Recently a forum member (gmphadte) told me that the LM317/337 combo will be more quiet than the L7815/L7915 combo. I looked at the datasheets and don't know how to interpret the 317 ones because they are adjustable. He also mentioned that the quietest option would be a shunt regulator. I looked at the datasheet for a couple of zeners but found nothing about noise in them.
It got me thinking about finding the best +/-15V regulator chips for my op amps. So far, I've found a LM340 that looks to be a better option by a small margin to the L7815 which is what I have been using.
I have mainly been looking at "load regulation", "output noise voltage in uV" and "ripple rejection in dB". Are all voltage regulator chips going to have the same specs or is there a plus/minus pair that is the best for low noise? Should I be focusing on different specs?
Is it true that a zener will have the lowest noise of all regulator chips and the only advantage of an IC regulator is for it's safety features like short circuit protection and thermal shutdown?
It got me thinking about finding the best +/-15V regulator chips for my op amps. So far, I've found a LM340 that looks to be a better option by a small margin to the L7815 which is what I have been using.
I have mainly been looking at "load regulation", "output noise voltage in uV" and "ripple rejection in dB". Are all voltage regulator chips going to have the same specs or is there a plus/minus pair that is the best for low noise? Should I be focusing on different specs?
Is it true that a zener will have the lowest noise of all regulator chips and the only advantage of an IC regulator is for it's safety features like short circuit protection and thermal shutdown?
Last edited:
I have read the datasheets as well as reports on 317/337 and 78xx/79xx 3pin regulators.
What I have seen is that the standard 317/337 is not any quieter than the fixed voltage 78/79 types.
It's only when the extra capacitors, to filter the noise, are added that the 317 can outperform the 7805. Then we are not comparing like for like.
The 317 does seem to respond very well to enormous capacitance across the Vref on it's adjust pin. 220uF seems to be about the biggest I have seen and used.
What I have seen is that the standard 317/337 is not any quieter than the fixed voltage 78/79 types.
It's only when the extra capacitors, to filter the noise, are added that the 317 can outperform the 7805. Then we are not comparing like for like.
The 317 does seem to respond very well to enormous capacitance across the Vref on it's adjust pin. 220uF seems to be about the biggest I have seen and used.
LM317 and LM7805 are 40 year old obsolete parts (originally designed in the 1970s).
There are more modern devices with significantly less noise available.
Linear Regulator (LDO) Power Management Solutions | New Power Regulators
There are more modern devices with significantly less noise available.
Linear Regulator (LDO) Power Management Solutions | New Power Regulators
I've built regulator using TPS7A4700, connected two of them to get symmetrical output voltage. Comparing to LT108x regulators or LM317, this one is much more cleaner and detailed in sound.
BR,
Ales
BR,
Ales
Like Andrew says.
Read the classic ap note, Understanding and Reducing Noise Voltage in 3-Terminal Regulators. It's on-line somwhere, not sure where. Also, this page is pretty good. The 317 is ancient, but still hard to beat if you choose the extra components correctly. There are certainly some newer high performance parts, but often in SMT only. Heck, I've still got some 317 parts in TO-3 cans somewhere! It starts to be a parts jungle, but you can clean up any regulator- http://www.wenzel.com/documents/finesse.html
Read the classic ap note, Understanding and Reducing Noise Voltage in 3-Terminal Regulators. It's on-line somwhere, not sure where. Also, this page is pretty good. The 317 is ancient, but still hard to beat if you choose the extra components correctly. There are certainly some newer high performance parts, but often in SMT only. Heck, I've still got some 317 parts in TO-3 cans somewhere! It starts to be a parts jungle, but you can clean up any regulator- http://www.wenzel.com/documents/finesse.html
I'm curious if super low regulator noise is something we should look at all except in some special cases like DAC clocks? Does it have any impact on sound? Especially considering far from zero PSRR of most of the stuff they typically feeding. I found disturbing how easy some of the modern low-noise LDOs can oscillate.
Low noise o.e.m. "chip" regulators: ADM7150, LT1761, TPS9xxx
Zener is best operated near 10mA -- and true zeners are lower noise than avalanche diodes.
Zener is best operated near 10mA -- and true zeners are lower noise than avalanche diodes.
found another thread on the same topic: http://www.diyaudio.com/forums/power-supplies/247643-modern-ultra-low-noise-ldo-adm7150.html
The need for a regulator to have the very low self noise is something that is, as you point out, application dependant. If the signal amplitudes of the circuit the regulator is powering are small, such as for an phono stage input circuit, or simply highly noise sensitive, like DAC clocks, then low regulator self noise can be critical. In a 2VRMS line level circuit, however, I tend to feel that typical regulator self noise is not much of an issue. So, I tend to agree with your view.
Regulator Zout is often more of an issue in practice than its self-noise, particularly when powering opamps.
Where in the datasheets do you find this? Is this also a reason to not put a resistor in line with the power before it feeds the opamps, like for a filter?
Last edited:
Why would it be in the datasheets? Its an application issue for the designer to be aware of.
There's more to it - if you put a resistor in line then you're increasing the opamp's self-generated noise by virtue of increasing the supply impedance. But you're also giving some isolation from that noise to the regulator. For myself I use an inductor so as not to significantly increase the supply impedance seen by the opamp.
@ALD - 0.3ohm Zout up to 1MHz looks pretty good - probably better in practice than a TL431 shunt. I'd want to check though what output current that figure was found at - probably at the top end of its range. Ah I see that's shown in the lower plot - I'd be a bit worried about the regulator's noise - 1000nV/rtHz is considerably worse than a TL431 and that's well known as a noisy beast.
Last edited:
obviates the utility of a low Z regulator...but...
just my 2 pfennigs, but I think that PSRR is more important than noise or Zout.
PSRR of the regulator you mean? Aka line regulation?
PS you're right that anything in series would render a very low Zout reg redundant. I make do with the 0.2R Zout of a TL431 typically.
PS you're right that anything in series would render a very low Zout reg redundant. I make do with the 0.2R Zout of a TL431 typically.
I'd just use one of TI's swift series switchers.
With frequency options from 600KHz to 6MHz.
Putting the noise way up there makes it easy
to trap with a low pass filter.
The lower frequency Swifts can be simple as
5 pins, and a little more forgiving of layout.
With frequency options from 600KHz to 6MHz.
Putting the noise way up there makes it easy
to trap with a low pass filter.
The lower frequency Swifts can be simple as
5 pins, and a little more forgiving of layout.
Last edited:
The issue with switchers is that the noise, is easily radiated and easily rectified by the PN junctions in IC's and discrete components. I believe that Analog Devices called this the reason for unexplainable DC offsets and THD. I don't know if it's as true in practice as it is in reality.
Does, however, make a case for the quiet switchers from Linear Tech -- these switch more slowly, so that di/dt is lower. Take a look at the LT3439.
Hey, looks like we've done some of this already ...
http://www.diyaudio.com/forums/power-supplies/143539-another-look-lm317-lm337-regulators.html
"So I set up to measure the noise of these program resistor/adjust capacitor combos, using an Amber 3501a with 30kHz lowpass filter engaged. (All have 1000uF || 1k2 ohms to ground on the output.)
240/2k7 || 82nF (82uV)
120/1k3 || 10uF (40uV)
1k/11k || 100uF (30uV)
Given the huge noise difference, I rejected the first one and decided to try the two others in the D/A: first 120/1k3 || 10uF, and then 910/9k1 || 100uF. Listening impressions compared to the 78xx/79xx: ..."
...
"Hold on until the next installment is posted and you'll see where this is going. My main interest is in the output impedance, and am using the noise/impedance correlation for clues into it, at least until I can get a rig up to measure impedance with DC present. "
...
"Now we get to look at some noise spectra. (Until I get confidence in how to post photos, I'm going to post this in chunks.) The following pics are all 0-5kHz sweeps, 10dB/division, -70dbV full scale, 30Hz filter BW with smoothing, and a painfully slow sweep rate. I didn't show higher than that because these voltage set components didn't change the noise spectra that much above 3kHz. With a smaller output cap they might.
The first pic shows the the extremes. Top trace is 120/1k3 with no cap. Bottom trace is 950/11k with 100uF cap. The top trace is the high trace all across, the two intersect right about 2.5kHz. The lower program current gives 2-3dB less noise at high frequencies. The real benefit of the adjust cap is below 1kHz, with a 15 to 25dB reduction in noise in the midband. "
...
"The pic on the left is the impedance of the "optimized" LM317 and LM337 into 1500uF ||1k2 ohm loads. The impedance range is 0 to 0.2 ohms vertical, DC-5kHz horizontal. And, what do you know, the impedance curves bear a strong resemblance to the noise curves. Who woulda thunk it. "
...
"The bad news is, I could find no combination of set resistors, adjust cap, or load cap that makes the 317 impedance anywhere near as linear as the 337."
..."So, to recap, where this has landed so far is, for ±15VDC:
LM317 - 80 / 880 with 10uF adjust cap (bypassed of course)
LM337 - 120 / 1k3 with 10uF adjust cap
1000-2200uF output cap, with as large a film bypass as you have room for (I like to put 10uF film across these)
Protection diodes as per data sheet.
Separate leads from each adjust circuit to the load ground. "
..."Here is what walt jung measured for the LT1085:
Click the image to open in full size.
also, i think the 1085 has better load regulation:
LT1085 load regulation = 0.1%
LM317 load regulation = 0.3%
those numbers are from the data sheet "headline". it is hard to compare the graphs directly because of scaling issues.
My understanding is that the LM337 and the LT1033 are pretty equivalent - my transient response testing is in agreement with this. However, the LT1085 is supposed to be superior to the LM317. I have not done my transient response testing on the positive regulators yet."
..."Now here are some VERY interesting LM317 impedance sweeps, varying the input-to-output voltage differential. This should put to rest some previous speculation about this. The steps were 8V, 6.5, 5, 4, 3, and 2V. Voltage was changed by putting an autotransformer in the line and varying the AC voltage to the transformer, so no circuitry around the regulator was changed.
As you can see the i/o voltage has a very well-defined affect on the output impedance around and below the critical resonance zone. This gives us yet another ingredient to optimize the output of this regulator. Optimum zone is in the 2 to 4 volts range.
An interesting angle to this is how this will change the sound of your system when your AC line voltage changes. As it goes up, the bass will get emphasized at the expense of the mids."
..."If your speakers have good phase behavior you will be blown away by the naturalness and holography.
LM317:
out to adjust - 240 ohms
adjust to gnd : 2.7k ohms || 0.12uF
output cap: 1500uF w/ 10uF film
minimum load: 100mA
LM337:
out to adjust - 240 ohms
adjust to gnd : 2.7k ohms || 0.033uF
output cap: 1000uF w/ 10uF film
minimum load: 50mA
Input to output differential : 2.5 volts"
..."Ok, here are some harmonic distortion measurements, comparing an opamp-amplified signal first with the 317/337 pair with standard 10uF adjust caps and more normal 5.4V input/output differential, and then with the optimized adjust caps with 3.0V diff. I don't think HD is going to adequately describe the effect that balancing and linearizing the PS impedance has, but it's a place to start."
"Like I said, I didn't expect a THD measurement would capture what I imagine to be more of a nonlinear distortion. Nevertheless, even a single jellybean opamp with a signal well within it's linear zone shows lower HD with better balanced and linear phase supply rails. "
...
"The regulator adjust circuits are:
Standard: (except for load R, both regs have the same values)
Vin-Vout = 6V
out to adjust - 240 ohms
adjust to gnd : 2.7k ohms || 100uF
output cap: 1500uF w/ 10uF film
load resistance: 165 Ohms (317) and 330 Ohms (337)
Optimized:
LM317: Vin-Vout = 2.7V
out to adjust - 240 ohms
adjust to gnd : 2.7k ohms || 120nF
output cap: 1500uF w/ 10uF film
load resistance: 165 Ohms
LM337: Vin-Vout = 2.7V
out to adjust - 240 ohms
adjust to gnd : 2.7k ohms || 33nF
output cap: 1500uF w/ 10uF film
load resistance: 330 Ohms
So the only difference between std and opt. is the Vin-Vout and the adjust caps."
http://www.diyaudio.com/forums/power-supplies/143539-another-look-lm317-lm337-regulators.html
"So I set up to measure the noise of these program resistor/adjust capacitor combos, using an Amber 3501a with 30kHz lowpass filter engaged. (All have 1000uF || 1k2 ohms to ground on the output.)
240/2k7 || 82nF (82uV)
120/1k3 || 10uF (40uV)
1k/11k || 100uF (30uV)
Given the huge noise difference, I rejected the first one and decided to try the two others in the D/A: first 120/1k3 || 10uF, and then 910/9k1 || 100uF. Listening impressions compared to the 78xx/79xx: ..."
...
"Hold on until the next installment is posted and you'll see where this is going. My main interest is in the output impedance, and am using the noise/impedance correlation for clues into it, at least until I can get a rig up to measure impedance with DC present. "
...
"Now we get to look at some noise spectra. (Until I get confidence in how to post photos, I'm going to post this in chunks.) The following pics are all 0-5kHz sweeps, 10dB/division, -70dbV full scale, 30Hz filter BW with smoothing, and a painfully slow sweep rate. I didn't show higher than that because these voltage set components didn't change the noise spectra that much above 3kHz. With a smaller output cap they might.
The first pic shows the the extremes. Top trace is 120/1k3 with no cap. Bottom trace is 950/11k with 100uF cap. The top trace is the high trace all across, the two intersect right about 2.5kHz. The lower program current gives 2-3dB less noise at high frequencies. The real benefit of the adjust cap is below 1kHz, with a 15 to 25dB reduction in noise in the midband. "
...
"The pic on the left is the impedance of the "optimized" LM317 and LM337 into 1500uF ||1k2 ohm loads. The impedance range is 0 to 0.2 ohms vertical, DC-5kHz horizontal. And, what do you know, the impedance curves bear a strong resemblance to the noise curves. Who woulda thunk it. "
...
"The bad news is, I could find no combination of set resistors, adjust cap, or load cap that makes the 317 impedance anywhere near as linear as the 337."
..."So, to recap, where this has landed so far is, for ±15VDC:
LM317 - 80 / 880 with 10uF adjust cap (bypassed of course)
LM337 - 120 / 1k3 with 10uF adjust cap
1000-2200uF output cap, with as large a film bypass as you have room for (I like to put 10uF film across these)
Protection diodes as per data sheet.
Separate leads from each adjust circuit to the load ground. "
..."Here is what walt jung measured for the LT1085:
Click the image to open in full size.
also, i think the 1085 has better load regulation:
LT1085 load regulation = 0.1%
LM317 load regulation = 0.3%
those numbers are from the data sheet "headline". it is hard to compare the graphs directly because of scaling issues.
My understanding is that the LM337 and the LT1033 are pretty equivalent - my transient response testing is in agreement with this. However, the LT1085 is supposed to be superior to the LM317. I have not done my transient response testing on the positive regulators yet."
..."Now here are some VERY interesting LM317 impedance sweeps, varying the input-to-output voltage differential. This should put to rest some previous speculation about this. The steps were 8V, 6.5, 5, 4, 3, and 2V. Voltage was changed by putting an autotransformer in the line and varying the AC voltage to the transformer, so no circuitry around the regulator was changed.
As you can see the i/o voltage has a very well-defined affect on the output impedance around and below the critical resonance zone. This gives us yet another ingredient to optimize the output of this regulator. Optimum zone is in the 2 to 4 volts range.
An interesting angle to this is how this will change the sound of your system when your AC line voltage changes. As it goes up, the bass will get emphasized at the expense of the mids."
..."If your speakers have good phase behavior you will be blown away by the naturalness and holography.
LM317:
out to adjust - 240 ohms
adjust to gnd : 2.7k ohms || 0.12uF
output cap: 1500uF w/ 10uF film
minimum load: 100mA
LM337:
out to adjust - 240 ohms
adjust to gnd : 2.7k ohms || 0.033uF
output cap: 1000uF w/ 10uF film
minimum load: 50mA
Input to output differential : 2.5 volts"
..."Ok, here are some harmonic distortion measurements, comparing an opamp-amplified signal first with the 317/337 pair with standard 10uF adjust caps and more normal 5.4V input/output differential, and then with the optimized adjust caps with 3.0V diff. I don't think HD is going to adequately describe the effect that balancing and linearizing the PS impedance has, but it's a place to start."
"Like I said, I didn't expect a THD measurement would capture what I imagine to be more of a nonlinear distortion. Nevertheless, even a single jellybean opamp with a signal well within it's linear zone shows lower HD with better balanced and linear phase supply rails. "
...
"The regulator adjust circuits are:
Standard: (except for load R, both regs have the same values)
Vin-Vout = 6V
out to adjust - 240 ohms
adjust to gnd : 2.7k ohms || 100uF
output cap: 1500uF w/ 10uF film
load resistance: 165 Ohms (317) and 330 Ohms (337)
Optimized:
LM317: Vin-Vout = 2.7V
out to adjust - 240 ohms
adjust to gnd : 2.7k ohms || 120nF
output cap: 1500uF w/ 10uF film
load resistance: 165 Ohms
LM337: Vin-Vout = 2.7V
out to adjust - 240 ohms
adjust to gnd : 2.7k ohms || 33nF
output cap: 1500uF w/ 10uF film
load resistance: 330 Ohms
So the only difference between std and opt. is the Vin-Vout and the adjust caps."
Last edited: