LM5088: -48V output not regulating and seems to be stuck resetting /hiccup mode (LM5088-2)

Also I have gnd and EP tied together.

Hi Kevin,

For inverting buck-boost configuration EP should be connected to -VOUT. All of the IC's GND references should be connected to -VOUT.

See application note:https://www.ti.com/lit/an/snva856a/snva856a.pdf

You are starting up to positive voltage correct? This is incorrect operation. Can you share your schematic?

-Orlando

Yes, I am using a regulated +20V supply input.

It seems to be going/stuck in hiccup mode, I measure the current at +20V input and it is spiking >2.6Amp with no load out.

I am using the same IC/design for -20V and that works fine.

I am also using the same IC for +12v, +6v designs with no issues.

yes, EP and gnd are on the -48V rail.

The compensation values have been adjusted accord to excel calculator supplies by TI.

I have pulled Q17 and made sure the system is enable.

I am a little confused about your statement "You are starting up to positive voltage correct? This is incorrect operation. " if I do not start the +20V rail how is the system going to work.

Sorry, I thought your nVOUT was rising to 20V, but looks like it goes ~-20V.

I cant see your schematic, can you upload as a zip file?

If it is going into hiccup during startup, then increasing soft-start time should help.

Also please share any oscilloscope shots of SW node during failure. 

-Orlando

ok, here is the schematic

-48poeRev1.pdf-48poeRev1.pdf

The problem is the LM5088 is a "buck only" controller. This controller can't regulate a boosted the voltage even if it is negative.

I had tried different and large inductors but no luck, then after discussion with co worker we realized the problem.

I will have to use a different solution, maybe the lm5118 which is buck/boost or ?

Hi Kevin,

No I dont think that's the issue, the LM5085 will see an input voltage of ( VIN+|nVOUT| ) and an output voltage of |nVOUT| so it will always operate as a step-down.

Did you try increasing soft-start time? I noticed is your output capacitance is very large (100µF) so charging up to 48V could trigger hiccup with your soft-start.

Can you try a startup with reduced output capacitance or increase soft-start?

Your schematic looks OK, my current thoughts are your C_RAMP might be too small and your FSW might be too large for this application. Too small C_RAMP will definitely trigger current limit early.

I'm glad to see you have a R_RAMP there though.

Send over your completed quickstart calculator as well.  

-Orlando