LM5158: boost a power from 24 v to 60v

Hi 

Since our expert of this device is out of office this week, please expect a late reply here. 

Regards

Johannes

Hi Chen,

Thanks for using the e2e forum.
I did not find the schematic attachment, so I am not sure which resistor/cap is which in the parameter list.
Could you attach the schematic or the quickstart calculator you filled in ( https://www.ti.com/tool/download/SNVR511 ) ?

I can then review the design.

Thanks and best regards,
Niklas

Hi NIklas

you can refer to the page 14 of LM5158EVM-BST User's Guide (Rev. A) (ti.com)

 thanks

LM5158EVM-BST

R5=51K

R12=10K

R14=21.1K

C25=22uF

C27=330pF

C26=15nF

R19=5.1K

R8=147K

R13=2.47K

L2=22uH/XAL5050-223MEC

Cout = 6.6uF

Cload = 1mF

Hi Chen,

can you please fill out the quickstart calculator?  https://www.ti.com/tool/download/SNVR511

This will make it much easier for us to review the design.

Thanks and best regards,

Moritz

hi Moritz

Due to security rule by company, we can't upload file, PDF for you reference, thanks

LM5157-58_Excel_Quickstart_Calculator_for_Boost_Converter_Design_V1_0_0.pdf

Hi Chen,

Thanks for providing the calculator. Are you using a LM5158EVM-BST where you just changed the parts you mentioned or did you build your own board?

Does the problem occur only at a load between 0.3A - 0.4A? So at no load condition the circuit works normally?

If you are using the EVM, what is the position of jumper J11? You can try to

place it between pins 1-2 to disable the hiccup feature and try if it works then.

Best regards,

Moritz

HI Moritz

jump change to pin 1-2, disable the hiccup function, the output is ok,

But DUT have a  1mF cap in back of the LM5158. The voltage output is very slow,

did you have any advise?

change to test ADI parts, ADI : LT3958(DC1569A), the work is ok for me

Hi Chen,

Do you mean the output voltage does not reach 60V in the softstart time of 1.5s?

Can you please provide scopeshots of your measurements?

For this high load capacity, the compensation network has to be adjusted. You can use the quickstart calculator and set the output capacity to 1mF.

It will recommend suitable values for the compensation.

Best regards,

Moritz

hi Moritz

RD follow the TI excel to design the Power stage, but it's still can't meet customer spec.

Hi Chen,

did you adjust the compensation network to the high output capacity?

Can you please provide the most recent schematic to me?

And if it is possible some measurements of the issue you see.

Best regards,

Moritz

hi Moritz

did you adjust the compensation network to the high output capacity? rd follow below table to adjust the compensation in TI EVB

4718.LM5157-58_Excel_Quickstart_Calculator_for_Boost_Converter_Design_V1_0_0.pdf

Can you please provide the most recent schematic to me?  all the test in TI evb

And if it is possible some measurements of the issue you see.

below figure show the Circuit trig one mos, the voltage is ok 

below figure show the Circuit trig two mos, the voltage is drop & ripple 

below figure show the Circuit trig four mos by ADI solution, the voltage is stable

Hi Chen,

Thanks for providig the scopeshots.

In the ecxel calculator you provided, the compensation is not adjusted. The output capacitance has to be set to 1mF. Then the compensation network has to be adjusted accordingly.

Please refer to below picture.

Best regards,

Moritz

HI Moritz

RD follow your excel, the result is same. 

did you have a evb which input is 24v, vout is 60v? can drive 1mFx4 ?

thanks

Hi Chen,

Unfortunately we dont have a reference design using LM5158 for these voltage specs.

Can you please again show measurements of Vin, Vout , switchnode and COMP?

"can drive 1mFx4 ?" Do you use 4mF in total ?

Could you please clarify your application? Does 1MOS, 2MOS, 4MOS mean that you are driving 1x, 2x, 4x The load with 1mF?

Thanks and best regards,

Moritz

hi Moritz

24v boost to 60v, in back of the LM5158 will have a 1MF cap, 

red frame have x1,x2,x3,x4 , ADI solution can drive x4 module, but TI only can drive x1 modue

Hi Chen,

Thanks for clarifying.

looking at the scopeshots of LM5158 again, it seems like the time Vout is dropping matches the 1.5s sofstart time.

Can you please try to reduce the softstart time and measure again Vout, SS, UVLO and COMP?

Best regards,

Moritz

Hi team,

I ask customers to change some parameters here. I assume the total cap load is 4mF(1mF*4).every channel load is 120A per 30uS. Total 

shows waveform just let 1 CH load work. 

C25 = 4.7uF

R19 = 2.2M

C26 = 680pF

C27 = 10pF(C27 = 1pF，can not work，vout = vin)

CH1 is Vo, CH2 is SS. 

It seems the parameter can not support pulse current load. 

But customer use C25=22uF,R19=9.1MF,C26=470pF,C27=none. 1 CH load Is okay, but when turning on the No.2 load, Vo will drop.

(when 1CH load work) 120A/30uS

(When 2CH load work) 

So seems the compensation will affect the test result. but based on the customer's condition, it seems can not find good parameters so far.

Hi Jimmy,

You can see in the scopeshot, that the softstart is pulled down, so the device starts from new every time. The question is why.

Please send scopeshots as i requested above: Vout, SS, UVLO and COMP.

Please do these measurements wit one channel, when everything works and with 2 channels, where the voltage drops. Pleace reduce the SS-cap to 3.3uF for the next try. Also please adjust the compensation to 4x1mF output capacity. (At least 60degrees of phase margin at the crossover frequency, please use the excel calculator).

Best regards,

Moritz