pushing 7805 over its limit

Thread Starter

mxabeles

Joined Apr 25, 2009
295
I might be thinking to far ahead but SAY I have a boatload of IC's controlling another boat load of synthesizer/lights, the amount of chips + devices will probably exceed 1A. What then?
Thanks,
M
 

SgtWookie

Joined Jul 17, 2007
22,230
It would be best if you simply used a switching power supply that is adequate to operate all of the items that you wish to power.

ATX form factor computer power supplies are quite inexpensive, and they can be converted to general purpose "bench" supplies reasonably easy.

Google "ATX bench supply" for lots of ideas.
 

rjenkins

Joined Nov 6, 2005
1,013
You could also just use a higher current three terminal regulator.

Example - LM323, 5V 3A.

I believe the 'current amplifier' mentioned above refers to using a PNP power transistor as a current booster with a 7805 (or other regulator).

The DC source positive connects to the emitter of the transistor, the base goes to the input of the 7805 and the collector goes to the output of the 7805.

Add a 10 or 22 Ohm resistor across the base and emitter of the transistor.

Once the current through the 7805 is enought to drop 0.6 - 0.7V across the resistor, the transistor starts to provide power direct to the load.

The voltage regulation is still controlled by the 7805, but the current limit is bypassed - you must fuse the input or add a separate current limit.

The transistor will need a good heatsink & an isolation kit. Fit the 7805 next to the transistor so it's thermal sensing also helps protect the transistor.
 

studiot

Joined Nov 9, 2007
4,998
I might be thinking to far ahead but SAY I have a boatload of IC's controlling another boat load of synthesizer/lights, the amount of chips + devices will probably exceed 1A. What then?
RJenkins beat me to offering a higher current capacity fixed 5v regulator.

You could also use an LM338 which goes up to 5 amps.

An aside I'm not sure what you mean by +devices in the above quote but the 5 volts will provide supplies for the controlling circuitry but not for the SCR or whatever driving the lights themselves. So you do not need to consider these as the supply here will come from the mains. ie you will be controlling the mains input to the lights.
 

eblc1388

Joined Nov 28, 2008
1,542
There is one important point that need to be clarified.

The use of LM338 over 7805 does not give more current continuously if the heatsink is tiny, as shown in the OP's link. I doubt the current can even reach 1A, with 7W dissipation.

The rating of a voltage regulator depends largely on how much heat it has to dissipate and ultimately the size of the heatsink. If those heat energy cannot be dissipated efficiently then the regulator will heat up. To protect itself it will go into current limit to reduce the heat dissipation.

It is not commonly known that if mounted on a huge huge heatsink in the Antarctic, the 7805 can supply up to 2.4A as stated in the datasheet.

If your 7805 cannot provide higher than 1A using a certain size heatsink because it become hot, then changing to LM338 will not help. It will go into current limit soon and one is left with the same current.

The way to get more current is to use a bigger heatsink and then look at how much the regulator can output. Better still, change to a SMPS as Sgt has suggested.
 

R!f@@

Joined Apr 2, 2009
10,004
L chung,
which company makes a 2.4A regulator.
Max one I have seen is 2.2A and it is peak, means not continues.
This is something new to me.
 

eblc1388

Joined Nov 28, 2008
1,542
which company makes a 2.4A regulator.
I'll let you guess from the image of the datasheet.



Max one I have seen is 2.2A and it is peak, means not continues.
This is something new to me.
Peak for how long? Why can't the regulator maintain the peak current?

What is your understanding of the following:

eblc1388 said:
if mounted on a huge huge heatsink in the Antarctic
 

Attachments

Wendy

Joined Mar 24, 2008
23,798
All of what has been said is true. You can also create 5V with discrete parts too, in this case the power is entirely up to the design.

SMPS (switching mode power supplies) tend to be a bit noisy, but they can create really huge currents with very little heat. It is part of the trade off. They are also converters, A 10V power supply might use 6Amps and output 5V at 10 amps. In this example the regulator is 80% efficient, the remaining 20% is wasted as heat. When compared to quiet linear regulators this is extremely efficient.
 

studiot

Joined Nov 9, 2007
4,998
The 78xx series of regulators are internally limited so attempts to continuously draw peak currents over 2 amps will lead to limiting or shutdown or both.

Obviously the more current you draw and the higher the input voltage the more power you dissipate in the regulator and the more cooling you require.

Say your input supply is 10 volts and you are drawing 2 amps. The dissipation is (10-5) x2 watts or 10 watts.

Your data sheet (I could not find that one on alldatasheet so I agree with RIFAA, who made them?) specifies a junction to ambient of 35 degrees per watt so the temp rise without heatsink would be 350 deg!

Junction to case is 4 deg per watt so with a perfect heatsink the rise would be 40 degrees and with a good TO3 heatsink of 2.5 deg per watt 65 degrees.
 

eblc1388

Joined Nov 28, 2008
1,542
Your data sheet (I could not find that one on alldatasheet so I agree with RIFAA, who made them?)
LM7805, National Semiconductor databook, a pretty old one.

Junction to case is 4 deg per watt so with a perfect heatsink the rise would be 40 degrees and with a good TO3 heatsink of 2.5 deg per watt 65 degrees.
What will be your guess of the temperature rise with a huge huge heatsink at ambient of -25°C ? :)
 

R!f@@

Joined Apr 2, 2009
10,004
Aaaah.. what if I were to jetison a 5V reg into space and try to draw 2.4A. without a heatsink. Then what?
 
Last edited:

R!f@@

Joined Apr 2, 2009
10,004
I like our power master regulator.. that is 1388 :D
He has us brainstorming most of the time and we are trying to prove otherwise.

Oh by the way.
I have voodooyoyo, wrecked (Wunce) , Klatu, PMR.....hehehe guess.
new nicks are coming
I am still thinking one for loosewire...may be loosewire itself suffices :p
 

Audioguru

Joined Dec 20, 2007
11,248
Go to your favourite electronic parts distributor and ask for a 5V regulator that somebody selected to supply 2.4A. They will laugh at you because you get whatever they have which has a guaranteed minimum output of only 1A. Some might supply 2.4A.

If you don't cool it well enough then it will simply heat then shut off.
 
Top