Voltage regulator with least noise

[vicky001.iith]

Among LT3080 and MIC39100 which one is better option for voltage regulator?? Which will have the least fluctuating output.?

 

[Vengateswaran]

For your requirement, LT3080 can be used effectively which have less noise., so that your signal won't fluctuate much.

 

[vicky001.iith]

Currently i am using this one and getting little noisy output. But can we remove the noise completely by adjusting Capacitance values?? If Fluctuations happens,how can we reduce its amplitude ??

 

[Vengateswaran]

The capacitance value you have used is low.. Adjust it to 100uf and try... Also check whether the input voltage you have applied is an pure one...

 

[vicky001.iith]

Hey, currently i doesn't have 100uF so I tried with 47uF ...it reduced but not too much.... tomorrow will try with 100uF ..
Input applied is a pure one only

 

[KerimF]

You may like adding a 10uF in parallel with R21 (though not included on the datasheet), or perhaps from the SET pin to ground.


Added:
Since you have 12V as input, using a standard regulator (not low drop) as LM7805 may give less noise.
Could you describe the noise as average amplitude and frequency (or range of frequencies)?

 

[vicky001.iith]

Hey KerimF, i didn't calculated the exact amplitude of noise and freq. range.It is like i am getting a little floating DC output,it not constant all time..Min.value is 4.8V or sometimes 4.6 and max. value is 5.1 or 5V. I mean Vpp(peak to peak) should be less than 80mV but it is around 400mV and 240mV.
I will try with LM7805 also.

 

[FvM]

I mean Vpp(peak to peak) should be less than 80mV but it is around 400mV and 240mV.

I mean Vpp(peak to peak) should be less than 80mV but it is around 400mV and 240mV.

That's no regular voltage regulator "noise". Specified noise level is in a µV range. Some possible explanations:

- defective parts or bad solder joints
- self oscillation because LT3080 capacitor requirements aren't kept, e.g. output capacitor is electrolytic type with ESR > 0.5 ohm
- noise is generated by load current transients
- any kind of measurement artefacts

 

[vicky001.iith]

Oh.Then i have to check the circuit again for defective parts...
and LT3080 capacitors requirements ??

 

[KerimF]

I recall my first circuit I designed after graduation (about 38 years ago). I had just a multi meter for my tests and I spent about 2 months in building it and recalculating the values of its few resistors/capacitors. It was a simple astable oscillator using 2 transistors I finished it when I discovered that the main electrolytic capacitor of its regulated power supply was connected in reverse polarity My voltmeter measured always Vcc as being a constant 5V while in reality it had a ripple of more than 1V. This taught me that if my calculations are likely right and I don't get the result I expect, I could be sure that I did one mistake or more in my connections and, as it happened to me many times, one component or more, I use, is not as specified. For example, once I bought a lot of BC546 NPN transistors to find out they are all PNPs And in another situation, their emitter and collector terminals were reversed (as being of another type)! I am saying this while I know that getting fake (if not faulty) components could mostly happen where I live unlike the case in many other places in the world. The solution is always... don't give up

 

[vicky001.iith]

You had experienced a lot Sir. I am new to these things(Graduated 3 months ago). Almost none of my circuits works at first attempt. I try to find the bug/fault and it takes days/weeks to me. Sometimes some component gets burned or short. Many times i use wrong components. It irritates very much.But in the end it has to work and it works Thats good part of circuits design.
And getting fake components is happening almost everywhere in this planet not only in Syria I too got a fake IR senor twice.

 

[KerimF]

Also during 38 years in designing electronic circuits, NONE of my circuits works at first attempt
Even now, I see myself lucky if I get all expected results (since my boards now are rather complex and include both hardware and software) after 50 attempts only