Problem with L4973D5.1 output voltage

[hmohs.1]

Hello

I bought a l4973D5.1 but I have a problem with its voltage output! For almost half an hour, the output of the regulator on the second and third pins is 5.1 volts, but after half an hour, the output of the regulator becomes 0 volts, but the output of the eighteenth pin of the regulator is 5.1 continuously! Where do you think the problem comes from? Has anyone had a similar experience?

[Peter BENSCH]

Welcome, @hmohs.1​, to the community!

What does your circuit look like?

What is the temperature of the L4973D5.1 when the output voltage drops to 0?

Regards

/Peter

[hmohs.1]

Thank you for your answer. The schematic photo of the circuit was sent in the attachment, and the temperature of the chamber does not exceed 30 degrees

[Peter BENSCH]

The circuit looks OK for now, except for minor details:

• what type is D11?
• L4973 can work with input voltages of up to 55 V (OK, the SS34, which is probably used to protect against polarity reversal, can only work up to 40V), so the TVS diode D13 should be dimensioned with a higher voltage. If you feed in up to 24V, of which approx. 23.5V remain after D12, D13 starts to conduct at typ 21.1V, possibly already from 20V. Either reduce the maximum input voltage to 20V or use an SMBJ22A or SMBJ24A for D13.

Next steps:

• what is the saturation current of L1?
• what are your parameters, i.e. at what input voltage and output current did you test?
• what does the layout look like?

[hmohs.1]

D11 is also SS34, and you are right about D12 and D13, but my voltage input is 24V at its maximum, and I did not assemble the Smbj diode at all, and about L1, I must say that it is a 68 μH power inductor, and it is 12 x 12, and My output current is 1.5 A and the input voltage varies from 9 to 24 volts, and also the photo of the assembled board was sent in the attachment.

While this board is placed on a Raspberry Pi, do you think the temperature of the Raspberry under the board can cause this problem?

[Peter BENSCH]

Well, on the one hand, the Raspberry with its heat generation may well contribute to the problem, because the temperature difference is added on top by the L4973's own losses. This is exacerbated by the fact that, apart from a few square millimetres around the GND pins, you have virtually no cooling surface.

Furthermore, even with the switching frequency of 150kHz, a better layout is necessary:

• all current-carrying lines as one-sided and wide as possible,
• all switched lines (especially the OUT-L1-D11 connection) as short and low-impedance as possible, and also
• the feedback line as short as possible and as far away from switched lines and inductors as possible.

You can find an example of this in the data sheet of the L4973 on page 11 (fig 5, 6).

Maybe the layout is still halfway sufficient for a function.

Probably the L4973 switches off because of an internally detected thermal overload. However, if the error occurs, you could spray some cooling spray on the L4973 to cool it down and get it working again. If that helps, you need to think about better cooling.

Regards

/Peter

[hmohs.1]

Thank you for your detailed answer.

As a last question, is it possible that there is a problem with the ground wiring? Because there are 3 max485 on the board! Does this mean that this problem may have arisen due to not observing the EMC and EMI guidelines?

[Peter BENSCH]

I don't think the MAX485 are a problem for the L4973, rather the switching regulator for the RS485 communication. But as already mentioned, the layout and especially the cooling still needs improvement.

Regards

/Peter