0
\$\begingroup\$

I am planning a hand-held design (similar to a remote control) using an STM32F030. It is low cost and has good low power options. There are alternative devices on the market which are pin compatible and if the design goes into higher volume, they are lower cost.
Now I am wondering which ones to use because their power specs are different. The STM32 has a higher active current, while APM32 has a higher standby current. Just for argument's sake let's say STM is 22 mA while APM is 14.5mA running at the same clock speed. In my target low-power mode (Stop-Mode), the STM is specified with 5 uA while the APM has a spec of 9 uA. My simple calculation is 99 x low power number + 1x power number for run mode. In this case the APM would last much longer. What else do I need to consider?
If I turn on the MCU 10 times / second to monitor external sensors, basically on - ADC conversion - off, is the startup time important in the power calculation?
What am I missing - if anything? Any experience with APM32? Thanks for your input

\$\endgroup\$
4
  • 1
    \$\begingroup\$ RobertT - Hi, What type of battery did you mean? Some site members think you meant Lithium-ion rechargeable (and so the question should have that tag), whereas others think you meant your original tag of Lithium i.e. primary cell. Since we don't want to have a tag war :) please clarify. Thanks. \$\endgroup\$
    – SamGibson
    Commented May 5, 2023 at 23:49
  • \$\begingroup\$ Startup time varies a lot depending on clock source etc. \$\endgroup\$ Commented May 6, 2023 at 0:04
  • \$\begingroup\$ Clock accuracy is not of highest priority, internal High Speed oscillator is the target clock source. \$\endgroup\$
    – RobertT
    Commented May 6, 2023 at 23:09
  • \$\begingroup\$ DETAILED battery spec needed. Chemistry, mAh, Voltage, form factor. & brand and model IF known. \$\endgroup\$
    – Russell McMahon
    Commented May 7, 2023 at 12:46

1 Answer 1

1
\$\begingroup\$

Something to start with is efficiency. You're going to lose a certain amount of power to the regulation of voltage(s) to run your MCU. The STM32 needs 2.4-3.6V, so unless you can find a Li battery that gives you that directly you'll need to regulate it down and lose some power there.

Battery life = Capacity / Load current
This only holds up for as long as your load is significantly greater than the self-discharge of the battery. For a Li-On of 100mAh, discharging at (0.01 x 14.5mA) + (0.99 x 5uA), I get 444.54 hours, 18 days. If we roughly assume that we're going to lose around 5% in the first 24 hours, then 2% per month then our answer should be about 6% out in this case, not terrible but possibly significant.
I do notice that Idd is typically 5uA in stop mode, but this is not a guaranteed figure (hence typical), you also need to add 1.7uA for VDDA

Figuring out your uptime is a bit difficult. For the STM32, wakeup from stop is 2.8-5us, though I couldn't find how long to put it into stop mode. If you're up time is 1%, with 10x waking per second, you're expecting a wake cycle of 1 ms, so this represents 0.5% (1% if we assume time to sleep == time to wake). You decide if you want to include it or not, but I probably would.
How long it will take to perform the action might be relatively easy to figure out, or complicated. It depends on what you're doing.

It's a great, or even essential, idea to get a development board or similar and do some testing. Doing the maths is a good start, and will help you know if your idea is viable, but there is no substitute for getting a physical device and measuring - things like the time to sleep that aren't listed in the datasheet. How much power a device will draw always depends on utilization. If you were making a commercial product and really wanted to get fine details, you might go as far as to rig up a current monitor, so you could answer questions such as "How does the current vary while powering up and down".

\$\endgroup\$
2
  • \$\begingroup\$ I followed your calculation and agree to the estimate for the STM. The APM32 comes out to 649h and I like that number better :) with 27 days I approach my desired target of 1 month. Turning it on 5 times / second should get me over the typical 1 month. \$\endgroup\$
    – RobertT
    Commented May 7, 2023 at 4:08
  • \$\begingroup\$ I am very much aware that the typical standby current assumes room temp and if the temperature goes up, so does the standby current. Has anyone used the Geehy APM32 and has gathered good or bad experience. Availability is better, power is better, CLK comes up faster and pricing is lower too!? Why not using the APM32? \$\endgroup\$
    – RobertT
    Commented May 7, 2023 at 4:15

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.