7805 regulator running hot!!

Archiver

Hi all,

Maybe someone will have some experience with 5-volt regulator. While I am
waiting for an LM2940CT

- Begin side note -
Make sure, if you plan to use the LM2940 you get the -CT termination, my
local store gave me one that had different pinout, to give -5V. This
resulted in about a 90% of the regulator input voltage (for 12V this was
around 10.8V) being delivered out of pin 3. This would have probably killed
the Stamp.
- End side note -

I am using a 7805 regulator with a small heatsink. I had it running for some
30 minutes, and the truth is that I couldn't touch the reg or heatsink, as
it was so hot.
I understand that heat dissipation is a function of input voltage, but is
there a way to reduce this dissipation, also, what heat levels can be
expected?
I am using around 200mA of the 1A the regulator delivers, so I don't see an
over-current problem.
One third question, are the capacitors at the input and output of the
regulator strictly necessary?

Cheers,

Mike

Archiver

Mike:
What input voltage are you using? These regulators have built-in protection
that shuts down the regulator on overtemp, but it still doesn't seem like a
good idea to run them that hot. You can "move" some of the heat out of the
7805 by using a resistor in series with the regulator input; the resistor
will share the excess heat. Pick the resistor so that there is 7.5 to 8
volts at the reg input under max load.

The input/output caps are needed to prevent regulator oscillation.

Good luck,
Ray McArthur

---

Original Message

---

From: Miguel Puchol <mpuchol@w...>
To: <basicstamps@egroups.com>
Sent: Monday, July 24, 2000 5:58 PM
Subject: [noparse][[/noparse]basicstamps] 7805 regulator running hot!!


> I am using a 7805 regulator with a small heatsink. I had it running for
some
> 30 minutes, and the truth is that I couldn't touch the reg or heatsink, as
> it was so hot.
> I understand that heat dissipation is a function of input voltage, but is
> there a way to reduce this dissipation, also, what heat levels can be
> expected?
> I am using around 200mA of the 1A the regulator delivers, so I don't see
an
> over-current problem.
> One third question, are the capacitors at the input and output of the
> regulator strictly necessary?

Archiver

Hi Ray,

Thanks for your reply.

> What input voltage are you using? These regulators have built-in
> protection
> that shuts down the regulator on overtemp, but it still doesn't
> seem like a

The input voltage is 12 to 14 volts, current consumption should be no
greater that 0.5A at max load.
I will do some test, and will try your idea of placing a resistor at the
input. I suppose that being an energy problem, dissipated heat will be the
same, just distributed over two components.

All the best,

Mike

Archiver

Miguel Puchol <mpuchol@w...> said:

> Hi Ray,
>
> Thanks for your reply.
>
> > What input voltage are you using? These regulators have built-in
> > protection
> > that shuts down the regulator on overtemp, but it still doesn't
> > seem like a
>
> The input voltage is 12 to 14 volts, current consumption should be no
> greater that 0.5A at max load.
> I will do some test, and will try your idea of placing a resistor at the
> input. I suppose that being an energy problem, dissipated heat will be the
> same, just distributed over two components.


this is a good solution, but try the capacitors first if you haven't already
done so. I've found regulators to get overly hot and then cool right down
after I remembered the caps.

Jason Lavoie

Archiver

At 12:04 PM 7/25/00 +0000, Jason wrote:

[noparse][[/noparse]snipped for brevity]

>this is a good solution, but try the capacitors first if you haven't already
>done so. I've found regulators to get overly hot and then cool right down
>after I remembered the caps.

I agree. Check your circuit for flaws or shorts. I'm running a 7805 at
~300mA with 24V going in and a small heatsink - it barely gets warm. Your
regulator shouldn't be running so hot that you can't touch it...

Duncan

Archiver

I guess I am getting in on this subject late but if you are using a 7805 in a
TO220 case and pulling 0.5A then you are pulling a pretty substantial load.
The amount of heat generated also is relative to the difference between the
input and output voltages. It is, however, NOT at all unusual for a 7805 to
run quite hot under a full load or continuous load. I have always made it a
practice to use a heat sink on them and even tho they run hot I have not had
any problems. Of course, if the circuit is enclosed you must have adequate
air flow to take the heat out of the enclosure. Don't know if this helps and
good luck.

Randy

Archiver

I am sure this is redundant but YES the caps at the input and output are
necessary for reliable operation as well as helping to reduce the heating
problem. Didn't see this post until after I wrote the other reply. And as I
said in the previous post, it is not unusual for them to run hot but with a
heatsink and only pulling 200mA it should be running fairly cool and you do
not seem to have that much of a difference betweent the input and output
voltages. The caps will probably take care of it.

Randy

Archiver

SOUNDS like too much voltage drop at one one location
could place an 9 volt reg in front of 5 volt reg and each one have their own
heat sink

get a bigger heat sink

blow on it with muffing fan

use a switcher

string together a few silicon diodes and get 0.7 volt drop from each one

good luck

richard

---

Original Message

---

From: "Miguel Puchol" <mpuchol@w...>
To: <basicstamps@egroups.com>
Sent: Tuesday, July 25, 2000 4:46 PM
Subject: RE: [noparse][[/noparse]basicstamps] 7805 regulator running hot!!


> Thanks to all who responded.
>
> I have now done the same that Parallax does in the Super Carrier Board,
> which is to place a 47uF electrolytic cap at the input and a tantalum 1uF
at
> the output. After adding this, the circuit seems to be running just as
hot.
> I've had it on for 12 hours since adding the caps, and the heatsink and
> regulator can't be touched.
> The rest of the circuit seems to be running just fine.
>
> I have a theory now, which may be completely incorrect, so please bash it
as
> you please :-)
>
> Suppose I placed 2 or 3 resistors (0.5W) at the input to the regulator, to
> step down the voltage from 12 to say 7 volts.
> In theory, current stays the same, but the voltage absorbed by each of the
> resistors and regulator is small compared to the voltage drop taken by the
> regulator alone.
> So, without going into detailed calculations, and knowing that power is a
> function of voltage and current, then, supposing that the regulator alone
> rose to a temperature of 80 degrees centigrade. Then, if the load was
split
> four ways (reg + 3 resistors) then the power dissipated by each would be
1/4
> of the total. Thus, the temperature rise should be, approximately, 20
> degrees on each component. This is VERY crude and does not take into
account
> ambient temperature etc...
> Put all this into a box, and then, the maximum temperature rise inside it
> would be 20 degrees plus the increase due to box insulation, but neve the
80
> degrees we had originally.
>
> Thanks again, all the best,
>
> Mike
>
> >

---

Mensaje original

---

> > De: orthner@s... [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=DOMT6_Hg7HWRHnJbWlmlbeNRSBNzZGobPgdHFA3su33rB9kt4YzhL2SKN6LxIh0TJN3KYGc2Y2AtH1kQ]orthner@s...[/url
> > Enviado el: martes, 25 de julio de 2000 14:16
> > Para: basicstamps@egroups.com
> > Asunto: RE: [noparse][[/noparse]basicstamps] 7805 regulator running hot!!
> >
> >
> > At 12:04 PM 7/25/00 +0000, Jason wrote:
> >
> > [noparse][[/noparse]snipped for brevity]
> >
> > >this is a good solution, but try the capacitors first if you
> > haven't already
> > >done so. I've found regulators to get overly hot and then cool
> > right down
> > >after I remembered the caps.
> >
> > I agree. Check your circuit for flaws or shorts. I'm running a 7805 at
> > ~300mA with 24V going in and a small heatsink - it barely gets warm.
Your
> > regulator shouldn't be running so hot that you can't touch it...
> >
> > Duncan
> >
> >
> >
> >
> >
> >
>
>
>
>

Archiver

Thanks to all who responded.

I have now done the same that Parallax does in the Super Carrier Board,
which is to place a 47uF electrolytic cap at the input and a tantalum 1uF at
the output. After adding this, the circuit seems to be running just as hot.
I've had it on for 12 hours since adding the caps, and the heatsink and
regulator can't be touched.
The rest of the circuit seems to be running just fine.

I have a theory now, which may be completely incorrect, so please bash it as
you please :-)

Suppose I placed 2 or 3 resistors (0.5W) at the input to the regulator, to
step down the voltage from 12 to say 7 volts.
In theory, current stays the same, but the voltage absorbed by each of the
resistors and regulator is small compared to the voltage drop taken by the
regulator alone.
So, without going into detailed calculations, and knowing that power is a
function of voltage and current, then, supposing that the regulator alone
rose to a temperature of 80 degrees centigrade. Then, if the load was split
four ways (reg + 3 resistors) then the power dissipated by each would be 1/4
of the total. Thus, the temperature rise should be, approximately, 20
degrees on each component. This is VERY crude and does not take into account
ambient temperature etc...
Put all this into a box, and then, the maximum temperature rise inside it
would be 20 degrees plus the increase due to box insulation, but neve the 80
degrees we had originally.

Thanks again, all the best,

Mike

>

---

Mensaje original

---

> De: orthner@s... [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=RFEykX7RMxyYMBD3ZWU0Pzni59cy_TjlBeZVyunuesPRI7oBAlld0Ko5qS_3kQhqU-sza3vw96FkTA]orthner@s...[/url
> Enviado el: martes, 25 de julio de 2000 14:16
> Para: basicstamps@egroups.com
> Asunto: RE: [noparse][[/noparse]basicstamps] 7805 regulator running hot!!
>
>
> At 12:04 PM 7/25/00 +0000, Jason wrote:
>
> [noparse][[/noparse]snipped for brevity]
>
> >this is a good solution, but try the capacitors first if you
> haven't already
> >done so. I've found regulators to get overly hot and then cool
> right down
> >after I remembered the caps.
>
> I agree. Check your circuit for flaws or shorts. I'm running a 7805 at
> ~300mA with 24V going in and a small heatsink - it barely gets warm. Your
> regulator shouldn't be running so hot that you can't touch it...
>
> Duncan
>
>
>
>
>
>

Archiver

Hello Miguel,

It sounds like you need to put (temporarily) a current meter inline with the
7805 output. You are obviously drawing far more current then you think you
are. Remove any major components, (i.e. chip, transistors, etc.) and recheck
the current bleed. It also sounds like you may in fact have a short across the
PC board. Find out what your "quiet" current draw is without the PIC.

Leonard Zerman

Archiver

Mike,
Some power calculations:

For max power calculations, you must know max allowed junction temp, and
thermal resistances, which are degree C rise per watt. My 7805 data book
for TO220 case lists max Tj (junction temp) =125 C, and TRjc (thermal
resistance-junction to case)=4 deg C/watt. Also, TRca (thermal res case to
ambient, no heat sink)=50 deg C/watt. Now we can do some calculations:

*No heat sink*, 14 volts in, 0.5 amp load, 25 deg C ambient;
14-5=9 volts across regulator. 0.5 watts*9V=4.5 watts reg dissipation.
Total TR (junction to ambient)=4+50=54 deg C/watt.
Temp rise, junction to ambient=4.5W*54 deg C/W=243 deg C.
Add ambient temp, Tjunction=243+25=268 deg C!
....spec is 125 C max so you need a heat sink.

*Infinite heat sink* This is a heat sink so large that the case is same temp
as ambient air.
The case = 25 C, so the junction rises TRjc above case.
4.5watt*4deg C/W=18 deg rise. Add case temp=25,
Tjunction=18+25=43 deg junction...no sweat.

This shows that a big enough heat sink will allow operation without a series
resistor.

*Now, try a series resistor*:
You should keep at least 2.5 volts across the 7805 for good regulation.
So you have 5V out + 2.5V=7.5V into 7805.
The resistor will drop 14V-7.5=6.5volts.
For 0.5 amp, R=6.5/0.5=13 ohms.
resistor pwr=0.5*6.5=3.25 watts.
Regulator pwr=2.5V*0.5amps=1.25 watts.
(notice we still dissipate 1.25+3.25=4.5 watts total).

Now check Tjunction, assuming no heat sink.
Total TR junction to ambient is 54 deg C/W as above.
Since the 7805 now only dissipates 1.25W,
1.25*54 deg C/W=67.5 deg C rise.
Add ambient of 25C,
Tjunction=67.5+25=92.5 deg C...no sweat.

This tells you that a series resistor would let you get away without a heat
sink. (For 25 deg air). You still have to dissipate the resistor heat, but
sometimes you can mount the hot resistor away from the electronics.

Hope this helps,
Ray McArthur

---

Original Message

---

From: Miguel Puchol <mpuchol@w...>
To: <basicstamps@egroups.com>
Sent: Tuesday, July 25, 2000 3:38 AM
Subject: RE: [noparse][[/noparse]basicstamps] 7805 regulator running hot!!


> The input voltage is 12 to 14 volts, current consumption should be no
> greater that 0.5A at max load.
> I will do some test, and will try your idea of placing a resistor at the
> input. I suppose that being an energy problem, dissipated heat will be the
> same, just distributed over two components.

Archiver

There is no reason for the Voltage regulator to over heat, If its working
according to the specifications. The 7805 v- regulator can produce 1 amp out
put and about 1.5 with the proper Heat sink...... Well if the regulator is
over heating you maybe pulling to much current from the regulator, You can
try adding a past transistor to the out put of the voltage regulator. this
will increase ur power handling capability ....try using the 2n3055
transistor it can dissipate about 115 watts...Why dont u place a amp meter
on the out of the regulator withe the circuit connected and see how much
current it drawing.......

Wendell

Archiver

nicely done,

but

0.5 AMPS*9V=4.5 watts

right ?

richard


> *No heat sink*, 14 volts in, 0.5 amp load, 25 deg C ambient;
> 14-5=9 volts across regulator. 0.5 watts*9V=4.5 watts reg dissipation.

Archiver

Thanks Richard, you are correct. At least I know someone read it[noparse]:)[/noparse]

Regards,
Ray McArthur

> 0.5 AMPS*9V=4.5 watts
>
> right ?
>
> richard
>
>
> > *No heat sink*, 14 volts in, 0.5 amp load, 25 deg C ambient;
> > 14-5=9 volts across regulator. 0.5 watts*9V=4.5 watts reg dissipation.