LM2594: LM2594-5.0

Petr Machacek

Part Number: LM2594

Dear,

During start up, a short overvoltage appears at the output, cca 1V/3ms above 5V. The input voltage is 24V/+-10%. What is the reason for overvoltage? How can this be solved? Thanks in advance for the suggestions. Best regards, Petr.

over 3 years ago

0 Elena Gao over 3 years ago

TI__Genius 13735 points

hi Petr,

Could you share you SCH and waveform .

Elena

0 Petr Machacek over 3 years ago in reply to Elena Gao

Prodigy 30 points

Hi Elena, SCH and waveform added. Petr

0 Elena Gao over 3 years ago in reply to Petr Machacek

TI__Genius 13735 points

hi, Petr

Looking from the datasheet ,the input and output cap could be larger than yours, could you try to increase input cap ,and check Vin Vout SW IL waveform, also need you could share PCB Layout for check .

Elena

0 Petr Machacek over 3 years ago in reply to Elena Gao

Prodigy 30 points

Hi Elena. Increasing value of input cap.does not help but progress is visible when we increase output cap.. We added to current 2x47uF one cap.with value 180uF, over voltage is now 0,4V during start up. It is accept for us. Please look at the layout if you see any problem. Thank you, Petr

0 Elena Gao over 3 years ago in reply to Petr Machacek

TI__Genius 13735 points

hi, Petr,

The output ripple or unstable is caused by many item, like input power stage has too fast transient ,or over spiked, we can increase input cap to improve. Out put cap also help with the ripple and loop stability, in your application it is a fixed Vout version so we could do nothing to add a CFF or RFF to improve the gain margin phase margin. The layout looks good, if you want to double confirm suggest to do a loop test and view the bold plot.

Elena

+1 Petr Machacek over 3 years ago in reply to Elena Gao

Prodigy 30 points

Hi Elena. Thank you for explanation and layout checking. So we will stay with increasing of output cap. Just for my interest, how can loop test be done, can you give me more information how to do it? Petr

0 Elena Gao over 3 years ago in reply to Petr Machacek

TI__Genius 13735 points

hi, Petr

share some guide and documents for referance.

https://www.ti.com/lit/an/snva364a/snva364a.pdf?ts=1663136810330&ref_url=https%253A%252F%252Fwww.ti.com%252Fsitesearch%252Fen-us%252Fdocs%252Funiversalsearch.tsp%253FlangPref%253Den-US%2526searchTerm%253DSNVA364A%2526nr%253D27

https://www.ti.com/lit/an/sluaa51/sluaa51.pdf?ts=1663136611213&ref_url=https%253A%252F%252Fwww.ti.com%252Fsitesearch%252Fen-us%252Fdocs%252Funiversalsearch.tsp%253FlangPref%253Den-US%2526searchTerm%253Dbuck%2Bloop%2Btest%2526nr%253D7900

Elena

0 Petr Machacek over 3 years ago in reply to Elena Gao

Prodigy 30 points

Hi Elena. Thank you for your support and have a nice day. BR,Petr

0 Elena Gao over 3 years ago in reply to Petr Machacek

TI__Genius 13735 points

My pleasure.

0 Petr Machacek over 2 years ago

Prodigy 30 points

Hi Elena

I have to go back to our issue. I found in the datasheet (9.2.1.2.2) that not posible use bigger output cap.than 220uF. What is the reason? What happened whan we use eg.330uF polymer output cap., KEMET T545Y337M10ATE35? Thank you for feedback. BR,Petr

0 Elena Gao over 2 years ago in reply to Petr Machacek

TI__Genius 13735 points

If the Cout capacitance is too large then the circuit may not have enough bandwidth and causes unstable loop.

0 Petr Machacek over 2 years ago in reply to Elena Gao

Prodigy 30 points

Hi Elena. Thank you for explanation. Let me summarize: layout is OK, Cin has no effect, L we tested sizes 220u/330u/470u and it has no effect, only Cout has an effect but the max. capacity of 220uF can be used, I am sending a picture. The initial overvoltage is now 5V+0.8V. Is there also a limit regarding ESR min? Do we have any other option to reduce the output overvoltage when the source starts up? Petr

0 Elena Gao over 2 years ago in reply to Petr Machacek

TI__Genius 13735 points

hi Petr,

Please try to compare the Vout difference before and after increased. if you could also have a loop test, check phase and gain margin, and share the result here to confirm the loop is stable ,then if is fine to increase Vout cap, since we already know it works .

Elena

0 Petr Machacek over 2 years ago in reply to Elena Gao

Prodigy 30 points

Hi Elena,

I'm attaching measurements with an output capacity of 2x150uF and Gain and Phase measurements for a 100mA load. The output voltage is now below 6V with some margin. Can you please evaluate whether this combination of output capacitor can be used?

Thank you. Petr

0 Elena Gao over 2 years ago in reply to Petr Machacek

TI__Genius 13735 points

hi, Petr,

The Vout waveform looks good, and could check under 100mA the device is CCM mode , usually we test loop stability in CCM, if it is DCM , the it will not have loop issue.

Elena

0 Petr Machacek over 2 years ago in reply to Elena Gao

Prodigy 30 points

Hi Elena

Thank you for feedback. You mentioned under 100mA, but I think to CCM mode we will get with current load above 100mA. Pls. confirm.

BR,Petr

0 Elena Gao over 2 years ago in reply to Petr Machacek

TI__Genius 13735 points

Petr,

Usually we required phase margin 45 degree and gain margin -10db around, checking the test result , it looks good .

BTW may I ask if this is an exist project or a new design, and on which customer side, Since the LM2594 is really old part, we are willing to recommend new device with same function and better performance. you could share more infor to elena-gao@ti.com if you want. Thank you for post on E2E.

Elena

Power management

Power management forum

LM2594: LM2594-5.0