LM10504: Force PWM switching waveform

Hi,

I have assigned your request to responsible Applications Engineer and we will get back to you as soon as possible.

Regards,

Murthy

Please share as much information as possible on the e2e forum. Otherwise, this thread will not be valuable to the rest of the community.

Hello Brian-san,

The switch node of waveform is like this. It seems not PWM.

The waveform of SPI write to force PWM is below. I think the write is valid.

It seems the fPWM mode seems valid but the switching waveform seems not PWM. I want to know the possible reason.

Best regards,

Toshihiro Watanabe

Watanabe-san,

This issue has been tested and resolved on the LM10506 in this e2e thread: e2e.ti.com/.../661885
The LM10504 is very similar to the LM10506.

In this other thread, the SPI Write was also assumed to be valid, but in reality the SPI Write was invalid and modifying the SW/FW of the MCU showed that the LM1050x device was behaving correctly when the SPI Write code was modified.

I believe this is the same problem your customer is experiencing.

Hello,
Thank you for your support.
If inductor changed from 1.5uH to 2.2uH, the waveform became PWM mode in light load.
There are some questions.
1) Is it thinkable if the switching waveform becomes posted before when inductor value is 1.5uH?
2) Only Buck1 inductor was changed but all Buck1 to Buck3 waveform seem to become PWM mode.
If Buck1 waveform isn't PWM waveform, is there possibility to affect it to Buck2 and Buck3?



[The default configuration]
Inductor value are 1.5uH on Buck1 to Buck3
Capacitor value are 22uH on Buck1 to Buck3
Buck1 is light load like PFM mode if not fPWM mode.
Buck2 and Buck3 load are not light like PWM mode even if not fPWM mode.
Buck1 ripple voltae is bigger,
Ripple could see on Buck2 and Buck3 which is not so big as Buck1.


Best regards,
Toshihiro Watanabe

Hello,
The customer changes only Buck1 inductor value but it affects all buck operation. I think this is wierd but is this kind of thing possibly happen?

I have not heard of this behavior before, but I can test on the EVM.

It may take 1-2 days to get the time in the lab to examine this.

In this most recent reply you said "customer only changes Buck1 inductor"

but in previous email it says:
"[The default configuration]
Inductor value are 1.5uH on Buck1 to Buck3
Capacitor value are 22uH on Buck1 to Buck3"

This makes it sound like all Inductor values are changed to 1.5uH and all capacitor values are changed to 22uH.

Please confirm which statement is true.

Please confirm exact part # of all Inductors (L1-L3) and output Capacitors (C1-C3) on the customer's PCB.

I will need to look at the detailed specs of the changes the customer made to passive components.

In my testing so far, fPWM mode has worked correctly at no load on both LM10506 and LM10504 devices. The defaults L's & C's are still installed on the LM10504EVM (L1-L3 = 2.2uH, C1-C3 = 10uF).

I will look into the need to modifying the EVM PCB after I receive your reply.

Hello Brian-san,
The Inductors are DFE201610E-1R5M=P2. The capacitors are C2012JB1A226MT. All buck1 to buck3 are the same parts used.
Best regards,
Toshihiro Watanabe

Thanks you for providing this info.

I will need to check if we have equivalent inductors in the lab to put on an EVM and measure the effects.

These are Murata multi-layer power inductors. 

The closest inductor available in the lab here at TI is the MLP2520V1R5M (TDK 1.5uH multi-layer inductor rated for 1.4A)

I tested the LM10504 EVM with the default VLF3010ST-2R2M1R1 2.2uH inductor, then swapped in the 1.5uH inductor at L1 and saw very little change in Force PWM performance at no load.

The following scope shots were taken with:

• Ch1 = BUCK1_S (DC output of Buck 1, Sense line)
• Ch2 = BUCK2_S (DC output of Buck 2, Sense line)
• Ch3 = BUCK3_S (DC output of Buck 3, Sense line)
• Ch4 = Measure at switching node of L1 (close to LM10504 IC)

Image 1: L1 = 2.2uH Force PWM Off

Image 2: L1 = 2.2uH Force PWM On

Image 3: L1 = 1.5uH Force PWM Off (timescale changed to 1us/div)

Image 4: L1 = 1.5uH Force PWM On

My conclusion is that the value of L1 (1.5uH or 2.2uH) does not impact the functionality of entering Force PWM mode.

Hello Brian-san,
Thank you for your measurement.
I think your test is buck1:buck2,buck3=2.2uH:2.2uH:2.2uH -> buck1:buck2,buck3=1.5uH:2.2uH:2.2uH.
The issue observed when buck1:buck2,buck3=1.5uH:1.5uH:1.5uH.
Suggested inductor value will be 2.2uH but I am wondering if LM10504 is in something like this condition when lower inductor value used.
If all 1.5uH and no issue, we couldn't reproduce the same phenomenon and suggest to use 2.2uH inductor if it is possible.
Best regards,
Toshihiro Watanabe

Watanabe-san,

Your understanding of my test setup is correct.

Murata has 2.2-uH inductor in same family, with part # DFE201610E-2R2M=P2 (www.digikey.com/.../8539966)

It would be best comparison if customer test the "buck1:buck2,buck3=2.2uH:2.2uH:2.2uH" case with the 2.2-uH inductors from Murata.

I can also test this case: buck1:buck2,buck3=1.5uH:1.5uH:1.5uH, but I will still be using TDK inductors so it will not be a 100% 1-to-1 comparison.

Hello Brian-san,
This issue is difficult to reproduce. The system works correctly by using 2.2uH so asked to go by using 2.2uH.
Best regards,
Toshihiro Watanabe