dual buck LTC3892 one channel not switching

[ferbo]

Hello,

I have designed a dual buck with LTC3892.
40 to 50 v input
vout1: 19v 10 A
vout2: 12v 6 A
my EXTVcc is connected to the second output.
for FETS I am using: BSC0804LSATMA1
For L in 1st channel: PA4344.223NLT
For L in 2nd channel: SRP1770C-330M

I connect the input of 48 volts. I have 4.95 on INTVcc, 9.95 on DRVcc I have 11.5 volt on VOUT2 and 7.5 volt on VOUT1. PGood2 is high and PGood1 is low which shows the programmed output voltages are almost right.

when checked with oscilloscope The channel 2 gate drivers are switching but not the channel 1.

Do you have an idea where is the problem coming from?

Thank you.

 

[wwfeldman]

i did not examine your schematic in any detail.
i assume channel 1 and channel 2 are (essentially) identical.
i would look for a missing connection.
suggested things to look for: missing power, missing ground, shorted gates

run1 and run2 enable their respective channels and appear connected to some circuits in the RUN box.
maybe that 2M resistor is too big and the voltage at run1 and run2 is marginal, so one channel is on and the other off.
what are J5 and J7 connected to? what should they be connected to for correct operation?

 

[ferbo]

i did not examine your schematic in any detail.
i assume channel 1 and channel 2 are (essentially) identical.
i would look for a missing connection.
suggested things to look for: missing power, missing ground, shorted gates

run1 and run2 enable their respective channels and appear connected to some circuits in the RUN box.
maybe that 2M resistor is too big and the voltage at run1 and run2 is marginal, so one channel is on and the other off.
what are J5 and J7 connected to? what should they be connected to for correct operation?

Thank you for your help. The top Fet gates seem to be shorted to the switching pin of inductor in both channels however. The jumpers in open situation are designed to have the channels running. if conneced that channel would turn off.

 

[wwfeldman]

i assume you corrected the shorted gates
if the gates were shorted, why did one channel work (appear to work??) while the other did not work?
did that fix the problem?

 

[ferbo]

i assume you corrected the shorted gates
if the gates were shorted, why did one channel work (appear to work??) while the other did not work?
did that fix the problem?

No I cannot find the source of the short even when I remove the zero resistor that connects the TG pin of IC to the mosfet, the short is still present on the IC side of net. but when I remove the IC it's not present anywhere. I think that's just present when there is no input. or else the channel 2 would not be switching.

 

[wwfeldman]

some years back, i watched a technician using a microscope to find a short on a PWB
he did not.
but when he took an x-acto knife and ran it between each surface mount pin, he removed an invisible solder splash
the board worked.

you might try that
and you might try a new control IC