LM7805 unstable

[bogdant]

Do you think, replacing the LM7805 with a LM317 would  be an option? I am thinking 3k//1k on the adjust pin and a load of 1kohm rezistor, 1uF to input and 0.1uF to output capacitor for stabilization.
The Vin=35V and the target is to have Vout=5V. As consummers I have a micro + LCD, should be no more then 50mA.

[bd139]

LM317 will do the job. Buy a TI branded one, make sure it has a 10mA minimum load current either from the load or the voltage setting network and make sure it is adequately heatsinked. It'll still kick out over a watt there. A small clip on will probably be enough heat sink for that.

[mariush]

You should also look into AP7381  or LM2931.

LM2931 is available on
* Farnell : https://ro.farnell.com/search/prl/results?st=lm2931&sort=P_PRICE
* RS-Components: https://ro.rsdelivers.com/productlist/search?query=lm2931&tag=&family=9492#PI-41|PS-99|viewtype-2|sortBy-Price|sortOrder-asc

AP7381 is available on

* RS-Components : https://ro.rsdelivers.com/productlist/search?query=ap7381&tag=&family=9492#PI-1|PS-21|viewtype-2|sortBy-Price|sortOrder-asc

Note that the cheapest on RS-Components require buying certain quantities.

Digikey has a larger selection but being in US would be more expensive to ship : https://www.digikey.com/products/en?keywords=ap7381

[bd139]

LM2931 discounted because max input voltage is too low (26V)

AP7381 outside SOA ... user needs 1.6W dissipation.

High dropout voltages are a special problem space.

[Ian.M]

A LM317 *NEEDS* that 10mA minimum load current to guarantee regulation.  If the real load current can drop to zero, as the regulator acts to maintain Out 1.25V above Adj, the upper resistor of its potential divider should always be under 125 ohms to guarantee 10mA minimum load current for the regulator.   Use 120R and calculate the lower resistor for the desired voltage (96 ohms per volt at Adj).

References to a 5mA minimum load current and a 240 ohm upper resistor only apply to the LM117 which is on the same National Semiconductor datasheet as the LM317.

If there's any chance of spikes or ringing on the 35V supply, get a LM317HV which can withstand up to 60V input to output differental.

[StillTrying]

The Vin=35V and the target is to have Vout=5V. As consummers I have a micro + LCD, should be no more then 50mA.

Max. 50mA @ 30V is 1.5W, you could just about get away with a 15V 2W zenner in series with the 35V, to share the max. dissipation to about 750mW each, in practice a little heat sinking for the reg.

Sometimes data sheets can be a little unclear

It's clear they go to great lengths not to mention what happens below 5mA or above 25V.

[bd139]

It's clear they go to great lengths not to mention what happens below 5mA or above 25V.

That's always something my first EE mentor said to me. Look out for what the datasheet doesn't say!

One reason LT got a good reputation is they were honest where stuff didn't work.

[Cliff Matthews]

Aren't some brands of 7805 rated for Vin 40V? I up'd the 7805's heatsink on my ADS200 (fighting back 33v at ~200ma)
AFAIK, it was running close to 80c (20c below its thermal protect) with the standard heat sink shown on the right..

[ebastler]

Aren't some brands of 7805 rated for Vin 40V?

Only if you interpret the term "7805" broadly, to my knowledge.

Murata has (had?) switching regulators with 78xx pinout which can work up to 40V input voltage, with reduced current capability:
https://power.murata.com/data/power/dms-78xxsr.pdf

[Zero999]

The LM78xx or LM317 can be cascoded with an external transistor, to work at higher voltages. At higher currents, Tr1 should be a Darlington pair and R1 can be higher.

[Jwillis]

Regardless whether you use a 78xx or a lm317 you still need to remember that  Power in Watts  = (Volts in - Volts out) X I . The LM317 won't run any cooler than a 78XX so adequate heat sinking is still required.

[Km]

you use a 100  oham resistor half watt input side of 7805 its drops small voltages its cheap solution 
i use it many time

[jackthomson41]

Yeah its better to use heat sink with LM series as they get hot and burnt out, happened to me a lot. Moreover, use 7805 instead of LM317 as there's no need to design the small basic circuit but it will give fixed voltage of 5V. LM317 gives variable output.

[Siwastaja]

Note that it's difficult to completely eliminate input voltage peaks and ringing during powerup, or sudden load changes. This ringing originates from cable stray inductance oscillating with the input capacitance of your circuit, and can go up to 2x the supply, i.e., 70V in this case! This happens easily if you have a big ceramic capacitor on the input.

With electrolytic (high ESR) bulk capacitance in place, this peaking is easily damped, but even then, when running any part at the "absolute maximum ratings", it means that even a very slight input voltage peak may permanently damage the part the first time you apply power; or at any point in time later. You have zero margin for such inrush voltage peak!

[bogdant]

Note that it's difficult to completely eliminate input voltage peaks and ringing during powerup, or sudden load changes. This ringing originates from cable stray inductance oscillating with the input capacitance of your circuit, and can go up to 2x the supply, i.e., 70V in this case! This happens easily if you have a big ceramic capacitor on the input.

With electrolytic (high ESR) bulk capacitance in place, this peaking is easily damped, but even then, when running any part at the "absolute maximum ratings", it means that even a very slight input voltage peak may permanently damage the part the first time you apply power; or at any point in time later. You have zero margin for such inrush voltage peak!

I think it is clear I need to do same redesigne. It is clear the LM7805 will not work at 35V input

Sent from my VTR-L29 using Tapatalk