Problems with on-board voltage regulation

[kek7u]

Hi,

I have a sensitive analog/RF circuit (1.2V supply, draws max. current 300mA) and for testing purposes I need an adjustable 1.2V (+- ~100mV) supply voltage.

I would prefer on-board voltage regulation, but I cannot find any adjustable linear regulator configuration that could produce such a small output voltage and I don't dare to use switching converters (due to noise issues, right?). There's no on-chip voltage regulation, just some bypass capacitor.

What I'm left with is to take raw DC-supply connection to the chip? If so, do I need to consider some kind of on-board LC-filtering, or just place some bypass capacitors (both close to the chip and power connector)?

I'd appreciate any help with this..

 

[ericgibbs]

hi,
You could use a zener reference diode driving a power OPA.
Place a variable resistor across a 2.5Vref and adjust the input to a non inverting OPA. set it for +/-0.1V on 1.2V

Do you follow that.?

 

[kek7u]

Could you please explain this in more detail cause I'm not quite sure what you mean.. thanks! =)

 

[ljcox]

If you look at the data sheet for the LM317 voltage regulator, it can be set to regulate at voltages between 1.2 & 25 Volt.

For 1.2 don't insert R1 and put a short circuit in place of R2. That will give you around 1.2 ~ 1.25 Volt.

Let us know if you have any problems.

 

[kek7u]

LM317 was actually my first approach, but it had problems getting below 1.25V. I really would like to have quite accurate tuning range around 1.2V, or at least from 1.2V to 1.3V.

 

[k7elp60]

You may want to consider still setting the LM317 up as an adjustable regulator and putting a diode on the output terminal which will drop less than one volt, but then you could adjust the output to where you want it.
The forward voltage drop will vary a little with the diode. You could use a 1N4001 or a 1N5817. The 1N5817 is a schottky diode so the voltage drop would be lower.

 

[Hero999]

Diodes are a bad idea as they'll ruin the regulator's good regulation. The voltage can be reduced all the way down to 0V by adding an op-amp and a few resistors.

https://www.electro-tech-online.com/threads/variable-bipolar-power-supply.24291/
https://www.electro-tech-online.com/attachments/0v-to-15v-lm317-pdf.9790/

 

[Roff]

Will this help? It's quick and dirty, and a short to GND on the output will destroy the TIP31, but it is short-term stable, and has good source and load regulation.

 

[k7elp60]

Diodes are a bad idea as they'll ruin the regulator's good regulation. The voltage can be reduced all the way down to 0V by adding an op-amp and a few resistors.

Yes that is true if the load is varying. If the load current is fairly stable the
voltage across the diode will be stable.

 

[Roff]

Diodes are a bad idea as they'll ruin the regulator's good regulation. The voltage can be reduced all the way down to 0V by adding an op-amp and a few resistors.

https://www.electro-tech-online.com/threads/variable-bipolar-power-supply.24291/
https://www.electro-tech-online.com/attachments/0v-to-15v-lm317-pdf.9790/

And a low-current negative supply, which could be a battery or a wall wart.

 

[Blueteeth]

I'm always up for designing things in a clever way....but sometimes I get lazy and just look for a chip

Look at https://www.linear.com

devices: LT3021, LT3021ES8-1.2

Also national semi have a few ultra low noise regulators:
LP38501-ADJ for example goes down to 0.6V @0.1% line regulation to 1000mA

It does seem that many regulators have an internal reference of 1.2-1.25V, but there are so many nowdays its just easier if you have access to internet distributors. The previous answers covered the DIY approach quite well, and if you do RF I'm sure you have enough knowledge to carry it out. GOod call with avoiding the switchers, I've been stung recently on an RF project by them....nothing filters them completey.

Just wanted to add something different.

Blueteeth

 

[Hero999]

And a low-current negative supply, which could be a battery or a wall wart.

I'd use a 555 negative charge pump.

This has already been discussed before.
https://www.electro-tech-online.com/threads/how-to-get-1v-by-using-regulator.31659/

 

[ljcox]

LM317 was actually my first approach, but it had problems getting below 1.25V. I really would like to have quite accurate tuning range around 1.2V, or at least from 1.2V to 1.3V.

Do you have a negative voltage available?

If so, you could supply a small negative voltage to the adj terminal of the LM317.