

Ph: +91-11-40524822

Mob: +91-9811064753

We normally use double-sided PCBs, with bottom layer primarily comprising of GND plane. We try keeping the signal traces on the top layer.

Yes all the capacitors on the board are ceramic

I hope, when you say "not regulating"; you mean with feedback removed. I am attaching LX2 waveforms at no-load and with feedback removed.

Sorry, the i/p capacitors are tantulum capacitors

Hello Kim,

Am I being sent an EVM?

Jasneet.

Sorry delayed response. Still if you keep EVM, Send to me.

Address:

Texas Instruments.

13560 N,Central Expressway, MS3744.

Dallas, TX, 75243.

Thanks YG

Dear Kim,

Till date I have not received any EVM. As of now we are not working on TPS65251

rgds,

Jasneet

Jasneet.

I will ask again to our program manager of shipping one EVMs to you ASAP as below address.

Jasneet Dua

Director - Tech

Terron Microsystems Pvt Ltd

D-6/3, Okhla Industrial Area, Phase-1,

New Delhi - 110020

India

Ph: +91-11-40524822

Mob: +91-9811064753

Regard YG Kim.

Jasneet.

our program manager shipped one EVM  to you on Dec/01/2010. The tracking number is below. Please check and let me know.

Fedex: Tracking number – 8717-0567-6441

Regards YG

Dear YG,

We have received the EVM. However, we are not using TPS65251 in our design as of now.

regards,

Jasneet

Jasneet.

Did you received TPS65251 EVM?. and double conform pin25, LOW_P should be connect to GND.

Regards YG Kim.

Hi

Can I close it

Regards YG Kim

yes, because of delays in your response we have shifted our deisgn to another DC-DC converter