When developing a new automotive electronic component, the component needs to be certified to withstand certain temperatures. There are several standards that list methods of testing, such as ISO 16750-4 and AEC-Q100. However, these standards list a number of 'grades' or 'temperature ranges', and it seems to be up to the reader of the standard to determine which grade applies.

In ISO 16750-3, which deals with mechanical loads, each requirement states the purpose, such as "These tests are applicable to equipment to be mounted directly on the engine" or "This test is applicable to devices under test intended for mounting on sprung masses". For temperature ranges, I can find no such guidance.

One might say that it is up to the customer to specify the required range, but how is the customer to know? Googling the problem doesn't help, as all web pages do mention 'extended ranges' such as -40°C to +125°C, but omit how they arrived at that number.

How do I determine what the temperature range should be for a new component to be developed?

  • \$\begingroup\$ It's up to you to decide what temperature ranges you're going to design your products for, there's no fixed formula and it's really down to what uses you're attempting to target as well as marketing and cost issues. There's a big difference between something that has to work in someone's living room and something that has to work in a satellite. \$\endgroup\$
    – Finbarr
    Apr 26, 2017 at 9:23
  • \$\begingroup\$ The most obvious step for me is to figure out where tat part will live and what the temperatures are there. Rooftop and exhaust manifold are two very different places. \$\endgroup\$
    – PlasmaHH
    Apr 26, 2017 at 9:40
  • \$\begingroup\$ Rooftop and exhaust manifold are two very different places. I understand. But for mechanical loads this same issue holds, and there it is defined quite precisely in the standard. \$\endgroup\$ Apr 26, 2017 at 9:43
  • \$\begingroup\$ Are we talking about part-level (resistors, capacitors, etc.) or module/subsystem/component-level temperature ranges? \$\endgroup\$ Apr 26, 2017 at 10:12
  • 1
    \$\begingroup\$ In North America, car manufacturer homologates their owns vehicle so that they can go on the market. That means they define their own standards of security, therefore the grade of their components. To have the right to homologates a vehicle, the manufacturer must comply to a set of requirements, but once they earn that right, they can do the homologation as they want. Of course, if something goes wrong, they needs to prove that they did due diligence and failure to do so leads to very strict sanctions. In the end, the customer (car manufacturer) will define his own requirements. \$\endgroup\$ Apr 26, 2017 at 10:56

3 Answers 3


The answer is given in Annex A of the ISO standard on climatic loads in road vehicles (ISO 16750-4):

ISO 16750-4 Annex A

For every mounting location, an operating temperature range is recommended. A dashboard, which is "exposed to direct solar radiation", receives a recommended code G, which translates into a range of -40°C to +90°C.


One might say that it is up to the customer to specify the required range, but how is the customer to know?

Well, the customer has the whole picture and is the one who knows better from its own thermal analyses.

If the component is to be mounted in a place where the thermal interface or the ambient will yield a high temperature (i.e.,located near the engine, etc.) the customer might want to specify a tougher requirement ("grade").

These are the grades according to AEC-Q10, as you probably know:

enter image description here

Temperature grades might be already standarized as seen above, but this doesn't mean that the design engineers of your customer are freed from carrying out their analyses and telling you which grade should your component be in order to be used in their assembly. One motor may run hotter than another, even under the same external environmental conditions, due to specification or design differences.

If your component is not project-specific (i.e., no customer specification applies), then you might want to design it for the wider operating temperature grade actually feasible, so to be able to market it to as many potential customers as possible. But you have to balance that with costs as well.

  • \$\begingroup\$ "from its own thermal analyses" - it doesn't make sense for each manufacturer to make its own analysis of what the temperature in the cabin will be when it's parked at noon in the Nevada desert - I presume that it will be similar for most vehicles. \$\endgroup\$ Apr 26, 2017 at 9:47
  • \$\begingroup\$ @FrankKusters In the aerospace world, thermal analyses are fairly ad-hoc, so as not to over-specify the components/subsystems. Maybe the expected temperature range inside in a car cabin can be standardised as it's due to external environmental conditions, but won't the temperature of components close to the engine be almost entirely depend on the thermal design of it under operating conditions? The best shot at standardisation here would be to specify a temp range and then rely on the thermal design of the top assembly to guarantee it. Is that the case? If yes, then you have your answer. \$\endgroup\$ Apr 26, 2017 at 10:02
  • 1
    \$\begingroup\$ It might not make sense to you but that's what happens. Who's going to do independent research into how hot a car gets in the Nevada desert and publish it? \$\endgroup\$
    – Finbarr
    Apr 26, 2017 at 10:02
  • 1
    \$\begingroup\$ I'm sure there are as lots of aspects of car manufacture that have standards that can be applied and lots that don't. Who decides how fast a car should be able go and how much fuel it consumes? \$\endgroup\$
    – Finbarr
    Apr 26, 2017 at 10:30
  • 3
    \$\begingroup\$ We used -40C to 85C for most of our parametric testing, with the upper number climbing to 105C or 125C for engine bay operation. The numbers came from the customer (Ford, GM, etc.) and make sense when you think about margin, and picture a car baking in the AZ sun, or sitting outside in the MN winter. \$\endgroup\$ Apr 26, 2017 at 15:49

It's really up to YOU, or whomever else is dictating the product requirements, to decide which environment the thing must...

  • A. Survive in, and
  • B. Remain functional in.

Those may be different values. For example, a standard LCD is not much use below 0C since the liquid becomes too sluggish to react to the applied voltages. However it will work when it warms up. If however you take it down to -40 the liquid can actually freeze and destroy the unit.

However, the above range is also dependent on installation location. Direct sunlight, proximity to other heat sources, poor ventilation, etc. can drastically change those numbers.

What you enclose your electronics inside also matters. The electronics inside an unventilated black box left on the dashboard in the sun in Death Valley, USA, may reach temperatures hot enough to melt some devices.

In addition, you ALSO need to worry about the enclosure itself. If it is that black box I mentioned above, it has to be made of a material that will not melt, crack or degrade in sunlight. It must also be constructed in such a way that it can expand and contract without stress failure.

Sometimes extreme conditions are required. Military specification tend to be much more extreme than domestic, but wider examples exist. For example, devices to be used outside in Antarctica, or a device that will be installed next to a blast furnace. In those cases extraordinary measures need to be taken. Heaters need to be added to the box for the cold example, and liquid cooling piped to the latter from a more distant location.

For automotive applications it really depends on what the device does. If it is critical to the safe operation of the vehicle, e.g. a brake sensor, it needs to work better than the low levels of the standards, perhaps as low as -60°C. If it is that little display the kids watch "Sponge Bob" on in the back seat, 0°C to +50°C may be sufficient.

The automotive "standards" indicate...

enter image description here

Where cost goes from high for Grade 0 to low at Grade 4. You have to decide which one of those you need to beat for the functionality of your design and to be within your cost budget.

  • \$\begingroup\$ He probably meant that. -40 to +125 C is Grade 1 according to AEC-Q100. There's even a Grade 0 that goes up to 150 C for maximum temperature. \$\endgroup\$ Apr 26, 2017 at 16:52
  • \$\begingroup\$ @EnricBlanco could well be. Still Grades 0 and 1 are pretty extreme. Your electronics would need to be pretty close the to engine or exhaust to need to withstand those kind of numbers. \$\endgroup\$
    – Trevor_G
    Apr 26, 2017 at 17:02
  • \$\begingroup\$ Probably grades 0 and 1 are for some part-level specifications, and grade 2 and higher for module/assembly-level specifications. I've seen this all the time in space electronics. \$\endgroup\$ Apr 26, 2017 at 17:08
  • \$\begingroup\$ @EnricBlanco yes the spec is not very clear. \$\endgroup\$
    – Trevor_G
    Apr 26, 2017 at 17:09
  • \$\begingroup\$ "You have to decide which one of those you need to beat for the functionality of your design and to be within your cost budget." That is exactly what I'm asking this question for. Is there a standard mandating for example if a dashboard component needs to be grade 4, 3 or 2? Or an engine compartment component needs to be grade 1 or 0? \$\endgroup\$ Apr 26, 2017 at 18:28

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.