LM5066I: LM5066I not turning off properly

Hi Rakesh,

I recorded some waveforms during startup and shutdown, triggered by the UVLO/EN-pin. Once with a 470 Ohm load, once without any load

Startup with 470 Ohm load

Shutdown with 470 Ohm load

Startup without load

Shutdown without load

Hi Christian Waechter,

Vout =~Vin at no-load in disable mode concerns me. Is there any other path feeding the Vout and causing to drop when loaded with 470 Ohm load.

Is RC43 mounted?

Few things to try: Please look at input current with 470 Ohm load to check from where the Vout is getting powered; Next replace T23, T24 MOSFETS (one at a time) and verify.

Best Regards,

Rakesh

Hi Rakesh, thank you for your answer,

I cannot see or measure any other path from Input (DCDC_V+) to Output (POWER). I also do not think that this is a random short circuit, because this problem is on multiple of these boards each with 3 LM5066i.

RC43 is assambled so that DCDC_V+ is connected to V_TL_FB. The third pad to POWER is not connected on RC43.

One thing about the DCDC_V+-voltage. It is generated by a DCDC-converter, that itself is turned on and off. This means, that the LM5066i is not always supplied via the Vin pin (pin 5). Also it is guaranteed by a hardware logic, that  the UVLO/EN Pin of the LM5066i is never on a high level, when the supplying DCDC-converter is turned off or starting up. The supply of the other components (+3.3V, +12V, ...) is always on.

I recorded the currents through with and without a 470 Ohm resistor in the on- and off-state of the LM5066. It looks like the output is only supplied by the DCDC_V+ input as intended.

I'm going to try changing the FETs and will report as soon as I have anything.

Best regards,
Christian

Hello again,

I think i found the reason for the output voltage, even with disabled power switch. As I wrote in the first post, I'm using a circuit for reverse current protection from Application note SNVA673. There is a resistor R4 which originaly is 10k and is connected to Vin via the Diode D1. The resistor was changed to 1k in my circuit, which than will cause a voltage drop of ~2/3 when a 470 Ohm resistor is connected externally (remaining ~1/3 voltage drop).

Is there any idea on how to avoid this behaviour when using this kind of reverse current protection? Changing the resistors is required as they adjust the trigger voltage/current.

Hi Rakesh,

I'm expecting continuous 10 A at 48 V and reverse voltage of 60 V. I would like to avoid a integrated solution and use the already designed discrete circuit shown with the LM5069.

What do you think, might a single diode in series with the resistor R4 do the trick? Anode connected to VOUT, Cathode to R4.

Best regards,

Christian

Hi Rakesh,

I was not yet able to try serializing a diode with R4.

The problem is not the Opamp. I'm using a TLV333 with only 15 uV Offset.

The problem is that the input voltage can be measured on the output, even when the LM5066 is turned off. This is because there is a conducting path from VIN through D1 and R4 to VOUT. Increasing R4 would not solve the problem completely as this would still require a load on the output.

The question is how to improve this circuit for the reverse current protection. Has noone else ever build this and faced the same problem?