LM5085: Switching power converter not regulating

Hi David,

That is correct. The voltage at the FB pin is greater than 1.25 V, but I'm not operating this in an open loop. There is just a potentiometer as a part of the feedback path. If Vfb > 1.25, shouldn't the controller turn the PMOS off?

Sorry for the confusion Alexander, 

The PGATE high means that the PFET is off.  You are setting voltage at FB pin to VREF and the device is doing what it is supposed to do by turning off the HS MOSFET.

Can you make the PGATE go low by reducing the voltage at the FB node?

Thanks.

David.

If the PMOS is off, then how come Vout is roughly equal to Vin?

That would indicate that it's behaving like a short. I just checked Rds, and it's on the order of several megaohms. I also checked the markings on the part, and they are correct.

I measured Pgate = 48.19 V. Vin = 48.2

I also measured Rdg and Rsg, and they were several MO

It turns out that the source and the drain were inverted. That sort of solved the problem... We're able to regulate the voltage to a certain degree. As I mentioned, there is a potentiometer in the feedback path to adjust Vout between 32 V and 45.7 V. Things seem to work well at 32 V and as we increase the voltage to 38 V, but once we exceed 38 V, Vout drops to 33 V and stays there.

I probed Vfb in order to get a handle of this. At the lower values of Vout, Vfb = 1.25 V as expected, but once we exceed 38 V, it drops down to 0.9 V and stays there.

I've made three peculiar observations:

1) If I probe the transistor gate signal with an oscilloscope probe, we're able to get up to 42 V.

2) The circuit behaves differently if we start at Vout_min and rotate the potentiometer towards Vout_max than if we start at Vout_max and rotate the potentiometer towards Vout_min. I've already described the former case. In the latter case, Vout stays around 33 V the whole time.

3) When the circuit is unpowered, and I measure the resistance between the potentiometer wiper and one of its other terminals, and even without adjusting the potentiometer, the measured resistance would slowly drop. This is similar to what I see when measuring the resistance of a capacitor, but much slower.

Alas, the load is practically open. That would also not explain the asymmetry between how the circuit behaves when adjusting the potentiometer from high to low versus low to high. It would also not explain why I can get to 42 V if I'm probing the transistor gate with an oscilloscope probe.

Although I will note that I did test it with an ELoad. At low output voltages, it performed well up to 500 mA (I didn't test further).

Hello Alexander, 

The asymmetry is probably not related.  Are you sure that the devices FB pin did not see greater than 7V when you were trying to get the device to work with the MOSFET in back to front?  If FB>7V the device will be damaged.  not sure what you did differently from the above schematic.

Also, please can you probe the gate voltage before, during and after the point of  question? 

Thanks, 

David.

Hi David,

We replace the device, so I can guarantee that it never saw more than 7 V. I see very little at the gate. It's pretty much just 48 V. It doesn't seem as though there is any switching.

I will note that our technician installed the big electrolytic output capacitor backwards (ie. wrong polarity). We definitely saw switching at the gate before reversing it to the correct polarity. Normally capacitors with reversed polarity are only a danger to themselves. Is it possible that it may have damaged the FET?

Hi David,

I'll ask our technicians to do that today.

Thank you!

Hi David,

Our technicians replaced the transistor, the diode and the regulator, and we're still seeing exactly the same problem. I don't see any switching, and yet there is a substantial voltage drop across the FET.

I probed the pins at both extremes of the potentiometer. Everything seemed pretty good. Here is what I found:

        "Low"       "High"

ADJ: 48.06      48.07

RT:   1.752       1.745

FB:   1.248       0.920

ISEN 48.21      48.22

PGATE: 48.21  48.20

VCC: 40.42      40.42

VDD: 48.22      48.22

Vout: 32.11       34.07

I did notice one strange artifact. Throughout most of the power side of the device (PGATE, VDD, VIN, ISEN, .etc) I saw periodic impulses. The frequency of these impulses was dependent on the position of the potentiometer and varied between 400 Hz and 800 Hz (relatively low frequency). Of course, I checked my power supply, but it wasn't present when not connected to the PCB.

Thank you!

Problem solved! Something was soldered to C7 (bypass capacitor for Vin) that was definitely not the 1 uF capacitor that it was supposed to be. I have no idea what it was but right now, I don't care. We're able to get our full and expected range.