LM5118

Hi 

Looks like your FB pin is floating now. 

Regards,

Sorry, a line is missing.

FB is connected between R17 and R19.

Alessandro Campanella said:

Sorry, a line is missing.

FB is connected between R17 and R19.

I mean, the line is missing on the schematic, not on board.

Do you need BOM?

Schema:

6712.DCDC2.pdf

BOM:

4505.BOM_DCDC2.pdf

Please help me, something should be wrong!

Hi 

Would you  please capture Q1 source-to-ground and Q2 drain-to-ground waveforms at the same time (with load when Vin is 12V, 15V, 20V and 48V) ? 

Please give me the information about the behavior as much as possible.

Regards,

Hi,

sorry for pictures.

Q1, 12V, source-to-ground:

Q2, 12V, drain-to-ground: 

Q1, 15V, source-to-ground:

Q2, 15V, drain-to-ground: 

Q1, 20V, source-to-ground:

Q2, 20V, drain-to-ground: 

Q1, 48V, source-to-ground:

Q2, 48V, drain-to-ground: 

Thank you.

Hi 

Would you please capture two waveforms together using two probe ? 

Also, would you please let me know the load condition when you testing. 

Regards,

Hi,

Load is about 5W, if I connect a bigger load VOUT falls down.

Q1 source-to-ground (CH2), Q2 drain-to-ground (CH1)

12V:

15V:

20V:

48V:

Hi 

Looks like something wrong on your board. 

Source-to-ground voltage goes up to about 15V when it should be 20V, it goes up to about 35V when it should be 48V. 

I guess there is a wrong connection or wrong component population on your board. 

Regards,

Hi,

I verified carefully each line/wire, board and connections are ok.

Could you please verify if mosfets PSMN025-100D,118 and diodes STTH3002G-TR are compliant with my application?

Thank you.

Hi 

PSMN025-100D is 100V rated and STTH3002G-TR is 200V rated. Both are enough to handle 60V input. 

Regards,

Hi,

so, in your opinion, diodes and mosfets are not the cause of this strange behavior.

What else?

Hi 

Source-to-ground voltage should go up to about 48V when the input voltage is 48V. 

If not, I think you should double check if any layout miss, wrong component population or any MOSFET damage. 

Regards,

We are getting crazy about this circuit, because seem that every test circuits ( same ) have different result.

We check all the components, mosfet seem ok, problem seem to be in the main voltage regulation.

Today i mount a new circuit, the mosfet seem open and close regularly, but the problem is the ... frequency. 10- 12 hz.

So i check the vcc, and vcc is also oscillating like a ramp at this value.

Vin is ok, vccx disconnected or ground shorted ( the same ). En is ok, shunted to vin. Rt is ok.

On sync i can see the right pulsations, but at 11 hz Seem something like that main voltage regulator cut off... for some reasons. we check components, schematic, pcb....

Hi Eric,

Could you please suggest some tests and/or checklist in order to troubleshoot this circuit?

We are forced to abandon TI for other manufacturer's IC if the we can't solve the problem.

many thanks.

Hi 

I would like to suggest to check VCC first. It is now only 3.36V while it should be around 6.8V~7.2V. 

Apply 60V supply voltage, adjust UVLO pin voltage to 0.7V and measure VCC voltage. 

Have you replaced IC ? 

Regards,

Hi,

could you please draw a table indicating voltages I should find on pins?

I will make each measurement asap.

IC is new, I can replace it if you think I should.

Many thanks for helping.

BR,

We found the problem: components population.

In your evaluation board documentation there is an error ( not in the hardware ); following it we was burning the ic at the power on.

Now  work good, very low heat  and stable out ; we have just a problem of very high ripple at high loads.

We have to take out about 50v 3 ampere. We will encrease the big out caps going up tp 1000 uF and we will try to change rt resistor in order to set a higher switching freq.

Any suggestion ?

Hi 

It is good to hear you found the cause. 

Just one comment is to increase soft-start capacitor when you increasing the output capacitor. 

Otherwise, a hiccup mode protection might be triggered. 

Regards,

Startup problem: without load OK

With load ic died. Can depend on soft start cap ? We should increase ?

if u start up without load and after is applied ... all ok

We have some problems with output power.

Without load, no problem.

With load and 50v output we have some difficulties with Vin.

We have to take out 50v and 2 ampere, with vin from 24 to 60, the problem i that when we start to move the vin more than 45 or 55 the vout start to fall down. We try to change the inductor in order to see if work better, but nothing.

Any suggestion for increase the performance ?

Alessandro,

Sorry but Eric is on vacation and will not be back until July 8th, I will try to help. Are you using the SYN input?  The voltage on the RT pin should be 2V DC.  Is the LM5118 running hot?  If not, then if VCC pin voltage is oscillating at 11 Hz, there must be an open pin, not soldered to the PWM.  I would re-check the connections.

Terry

Hi Terry,

thank you for  reply.

We found there was a problem with UVLO resistive divider that was burning the IC (more than 16v on UVLO pin).

Solved. Now the circuit has an output voltage of 48V exactly like we expect it to be.

We experienced following problems:

1. applying a load of about 60w it works only if input voltage is high, near 50v and only if load is applied not at power on.

2. at low input voltage it can handle only very little load.

We are recalculating each component but we need confirmations.

Vin should be 20-60v, Vout 48V and output power near 100w.

Can you please suggest inductance, ramp capacitor, ss capacitor and sense resistor?

Regards,

Alessandro.

Alessandro,

You can down load the LM5118 Quick start calculator from the TI website.  The inductor is 90uH, Ramp 3000pF, Rramp 37k, Rsense 27mohms.

Terry

Alessandro Campanella said:

Could you please verify if mosfets PSMN025-100D,118 and diodes STTH3002G-TR are compliant with my application?

I'm using in my prototype THT "cousins" of mentioned components, namely: PSMN9R5-100XS and STTH802FP and they works fine. Did you experiment  with Cramp capacitor and also something that is not covered in the LM5118 datasheet but appear in quick xls calculator: Rramp connected between pin 5  and 16? That is basically what Terry suggested.