1
\$\begingroup\$

I'm building a circuit at the moment which involves switching a 500ma load from a microcontroller to balance a bank of batteries. This circuit runs at 12V nominal. I've built a PCB with 1210 chip resistors in a 4S8P type configuration (32 resistors) each running 51ohm for an overall resistance of 25.5ohm.

I've selected 32 Panasonic ERJ14 1210 1/2W resistors which should be dissipating around 180mw each, however these chips are very quickly getting very hot, measured in at over 120˚C in a 20˚C room. The data sheet lacks ambient temperature delta / load data so my question is, how does one determine the temperature rise for a particular chip resistor? And less directly, what are the physical limitations to heat rise from dissipating a certain amount of energy? In this example, how hot is a particular surface area going to get to dissipate 6W?

\$\endgroup\$
1
  • \$\begingroup\$ How much copper are the resistors attached to on the board? \$\endgroup\$
    – Colin
    Commented Apr 16, 2020 at 8:10

1 Answer 1

Reset to default
1
\$\begingroup\$

First: A rise in temperature of 100K seems quite a bit for a 1210 resistor with 180mW net load. Did you measured the load or is it just calculated? Is the resistor packed in an additional package (housing or such) or is it operated at free air?

Second: The resistor has three ways to lose his dissipated energy:

1) conducted over mounted board - which for his parts loses the energy over 2) and 3)

2) Air convection

3) Radiation

It is not that easy to determine how hot which surface gets under given circumstances. For a simple rule of thumb obviously we can say: Make the part bigger if the part is getting to hot.

\$\endgroup\$
2
  • \$\begingroup\$ Thanks for your reply! Load is calculated and the board is currently just sitting on a bench. The resistors are in free air on a board with no thermal considerations (no thermal vias, etc). The 1210's in an an array of about 560mm^2. Will a larger component (say I 1W component) produce less heat provided we maintain the 180mW net load? \$\endgroup\$
    – Alex Turner
    Commented Apr 16, 2020 at 4:55
  • \$\begingroup\$ I would measure the actual load first to ensure the circuit is running properly. If you say the 1210s are near to each other and dissipate 180mW each? Then 100K for the ones in the mid are rather plausible. \$\endgroup\$
    – NoiseEngineer
    Commented Apr 16, 2020 at 5:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.