"Safe" temperature for LM7805CV

[soubitos]

I am testing my tp4056flexadv module which works quite well when powered via USB (as expected, its a typical TP4056+protection configuration) but i am concerned about the LM7805 temperature... at VIN 7.5Volt and with charging current set @0,5A, i get readings on the regulator surface as high as 70-75 degrees C and with VIN of 12V it raises up to 95-100 degrees C.

What is a "safe" temperature for the LM7805 (I am even using the CV variant which should be able to handle 1,5A max current, more than enough for this application)
I have soldered the regulator to the mounting hole on the PCBand its also grounded so it should work as a heat sink but i dont see any transfer of temperature to the PCB except directly around the hole.

The idea behind this module is to be able to use a variety of input sources and offer both battery output and boosted output on demand so it can serve as a real general purpose lithium battery charger/power supply for applications etc.

I could possibly test every charging current selection from 60ma-1A with a variety of input voltages and provide a safe settings chart but even for this, i would need to know what is considered a "safe" temperature for the LM7805 to operate... note that my current ambient temp is around 24C so operating in hotter environments or insides cases etc would also make things harder....

Should i go for another linear regulator and which one (needs to be low cost) or should i change design altogether and go for a step-down switching regulator??

I also get a significant voltage drop at "higher" currents over 250mA
Is it possible my LM7805CV is a fake??? Unfortunately all i got in hand are a few 7805CV from ST which i dont know where i got them from, they might be laying around for several years...

https://easyeda.com/catech75/TP4056FLEXADV_v2_10-1636385dbe504ca9a2cfd3a7722e5bc7

[danadak]

The datasheet has -

1) Max allowed Tj

2) Thermal R Junction-Case and Junction-Ambient

So compute max allowed Pdiss based on worst case Ambient
and your thermal path resistances.


https://forum.allaboutcircuits.com/threads/thermal-resistance-calculations-for-mosfet-help.40217/


file:///C:/Users/user/Downloads/application_note_calculating_junction_temperature_from_thermal_resistance_an027.pdf


http://www.ti.com/lit/an/slva462/slva462.pdf



Regards, Dana.

[T3sl4co1l]

You want to be at or below the minimum thermal limiting temperature, under the worst case condition (max Pdiss, ambient T, Rth, etc.).

Tim

[mariush]

Read the datasheet of your chip, it says there.

Here's a datasheet from ST : LINK

Go to page 5, and you'll find there:

Code: [Select]

Top Operating junction temperature range

for L78xxC, L78xxAC 0 to 125 °C
for L78xxAB -40 to 125°C

Table 2: Thermal data

Symbol Parameter D²PAK DPAK TO-220 TO-220FP Unit
R?JC Thermal resistance junction-case 3 8 5 5 °C/W
R?JA Thermal resistance junction-ambient 62.5 100 50 60 °C/W

So your regulator will dissipate the excess energy as heat. If you have 7.5v in , 5.0v out then at 0.5A your regulator will dissipate (7.5v-5.0v ) x 0.5A = 1.25w

Now if you look at the table above, you can see that a TO-220 regulator will heat 50 degrees C over ambient temperature with every watt of heat produced, and your regulator will produce 1.25 watts, so it's only natural that your regulator will reach ambient temperature + 50 degrees C without any heatsink.
Circuit board is not very good at dissipating heat, if you put DPAK regulator on maybe 3-4 cm of copper, maybe the regulator will dissipate 1w and stay below around 90 degrees C

So you need a heatsink and you can easily figure out how big of a heatsink you'll need.

First, pick a suitable ambient temperature. It won't necessarily be room temperature, because you may have your regulator in a closed box or a box with poor air flow, not always in the open air.
So for this example, let's say the ambient temperature in your product will be up to 40c
The regulator says it will function up to 125c - INTERNAL - temperature , which is NOT the temperature of the heatsink.

For this example, let's say we want the regulator to not go over 100c and your ambient temperature could be up to 40c and you want the regulator to handle up to 1A with 7.5v input.
That means you need to dissipate (7.5v - 5v) x 1A = 2.5 watts

So :

Pd = 2.5 watts
Rjc = 5°C/W (from semiconductor manufacturer) - for the TO-220 package
Tj max = 125°C (from semiconductor manufacturer) - but we're gonna use 110c for safety
Ta max = 40°C

You have 110c (maximum temperature you want) - 40c (ambient temperature) = 70 degrees C to play with.

Quick rough check if you need heatsink or not : If you dissipate 2.5w, then the temperature increase on the regulator must be below 70c / 2.5w = 28c/w - this number is less than 45w  ( 50c/w -  normal temperature rise for TO-220 -  MINUS 5c/w  - junction-case thermal resistance for TO-220) therefore we definitely need a heatsink.

So now the maximum thermal resistance of the heatsink without any forced air will have to be this calculated value, minus the resistance between the die and the metal of the chip (junction-case temperature) and minus the thermal resistance of the thermal paste or thermal pad between the chip and the heatsink, which is generally small, usually under 0.5c/w:

Rsa  = (Tj max - Tj amb ) / Pd  - Rjc  - R?paste

So you need a heatsink with less than  Rsa  = (110 - 40) / 2.5 - 5  - 0.5 = 22.5 c/w

Your heatsink must have a thermal resistance below 22.5 c/w in order to keep your regulator below 110c at 1A of current with 7.5v in and 5v out.

Here's a heatsink that would fit these constraints, barely, with 21 c/w  :  https://www.digikey.com/product-detail/en/assmann-wsw-components/V7237C/A10753-ND/3476145

The price differences are so small that you'd probably just want to go with something a bit bigger and much better unless you're space constrained.

For example this one has a resistance of only 13.6 c/w : https://www.digikey.com/product-detail/en/aavid-thermalloy/575002B00000G/575002B00000-ND/126063


edit : Check the PDF below for more explanations and formulas and stuff  (extract from Aavid Thermalloy catalogue, they make heatsinks of all kinds)

[soubitos]

Thank you all, esp. Mariush for your response.... It has been a long long time since my last projects and i seem to have forgotten a lot about reading datasheets and doing my math right.... On the other hand, i was expecting the regulator to heat just not THAT much... with temps of 100C at 12V/0.5A inp it makes no sense to pursue this any further with this type of regulators, i think i will be better of to redesign the module with a switching regulator instead... the extra cost and bulk of a heatsink for the LM7805 can go on the added parts of a step down switching desing and the pcb area should not change significantly if at all.

What made me stop and start checking this further was that with the modules powered by USB i draw steadily 0.89A@5.12V the TP4056 sits happy at around 40C while powering from the DC jack i cannot draw more than 500-550mA@7,5-12V input while the output voltage of the 7805 dropped even 800mV under 5V when drawing anything over 250mA. Could it be that the 7805 is fake (or does it have thermal protection, i dont remember 78xx series having thermal protection but like said, i was away from electronics for many many years and need to catch up.)