# Generating a solid 5V output from a buck converter whose output varies from 4.8V to 5.2V

I have a buck converter whose output varies from 4.8V to 5.2V. This buck converter is present in my VCU. Input for the buck converter can vary from 8V to 14V.

The 5 volt rail is powering many IC's present on the board, and for those IC's, this 4.8V to 5.2V is fine and there are no issues.

The same output from this buck converter is provided as a reference voltage for the vehicle's throttle.

Since the output of the buck converter is varying, the throttle output also varies, which is not acceptable.

What circuit can I use to generate a constant 5V from this varying 4.8V to 5.2V buck output?

• It would be much simpler to "generate", say, 4.5 V. Any chance to get voltage range at the buck output small enough? May 11 at 16:44
• The buck is designed for 5V output,but When I tested the buck it's out put varies from 4.8V to 5.2V.
– Hari
May 11 at 16:48
• That's terrible load regulation for a buck converter. Any chance you can just fix the buck? Otherwise you'd have to put a buck-boost converter on the output to get 5 V. As @greybeard said a lower voltage would be easier to use as a reference, then you could just get an actual voltage reference or LDO. May 11 at 16:53
• Frame-challenge: you're doing the A2D conversion of your throttle voltage wrong. You should be using the supply voltage as the reference for your ADC such that it performs a ratiometric conversion of the throttle voltage. If you did that it wouldn't matter that the supply voltage is not perfectly constant. May 11 at 17:52
• I'm surprised no one has asked for a schematic, frequency, current, components or screenshot of voltage waveform of output and inductor, which would be key to fixing your problem. What capacitance do you have on the output? May 11 at 20:24

Design away the problem. Measure both the reference voltage (that is, your 5V) for the throttle and the throttle output itself and compute a ratio to determine throttle position. Then you can reject variances in the reference.

Otherwise, create local, stable reference at a lower voltage to use for your throttle voltage reference. You'd still want to measure it as a ratio regardless.

Also consider redundancy for your throttle sensor. In FuSa (functional safety) -speak, it would be an ASIL-D (highest level) component since its failure would lead to an uncontrollable condition resulting in a possibly fatal crash.

I was looking at one of the Texas Instruments "Simple Switcher" switching regulators, and the jellybean 33063 or 33064 to solve your problem, but it all seemed like way too much work just to furnish your 5V reference.

Besides fixing the Buck Regulator so it works properly, which should really be your first strategy as belabored already in comments, it occurred to me that a completely separate regulator using the same inputs into the flaky buck regulator would be the "perfect" solution.

So... You said, "Input for the buck converter can vary from 8V to 14V." Use those same inputs and throw in a completely separate voltage regulator. Such as an LM317T...

From this datasheet on this part, the STMicro LM317BT,

You may not need the 2N2219 if you're not going to turn it on and off by microcontroller or another circuit.

I would still add the protection diodes as shown below:

And you may need a minimum load of >= 10mA to ensure regulation.

And don't forget the heat sink!