Hello,
There was a conversation on a similar question, but since specifics were different, I would like to ask it all again.
So I have a (pro mini based 3.3 v circuit) running on 3.7 v 360 mA battery. I have another circuit - a two stage crystal run radio-transmitter, that runs on 1.5 V, 160 mA, Silver Oxide battery. I would like to connect these two circuits, to run in sequences on the same Lithium battery. It is very important to me to keep the drain of the circuit low to extend the battery life as much as possible. And I can't use a larger battery, due to weight constrains. I am planning to program the second circuit to begin running closer to the end of the battery life, when voltage may be dropping. What shall I use to reduce the voltage from 3.7 V to 1.5 V? Shall I consider reduction from a lower voltage point, since I plan to turn the second circuit when that battery will be running out of power?
If you use a linear regulator, the efficiency will be 1.5/3.7 = 40%. However, the quiescent current will be low.
I think I'd look for a switching regulator that has a shutdown pin. Only activate it when the transmitter must operate. Then efficiency can be in the range of 90%.
What battery or cell is this? Lithium cells are pretty particular about not being drained beyond a certain point. But you'd need to know the cell chemistry.
As the transmitter operates on silver oxide batteries, the power must be really low. Buck converters may simply not work at such low current draws, or their quiescent current is more than the device takes...
Many linear regulators have an enable pin as well, and can go very low in quiescent current. The MIC5225-1.5 is one: 29μA quiescent, 0.1μA in shutdown mode. Or the RT9012-15, 25µA quiescent and 0.7µA in shutdown mode.
What will be managing the 2nd, Transmitter circuit? The Pro Mini?
a. If so, then, won't they both need to be powered?
b. If not, is their significance to us knowing anymore about what's going on?
Transmitter, using 160mA, on a Silver Oxide coin cell? I didn't even think 160mA could be delivered by such a cell. Am I to assume these are infrequent bursts, and perhaps, that a capacitor is supplying the instantaneous current?
Would it, perhaps, make sense to employ a switch mode solution, with two outputs: one at 3.2V [3.0V] and another at 1.6V [1.5V] (power of 2 voltages are easier to produce than, say 3.3V and 1.5V).
How about this: a 1.5V switch mode supply to run the transmitter, and run the ProMini from the battery, with a series diode, biased to have a forward voltage no lower than 400mV [Nominal] -- thus, cutting the 3.7V down to 3.3V...like this:
If at least 15µA is made to flow through the 1N914, there will always be at least 300mV, and 400mV nominal Forward Voltage on the diode, thus it will charge the Capacitor up to around 3.3 to 3.4V. Once the capacitor is charged, the only load on the battery, from 220k bias resistor,will be around 15µA, which will have little impact on batteries charge level. By, managing power use, in the Pro Mini, by lowering it's clock speed, and keeping it in sleep mode as much a possible -- waking it only to do whatever it will be doing. Here's an article on conserving power in an Arduino:
Also, since the ProMini, run at 3.3V, will draw a nominal 4µA, that bias resistor could be changed to 300k, which will expend even less power.
The value of C1 will depend on the overall power demand of the Pro Mini, which depends on how it will be applied. But, a 470µF might be a good ballpark assuming 10mS of activity every 5 or so seconds, at around 4mA current demand form the ProMini [a demand that will go up if the ProMini is driving anything significant].
What is the idle current [quiescent current] of the transmitter? I'm guessing it must be low, if this can be run off a Coin Battery, but if not, then connecting the SMPS [Switch Mode Power Supply] directly to the Transmitter, might not comply with the energy conservation stipulation. In which case my circuit would need a modification.
No need for that diode. The 3.3V Pro Mini runs perfectly fine on 3.7V - at least my cheap made-in-China clones do. After all it's the same processor as is in the 5V Pro Mini, just slower clock and different regulator (for if you power it at higher voltage). They even run perfectly fine at 5V when connected to the USB - then the Vcc pin actually produces 5V, which may cause other problems!
wvmarle:
No need for that diode. The 3.3V Pro Mini runs perfectly fine on 3.7V - at least my cheap made-in-China clones do. After all it's the same processor as is in the 5V Pro Mini, just slower clock and different regulator (for if you power it at higher voltage). They even run perfectly fine at 5V when connected to the USB - then the Vcc pin actually produces 5V, which may cause other problems!
Good point! Dang! I wanted so much to maintain my Bad A$$ MacGyver Hobbyist image!!
allanhurst:
I would avoid switchmode psu's in any rf application - the switching noise can cause big problems.
Allan
Doesn't that depend on the type of RF application? For instance, if it's simple "Amplitude Shift Keying" [ASK], then I doubt there will be a problem -- especially with a "crystal controlled transmitter" with little or no frequency deviation due to supply noise [i.e. high power supply noise reduction].
You might get away with it. The OP doesn't give any detail of the transmitter - might be FSK. Power output < -6dBm if in Europe in the 433 and 866MHz bands..
Given the idle current of a little swmode psu, compared to the linear reg losses, is the risk worth it?
The battery lasts a little less - tough. Do some sw tricks.
I've had a lot of trouble meeting ETSI specs with swmode psus. Adjacent channel noise is the big one. This would be a problem even with ASK.
But does an arduino hacker care? Is he/she going for type approval?
I might meet you at an accredited test facility, and watch you get the checkbook out.......
But probably won't.
Here in the UK in principle the RA could prosecute for an out of spec transmitter, but with such a low power level in an unlicensed band , no-one is likely to complain.
I'm not a fan of SMPS but I've recently been reading about some µConverters. I have not had any expereince with them but the specifications look interesting.
That's a really nice little device! 600 mA output is not too shabby, can indeed replace just about any linear regulator. The only downside is the cost, pretty expensive at about 20x the price of a standard LDO linear regulator.
polymorph:
I think I'd look for a switching regulator that has a shutdown pin. Only activate it when the transmitter must operate. Then efficiency can be in the range of 90%.
I'd avoid SMPS for a transmitter circuit, could be too noisy.
