MSP430 Powered by Li-Ion Not Possible

Hi George,

Although we don't offer a MSP430 which can directly accept a supply voltage of 4.2V, here are several resources about how to use either a LDO or DC/DC converter to provide the appropriate supply voltage to the MSP430. If minimizing power losses during conversion is required for your application, you'd want to select a device with high efficiency.

If your application only involves a MSP430 and has strict size requirements, I can understand your concern about needing an additional power part. However, many devices (other than the MSP430) in common applications such as an accelerometer, LCD, LED, etc. will require a 3.3V rail anyway. Obviously, the power part(s) will be selected based on your system-level power consumption at each voltage level.

Hope this helps.

Regards,

James

MSP Customer Applications

Hi James,

You bring up a good point about the efficiency. I was looking at their datasheet and in section 18 (page 167) they have some graphs showing current consumption vs. Vcc. Picking an 8MHz operation I see that at Vcc=4.0V, Icc=1.75ma while at Vcc=3.3V Icc=1.4ma. Looking at the idle current it's ~0.4mA and ~0.3mA.

How would this compare to the MSP430G?

300mA is more than enough current for this application so the TLV733P could work although if I was to use a regulator I probably would lower the Vcc to try and save more power if the number crunching worked out that way.

Thanks,

George

Hi George,

Thanks for the additional details. The MSP430FR2110 16 MHz Ultra-Low-Power Microcontroller With 2 KB FRAM, 1 KB SRAM, 12 IO, 8 ch ADC10 is an excellent option and has similar power consumption characteristics. This device is small, cheap, and allows seamless migration from MSP430G2xx devices.

For the power consumption values, these depend on many different things like operating temperature, execution memory, I/O configurations, etc., so I'd recommend taking a look at Sections 5.4 through 5.11 in the datasheet for a detailed break-down. At VCC=3.0V, 8 MHz operation, 85C, execution from FRAM, 0% cache hit ratio, the typical current consumption for Active Mode is 2.73mA. For a 100% hit rate and 25C, the typical current consumption is 0.57mA. Section 5.5 defines a value of 120uA/MHz as well.

For LPM0 (see Table 1-2 in the User's Guide for details about each mode) at 3.0V, the typical current consumption is 0.3mA. For LPM4.5, the typical current consumption can get down to 0.034uA at VCC=3.0V.

If you need capacitive touch for your application, some of our MSP430FR25xx/FR26xx devices feature CapTIvate technology.

Regards,

James

MSP Customer Applications

Hi James,

Thanks, I really do appreciate the time and effort you're putting in to answer my concern. I initially looked at the FRAM offerings since I figured they'd have the lowest power consumption. Unfortunately I came across some downsides. The first is that nobody has stock on the  MSP430FR2110. When comparing cost, the  MSP430FR2110 is at ~$0.50 and then I need to add a regulator which adds another ~$0.10 so the cost is ~$0.60 compared to the $0.29 of the ATtiny10.

I would love to use a MSP430 in this product as I have the tools and am familiar with the MSP430 devices. With everything being so cost conscious these days I'm not seeing how it's going to be possible. The MSP430G2210 is ~$0.35 so it's close to the ATtiny10 cost but would a MSP430G2210 + he TLV733P be equal to or better than the power consumption of the ATtiny10? The only chore of the MCU is listening to the power switch and changing the brightness of the LEDs by changing the duty cycle of the PWM signal.

To be honest I tried to come up with a way to accomplish this without a MCU. If I don't have a MCU then I don't need to incur the cost of programming the MCU in production. But just about anything I could think of ended up costing more that just buying a $0.30 MCU, sad but true.

Thanks,

George

Hi George,

I'm happy to help. Unfortunately, it doesn't look like a MSP430G2210 + TLV733P solution could provide better overall power consumption than the standalone ATtiny10, but thank you for considering a TI solution.

Your idea of accomplishing this goal without using a microcontroller is interesting. After looking through our LED Driver portfolio, I found some (like the LM3633) that offer an "analog dimming" feature, but this approach still seems to require a PWM input. I found the TPS61161 which is more efficient than the AP5724, but it may be more expensive. Since I'm not an LED Driver expert, I'd recommend asking the folks over in the LED Driver Forum if there's a way to do this without a microcontroller.

Regards,

James

MSP Customer Applications

Hi James,

Thanks again. I always do go to TI first for all of my needs but for some reason this project isn't a good fit. Yes, the TPS61161 is more than double the cost of the AP5724. All of the LED drivers that I've looked at either want a PWM input for dimming or an analog voltage change. I might post on the LED forum and see if anyone else has come across requirement.

Best Regards,

George