CC430 Hardware Question (Separate VCC)

goodarz kashian said:

i want to separate the battery used for general MCU application and RF application in my hardware.

To answer properly, it is good to know - why you need this?

Thx for reply

I am using the CC430 MCU in a metering application and the lifetime of the meter should be guaranteed

to be a minimum of 10 to 15 years. 

I achieved the minimum Average current of 20uA to 50uA  for main MCU applications (without RF Communication)

but for a 10dBm Rf  transmission  the current consumption is 30mA (in transmit mode)

and (15 mA in Receive) or (2mA in RF Idle Mode)

although i used the Wake On Radio for RF operation

but the average current is still high (Not low enough with my current battery i already have chosen)

since my RF port is not a vital feature of my design

i decided to separate the RF and the Main MCU Power (use separate batteries)

with separate batteries: if the RF Battery is reached to its critical capacity the meter still works.

and the RF battery could be replaced during the lifetime of meter.

goodarz kashian said:

lifetime of the meter should be guaranteed to be a minimum of 10 to 15 years.

What kind of battery you are going to use? (just curious)

goodarz kashian said:

although i used the Wake On Radio for RF operation

Don't. You shall shut RF completely down, including it's oscillator. Radio standby current specified is 0.2 uA, msp430+RF part combined standby current is 1uA (checked datasheets especially for you). 

Ilmars said:

What kind of battery you are going to use? (just curious)

There are some types of battery which have a large energy density called "lithium-thionyl chloride" 

or simply "Li-SoCL2"

Wikipedia says:

Lithium-thionyl chloride batteries are generally not sold to the consumer market, and find more use in commercial/industrial applications, or are installed into devices where the consumer does not replace them. The cell contains a liquid mixture of thionyl chloride (SOCl₂) and lithium tetrachloroaluminate (LiAlCl₄), which act as the cathode and electrolyte, respectively. A porous carbon material serves as a cathode current collector which receives electrons from the external circuit. Lithium-thionyl chloride batteries are well suited to extremely low-current applications where long life is necessary, such as wireless alarm systems.

they have a very low internal resistor and do not discharge when left open (in a good condition)

(or to be more exact they have a very low self discharge rate)

there are many capacity of them available in the market (Just google it)

Ilmars said:

Don't. You shall shut RF completely down, including it's oscillator. Radio standby current specified is 0.2 uA, msp430+RF part combined standby current is 1uA (checked datasheets especially for you).

that is been so kind of you (thank you)

but this meter act like a client(node) in a network and there is another high power RF Server that call these nodes

and asks for data.

Wake on radio periodically shuts the radio and put it in the receive state: (Switch between sleep and idle-receive mode)

if i shut the Radio core completely there are 2 issues:

1 - The time between  deep sleep and idle in RF Core is considerable (because of some initialization overheads and oscillator stabilization )

2 - when the Nodes(meters) are in sleep state they are not available in network.

i want the meters to be available as long as they have a good battery.

so do you think separating the RF Core Power and the other parts is not a good idea ?

Sorry i forgot to say that the sleep mode i use is : LPM3

in this mode the XTALL 2 (RF XTALL) is shut down automatically.

thank you :)

goodarz kashian said:

but this meter act like a client(node) in a network and there is another high power RF Server that call these nodes and asks for data.

Worst possible way of building low power sensor network - to let sensors continuously receive. Ever heard of synchronized network? - Timers of all nodes are synchronized and each node have it's unique "timeslot" to wake-up and quickly transmit it's data to (mostly listening) master. After transmission node briefly listens for master answer if any - to request more data or to update node clock sync or configuration.

goodarz kashian said:

so do you think separating the RF Core Power and the other parts is not a good idea ?

I simply cannot get what you are going to "fix" such way. Please explain.

goodarz kashian said:

but this meter act like a client(node) in a network and there is another high power RF Server that call these nodes

Usually, people think that sending consumes most energy. Which is wrong, Yes, sending requires a high peak current, but only for milliseconds. Constantly receiveing only consumes a fraction of it, bu tfor way longer time. Even with Wake on Radio, it is most likely much more than sending takes.

You should at least minimize it by using some sort of time slot mechanism. Once you have received a message form the server, you know when to go online for listen and when not. So go to sleep., wake up when time is due (and adjust the timing by the incoming message) and then go to sleep again. Just waking up in the hope of catching something is waste of energy.