[time-nuts] Low noise voltage regulators

Bruce Griffiths bruce.griffiths at xtra.co.nz
Thu Feb 25 02:07:44 UTC 2010


WB6BNQ wrote:
> Bruce,
>
> Why not just use a diode in place of the 2M3904 circuit ?  Seems like it would do the same
> thing with less parts.
>
>    
That can be made to work, however the reference divider impedance may 
then need to be decreased somewhat.
Using a transistor minimises the interaction between the reference 
divider and the loading due to the 3k7 connected from the reference 
filter cap to the external reference.
> Also, would not the Texas Instruments TL-431 be a better choice as it has a lower noise spec
> them the LM329 part.  In addition, with its adjustable nature, the TL-431 would be adjusted
> to make sure the single diode (in place of the 2N3904 circuit) is shut off when the TL-431
> comes up to speed.
>
>    
Thats doesn't appear to be true, see attached noise spectra  taken from 
the respective datasheets.
NB the L329 has a nominal 6.9V output whereas the TL431 has 2.5V output 
(for the noise spectrum shown).

However a lower voltage reference is more appropriate for lower output 
voltages.
Also the input noise of the LM723 error amplifier (50nV/rtHz??) has also 
to be taken into account.

> Obviously, this would be best with intended output voltages above the reference level.
>
> Bill....WB6BNQ
>
>
>    
Bruce
> Bruce Griffiths wrote:
>
>    
>> I meant something like the attached circuit schematic for an LM723 based
>> 15V regulator.
>> The circuit can be easily extended to use an external pass transistor
>> where more current is needed.
>>
>> The LM329 is biased from the regulator output which improves the
>> reference line rejection.
>> The 723's internal reference is used only during startup.
>> The 2N3904 disconnects the internal reference when the LM329 terminal
>> voltage exceeds about 5.4V.
>> Some optimisation of the circuit may be required.
>>
>> Bruce
>>
>> Arnold Tibus wrote:
>>      
>>> Hi,
>>>
>>> NS gives some informations about improvements in their AN-173.pdf
>>> http://www.national.com/ds.cgi/LM/LM117.pdf
>>> Audio freaks are discussing it in
>>> http://www.diyaudio.com/forums/solid-state/28978-improving-lm3x7-regulator-circuit.html
>>> Is that what you are looking for?
>>>
>>> 73
>>> Arnold
>>>
>>>
>>> On Thu, 25 Feb 2010 11:22:02 +1300, Bruce Griffiths wrote:
>>>
>>>
>>>        
>>>> You can improve the performance of the LM723 if one substitutes an LM329
>>>> for the internal reference biased from the regulator output.
>>>> The trick is to use the internal reference for startup and decouple it
>>>> with a diode or similar once the LM329 achieves its nominal output.
>>>>
>>>>          
>>>        
>>>> Currently, there appear to be 2 variants of the LM723 one (made by
>>>> National) that uses a noisy bandgap reference and another variant that
>>>> actually uses a quieter zener reference.
>>>>
>>>>          
>>>        
>>>> Bruce
>>>>
>>>>          
>>>        
>>>> Brooke Clarke wrote:
>>>>
>>>>          
>>>>> Hi neville:
>>>>>
>>>>> My old Gibbs rack mount 5 MHz standard used the LM723 linear
>>>>> regulator.  I believe it's one of the lowest noise regulators you can
>>>>> use.
>>>>> http://www.national.com/mpf/LM/LM723.html#Overview
>>>>>
>>>>> Have Fun,
>>>>>
>>>>> Brooke Clarke
>>>>> http://www.PRC68.com
>>>>>
>>>>>
>>>>> Neville Michie wrote:
>>>>>
>>>>>            
>>>>>> Hi,
>>>>>> I remember a reference, probably by Bruce, that LEDS provide a low
>>>>>> noise voltage reference.
>>>>>> I am proposing to build a voltage regulator for a thermally
>>>>>> controlled LPRO rubidium oscillator,
>>>>>> with the voltage regulator being mounted on the 0.5 inch thick
>>>>>> aluminium heat sink plate.
>>>>>> The LEDS would also be mounted on the plate, which has controlled
>>>>>> temperature.
>>>>>> The LPRO has internal voltage regulation, and by running it at ~40C
>>>>>> and 18Volts, the thermal
>>>>>> flux within the unit is minimised as is the power demand.
>>>>>> What I want to know is if a LM317 running on a stack of LEDs driven
>>>>>> by the LM317 output
>>>>>> would provide a low noise power source? What would be better?
>>>>>>
>>>>>> cheers, Neville Michie
>>>>>>
>>>>>>
>>>>>>              
>>>
>>>
>>> _______________________________________________
>>> time-nuts mailing list -- time-nuts at febo.com
>>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>> and follow the instructions there.
>>>
>>>
>>>        
>>    ------------------------------------------------------------------------
>>                                   Name: LowNoiseLM723Regulator.pdf
>>     LowNoiseLM723Regulator.pdf    Type: Portable Document Format (application/pdf)
>>                               Encoding: base64
>>
>>    ------------------------------------------------------------------------
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the instructions there.
>>      
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
>
>    

-------------- next part --------------
A non-text attachment was scrubbed...
Name: LM329(6V9)Noise.png
Type: image/png
Size: 9929 bytes
Desc: not available
URL: <http://www.febo.com/pipermail/time-nuts/attachments/20100225/f47e0b27/attachment-0002.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: TL431Noise(2V5).png
Type: image/png
Size: 34202 bytes
Desc: not available
URL: <http://www.febo.com/pipermail/time-nuts/attachments/20100225/f47e0b27/attachment-0003.png>


More information about the time-nuts mailing list