frex
New Head-Fier
- Joined
- Dec 1, 2006
- Posts
- 22
- Likes
- 1
Hello,
Because it is interesting and also because somebody ask me, i continue
to make some regulators noise measurements.
I have test the noise performance of :
LM317 (Motorola)
LT1085CT (LT)
LM2940CT-12 LDO regulator (NationalSemi)
The measurements methodology is same of my previous post.
(Vout=15V unless otherwise specified Rload=100 Ohms)
Results are (10Hz-100kHz bandwidth) :
LM317 -- 55µVrms
LM2940CT12 -- 190µVrms (Vout=12V)
LT1085 -- 40µVms
These results show that the LM2940 série (Low dropout regulator) are the worst !
Finally for low budget, the best choice seem to be the LM317 with large decoupling capacitor in ref input.
Because i'm curious and also to answer of somebody,
i have tested the simple noise reduction circuit from Wenzel Associates app notes.
You can read this app note here : Finesse Voltage Regulator Noise!
The first circuit use only one NPN transistor, and of course if the noise reduction is effective,
it can be a right way to supply some sensitive device, like oscillators.
Because of the high serie impedance (15 ohms) in the supply path, only very low output current can be draw by the load (about 20-30mA).
After many tests, with same values of the original schematics and with slight modification too, i don't gain very good noise reduction. Only about 10dB of noise attenuation (1/3).
(Note: Iout=20mA, Vin=15V, Vo~15V).
I have also tested the second schematic using 2N5192 power transistor.
After differents settings, i obtain in the best case 10dB of attenuation.
Wenzel say about 40dB of attenuation is possible, but i' can't get it.
Since my last post, I also purchased a couple of SSR01/SSR02 PCB from Peter Sjöström. It is high performance serie regulators PCB.
As you can think, when wiring was finished, i have start to make some noise test.
The result is impressing !
The output noise for 15V output voltage is about 1.5µVrms !
(20V dc input, 150mA out, LT1115 AOP).
Both SSR01 (positive) and SSR02 (negative) give the approximatively same results.
I congratule P.Sjöström for this very compact and clean design !
I will use it for build a /-15V 1A very low noise power supply.
It will used to supply my new design,an high performance 24/192k ADC.
Frex.
Because it is interesting and also because somebody ask me, i continue
to make some regulators noise measurements.
I have test the noise performance of :
LM317 (Motorola)
LT1085CT (LT)
LM2940CT-12 LDO regulator (NationalSemi)
The measurements methodology is same of my previous post.
(Vout=15V unless otherwise specified Rload=100 Ohms)
Results are (10Hz-100kHz bandwidth) :
LM317 -- 55µVrms
LM2940CT12 -- 190µVrms (Vout=12V)
LT1085 -- 40µVms
These results show that the LM2940 série (Low dropout regulator) are the worst !
Finally for low budget, the best choice seem to be the LM317 with large decoupling capacitor in ref input.
Because i'm curious and also to answer of somebody,
i have tested the simple noise reduction circuit from Wenzel Associates app notes.
You can read this app note here : Finesse Voltage Regulator Noise!
The first circuit use only one NPN transistor, and of course if the noise reduction is effective,
it can be a right way to supply some sensitive device, like oscillators.
Because of the high serie impedance (15 ohms) in the supply path, only very low output current can be draw by the load (about 20-30mA).
After many tests, with same values of the original schematics and with slight modification too, i don't gain very good noise reduction. Only about 10dB of noise attenuation (1/3).
(Note: Iout=20mA, Vin=15V, Vo~15V).
I have also tested the second schematic using 2N5192 power transistor.
After differents settings, i obtain in the best case 10dB of attenuation.
Wenzel say about 40dB of attenuation is possible, but i' can't get it.
Since my last post, I also purchased a couple of SSR01/SSR02 PCB from Peter Sjöström. It is high performance serie regulators PCB.
As you can think, when wiring was finished, i have start to make some noise test.
The result is impressing !
The output noise for 15V output voltage is about 1.5µVrms !
(20V dc input, 150mA out, LT1115 AOP).
Both SSR01 (positive) and SSR02 (negative) give the approximatively same results.
I congratule P.Sjöström for this very compact and clean design !
I will use it for build a /-15V 1A very low noise power supply.
It will used to supply my new design,an high performance 24/192k ADC.
Frex.
