[time-nuts] Parallel voltage regulators

Don Collie donmer at woosh.co.nz
Thu Oct 25 04:24:29 EDT 2007


----- Original Message ----- 
From: "Bruce Griffiths" <bruce.griffiths at xtra.co.nz>
To: "Don Collie" <donmer at woosh.co.nz>; "Discussion of precise time and 
frequency measurement" <time-nuts at febo.com>
Sent: Thursday, October 25, 2007 7:20 PM
Subject: Re: [time-nuts] Parallel voltage regulators


> Don Collie wrote:
>> Bollocs, Bruce! If National say it will do it, you can bet that it will. 
>> An
>> LM338K will do the job too, but in my opinion it`s overkill, and in the
>> event of a short circuit on the output of the regulator the current for 
>> the
>> LM338 will only be limited to [.......he gets the book......] 8 Amps 
>> [Typ],
>> as against 2.2 Amp [Typ] for the LM317T. This would probably be too much 
>> for
>> the transformer, rectifiers, and smoothing capacitor, effectively meaning
>> that you would have no current limiting. If the input/output differential
>> was kept in the range of 5 to 10 Volts, while the oven was stabilising, 
>> and
>> the LM317 had an adequate heatsink, it would do the job nicely [and 
>> cheaper
>> too!] Actually, it wouldn`t matter if the oven supply went unregulated 
>> while
>> the temperature was stabilising, because you wouldn`t be using it for
>> measurements during this time anyway - or is that a bit radical!?
>> All the best!,..................................................Don.
>> ----- Original Message ----- 
>> From: "Bruce Griffiths" <bruce.griffiths at xtra.co.nz>
>> To: "Don Collie" <donmer at woosh.co.nz>; "Discussion of precise time and
>> frequency measurement" <time-nuts at febo.com>
>> Sent: Thursday, October 25, 2007 5:42 PM
>> Subject: Re: [time-nuts] Parallel voltage regulators
>>
>>
>>
> Show me where that is actually guaranteed on the datasheet.
> Only the inexperienced and the gullible fall into the trap of assuming
> every regulator (or any other device) will meet its typical specs.
> The designer of this particular regulator actually cautions against this
> cavalier approach to design.
>
> If you worried about the transformer a simple fuse (resettable or
> otherwise) will surely cure that problem.
>
>
> Bruce

 Hi Bruce,
    I`ve got the National Semiconductor Corporation Voltage Regulator 
Handbook [1982]. On page 3-3, the leftmost graph shows the LM117/217/317 as 
having its current limit, with a junction temperature of
125 degrees Centigrade, at 2.25 Amps over the input/output differential of 5 
to 10 Volts.
    The point being, that if you use a higher current regulator, you loose 
the advantage of
the regulator`s current limiting, and perhaps, its thermal shutdown 
protection as well.
    A fuse *might* protect the semi`s down the line, but often it`s the 
semi`s that fail before the fuse, and the peak current that might flow 
before the fuse blows might be many times the current limit of the regulator 
[which is nearly instantaneous], and if so, damaging, so it is wise to run 
these regulators near their current limit, just as you would set the current 
limit on a bench supply to just above the working current.
    I find your use of the emotive words "inexperienced",  "gullible", and 
"cavalier" saddening.
Wishing you well,................................Don.
 




More information about the time-nuts mailing list