2
\$\begingroup\$

I have system that 98% of the time consumes less than 300mA. The remaining 2% it consumes 550mA. It is a USB system and works with a PC. I have looked at various options to stay withing the 450mA target power consumption but I couldn't bring it down. I cannot take risk with system malfunctioning due to longer cable or poorly designed USB power of a PC, therefore I was thinking of adding a battery with power path IC so that during the peak period the current can be augmented from the battery.

The down side of this is that the battery will always be in charging mode. I have no idea what this would to the battery. Any insights would be appreciated. I need this product to work for 3 years without a hitch.

\$\endgroup\$
  • 2
    \$\begingroup\$ How long can the 550mA operation last? If a few minutes or less, you may be within range of supercapacitors. \$\endgroup\$ – Brian Drummond Oct 27 '14 at 13:17
  • \$\begingroup\$ @BrianDrummond less than 100m sec. Super caps cannot deal with very fast ramps I heard. I will experiment with ECaps for sure before I take the plunge. \$\endgroup\$ – TGG Oct 27 '14 at 13:19
  • \$\begingroup\$ Then it's more or less 55mC, or 11mF@5V. It's not even a super-cap I'd say, just a fairly large one. \$\endgroup\$ – clabacchio Oct 27 '14 at 13:27
  • \$\begingroup\$ @clabacchio how do you calculate this? I couldn't do this. I would be grateful if you could point some light into this. \$\endgroup\$ – TGG Oct 27 '14 at 13:31
  • \$\begingroup\$ Well, 550mA*0.1s=55mC. Then, C=Q/V but of course as Brian points out, you need a larger capacity to limit the voltage drop. Let's say just 1V, then you need a 55mF cap. I think you can work it out from there. \$\endgroup\$ – clabacchio Oct 27 '14 at 13:43
2
\$\begingroup\$

Investigate the ESR (series resistance) of different supercaps. Some can take high currents, some (designed for memory backup) can not.

A 1 Farad capacitor (small for a supercap!) with low ESR could supply the missing 100mA for 10 seconds (losing about a volt) which greatly exceeds your requirement. So with suitable circuitry to limit the charge current they certainly store enough energy. Whether you can simply parallel it with the USB supply or need some switching arrangement to supply the extra power may depend on your voltage tolerance needs.

Searching Farnell for 1F capacitors shows 27 available with a wide range of ESR values :
0.5 ohm price £1.60 (or under £1 at 10 off)
0.031 ohms at a ridiculous price, probably useful for starting big motors!
and many with ESR of 30 ohms or higher.

I would consider the former (0.5 ohm ESR) with a suitable series resistor to limit charging current, and a Schottky diode across it (forward biassed when discharging) and ask if something as basic as this meets your requirements. Failing that, a MOSFET across the resistor switched by a brownout detector can give better voltage regulation, but with more complexity and expense.

EDIT : APOLOGIES : the highlighted capacitor is only rated at 2.5V. Further searching shows a more expensive 5V rated one to maintain simplicity, alternatively a boost regulator could generate 5V from this capacitor when demanded by a brownout detector.

\$\endgroup\$
1
\$\begingroup\$

You can always use an hystheresis to make sure that it will only charge when the battery voltage is below some threshold.

You would just need an hystheresis comparator that drives a series transistor, or even better a LiPo battery charger.

\$\endgroup\$
  • \$\begingroup\$ I am using a charger for sure. But battery will never go through the full charge/discharge cycle. It will always hover around 4.1V. I am woried this could cause batter life to shorten and battery can no longer hold charge. \$\endgroup\$ – TGG Oct 27 '14 at 13:16
  • 1
    \$\begingroup\$ @TGG What I meant was to use the comparator to turn on and off the charger itself (that I knew you were using :)). You can have the battery oscillate between e.g. 3.5 and 4 V, making it much less frequent. Anyway, it's recommendable to keep the battery voltage as much as possible to the middle of the range (around 3.7V) to maximise its lifespan. \$\endgroup\$ – clabacchio Oct 27 '14 at 13:18
  • \$\begingroup\$ I see. This may be a viable option. Do you have a sample schematic how I can solve this? \$\endgroup\$ – TGG Oct 27 '14 at 13:22
  • \$\begingroup\$ Also, what happens I don't do what you suggest. Just stick the battery and let it be charged to 100% all the time? \$\endgroup\$ – TGG Oct 27 '14 at 13:23
  • \$\begingroup\$ @TGG well I have an idea in mind, but seeing what charger you use and how much time your system is drawing 550mA would help getting a better view. Lithium-based batteries don't like to be at high voltages for long times, and they wear-out quicker. I don't know exactly why, but it's recommended to keep them in that - let's say - 20%-80% range. \$\endgroup\$ – clabacchio Oct 27 '14 at 13:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.