I have a noise problem at an oscillator starting at abt 100 kHz up to abt 1 MHz offset from the carrier (see picture).
I measured the spectrum of the power supply - and there is a remarkable noise increase around 500 kHz +/-.Problem located.
I realized the supply (10V/100 mA) with a REF01 and booster transistor 2N 2905A.To reduce noise further, I put an RC lowpass at the output (2,2Ohms, 4700 uF).
Looks as if the unit is very sensitive to capacitive loads. I need to increase the 2,2 Ohms resistor at the output of the REF01 to abt 10 Ohms to get rid of most of the noise.
Do you have any suggestion to deal with this problem without increasing the resistor? (Put a capacitor across the REF01?)
Maybe I should change to AD587 instead of REF01, or would there be the same problem?
Your suggestions are very welcome!
As I understand the question, you used REF01 as a power supply for your oscillator? is this correct? If that is correct, then why do you use Voltage reference in the first place, is there specific requirement for the application? Can you use LDOs?
The noise you are referring to which is a bump at around 600kHz can be cause by anything in the circuit, it could be the REF01 or the transistor or other components in the circuit and changing the REF01 to another voltage reference might help or not unless we figure out which component produces that bump.
There are lot of low noise LDOs available in our portfolio that we can suggest. But can you confirm first what type of voltage source is 21V (referring to your schematic) going in to the 7815 regulator? Can you confirm if it's the same clean power supply you use for the rest of the circuit.
Voltage references are not good at driving high capacitive load. There is also no available spectral noise density plot for REF01 to verify if that bump is coming from the REF01 itself.
no, the supply powers only the loop.The VCO has another (clean) supply.
But the noise is there without any load, and is gone as soon as I remove the 4700uF cap (with 2,2 Ohms resistor).
In a REF03 datasheet I remember a workaround for capacitive loads (with booster transistor) by putting a cap across the regulator.
The sideband noise graph was where I first Noticed the problem-the noise obviously modulates the VCO via the control voltage (from the loop).
I want to understand the problem.The setup is not so uncommon.
But of course,if there is a low noise 10V /150mA LDO that can deal wich capacitive load (preferably a leaded version), I am also open for any suggestion.
Gesendet: Mittwoch, 31. Mai 2017 09:33
An: Angst Günther
Antwort an: firstname.lastname@example.org
Betreff: Re: - Re: Noise problem with a REF01 with booster transistor
Re: Noise problem with a REF01 with booster transistor
reply from iamPhil<iamPhil> in Power Management - View the full discussion<https://ez.analog.com/message/303829?commentID=303829&et=watches.email.thread#comment-303829>
yes,I can confirm,the preceding supply is clean.I measured it and tested with another 21V source ,too. Same behaviour. I am pretty sure the REF produces the noise so I want to have a remedy. What about the cap across the regulator? Might that help?
Gesendet: Donnerstag, 1. Juni 2017 04:58
Antwort an: email@example.com
reply from iamPhil<iamPhil> in Power Management - View the full discussion<https://ez.analog.com/message/303989?commentID=303989&et=watches.email.thread#comment-303989>
Thanks for confirming it. I'll take a look at that option and confirm the problem with the team. I'll get back to you as soon as possible.