LM5165XQ switching around 1.25V instead of 5V

Hi Peter,

What is your Vin Voltage?

Regards,

Gautam

Vin will be 24V. I have my prototype devices connected to a power supply that I've varied down to 7V. I can see the voltage ripple decrease at lower voltages.

Hi Peter,

It is possible that you may have recieved an adjustable output version of the part. Do you have provisions to add a resistor divider externally to set up 5V or could you place another device that may be configured correctly? Either attempts should confirm the case.

Regards,

Gautam

The switching regulator works with the feedback resistors! I followed the example schematic on figure 32 and equation 5 from the datasheet. I used RFB1=10KΩ and RFB2=3.3KΩ. This causes Vout to be 5V. 5V is suppose to be the fixed output for LM5165XQRCTQ1. The functional block diagram in section 7.2 of the datasheet has a note that says the feedback resistors are implemented in the fixed output versions. 

I will order the non-automotive version of the fixed 5V regulator and put it on my PCB. It seems there is a problem with you fab or a problem with you chip markings for the automotive grade version. I am still interested in automotive grade. Will you fix this soon?

I can supply pictures of the device marking and Digi-Key package to prove I ordered the correct part if you don't believe me.

Hi Peter,

Were all 20 parts behaving the same way or was this a one-of? We are working to get this straightened out,

Regards,

Gautam

The problem is present on two of my prototypes that I have inspected. I have not checked the rest. I only have 10 prototypes. The other 10 regulators are extras that the contract manufacturer ordered. The unused 10 are still in the reel tape packaging.

I inspected a third prototype and it produces the same 1.25V. I assume the other 7 prototypes also produce the same 1.25V. To be clear, 3 out of 3 devices function incorrectly; 100% failure rate. (I hope I didn't give the impression that only 2/10 were faulty in the previous post).

I have ordered the non-automotive version of the 5V fixed regulator, LM5165X, and the 3.3V fixed regulator, LM5165Y. I will let you know how those function when they arrive in 2 days.

He is the packaging the regulators arrived in:

Hi Peter,

The frequency setting resistor R23, currently 100 Ohms, is incorrect. Based on resistor R22 in series with the output capacitor, this is a configured as a COT converter. Then the resistor at the RT pin sets the PWM on time (and thus the switching frequency).

Connecting RT to GND with a short or low resistance configures the LM5165 in PFM mode (in which case R22 should be 0 Ohms).

Perhaps you can use the Excel quickstart design tool available in the LM5165 product folder to determine the correct RT resistance. Alternatively, let me know the desired switching frequnecy and I will provide a recommended design for you.

Regards,

Tim

erreur on schematic : real value for the IC is LM5165YQ

Hi Timothy,

I've notice the incorrect value for R23 as well, but I do not think the switching frequency is related to the topic at hand. The main focus is the automotive grade fixed 5 volt version of the LM5165X does not produce fixed 5 volts.

The non automotive grade version in the exact same circuit produces 5V. I used hot air to rip out the LM5165XQ and placed in a LM5165X with no other edits. This produced 5V.

The mistaken value for R23 comes from equation 2 in the datasheet. I didn't notice that the uints were KΩ instead of Ω so I was off by a factor of 1,000. Shorting R23 to ground for PFM mode and changing R23 to 68K for a frequency of 460KHz had no impact on the output voltage.

I used the exel sheet you mentioned to select R22.

Your offer to help with frequency is appreciated, but I would rather you help with understanding what's wrong with the voltage.

JP.Viaut, would you mind using equation 5 on the datasheet (page 17) to select feedback resistors? Rough numbers would be RFB2=56K and RFB1=100K. You can rip up a few traces to create the circuit shown in figure 32 B or figure 31 B. If this works for you, then the 3.3V automotive version of the part is broken too. You can disregard RUV2 and RHYS. Also try ordering the LM5165 Y (non automotive version). You can drop that IC into the footprint and it should work.

Recommendation is to return the units via RMA to the original distributor (for example, Digikey). TI will get notified, allowing free replacement samples to be shipped out.

Regards,

Tim

The units were part of a turnkey order from a contract manufacturer. The ICs are already mounted on PCBs. Your suggestion is not the easiest thing to do... I have already thrown away 2. 

I only ordered 20. Could you send me that many samples without an intensive process? 

Hello

the 4 unit are out and we sold 4 new LM5165YQ , it's now ok, we can read 3.28 Volts at the output.

BUT WE CAN'T RETURN OLD IC 

This story make us lost 2 weeks for our development and lot of questions about real quality chain of your products.

What happen is the same thing arrive during production with 1000 or 5000 pcs ? Who is paying for all the damages ?

My company want to have warranty and ask me to search another way is you can't engage about this warranty.

Regards

JP.Viaut

"it's now ok"

What does that mean? How did it get to OK status? What did you change to get the 3.28V?

It's not a problem for your distributor but a factory problem !!!!

wrong reference on a chip make your company responsible for all problem of your customers.

if you don't want to take this in consideration we will search other way for factory. You are not alone on the market.

We can understand a mistake, but we can't understand we don't want to repair your mistake and the most important 

you have to give warranty for future delivery and bulk factory order. Without these warranty my boss want i search other product in other company.

Waiting your reply.