SparkFun Forums

[UNTEngineer]

Hmm, I wonder if I should provide both a Li battery backup for the entire circuit, and a capacitor backup for just the RTC.

Anyways, thanks for the calculation. It'll go a long way.

[UNTEngineer]

Can you find a 1F supercap in SMT at that voltage rating (w/margin) ? I didn't think you could do even 50% of that.

I was going to use a 1F at 5.5V, but all I can seem to find is through-hole radial caps. Having a hard time finding any SMT Supercaps at 1F.

[MichaelN]

I was going to use a 1F at 5.5V, but all I can seem to find is through-hole radial caps. Having a hard time finding any SMT Supercaps at 1F.

Even if you could get them in that rating, I’d be hesitant to use SMD supercaps for a couple of reasons:
- They would be more heat sensitive than other components, so you’d need to be very careful soldering them.
- I don’t like using “large” SMD compoents that are only supported by a couple of small pads. The weight of the component can pull the pads off due to shock or vibration. SMD connectors are the worst (yes, I’ve seen this happen), and I tend to avoid them like the plague.

[Mee_n_Mac]

And the above should raise the question that the OP needs to have answered up front ... what's the trade study for using SMD supercap vs good ole Li battery backup ? I can understand the desire to avoid the space and cost of the normal Li charging circuit but does that still hold up ?

[UNTEngineer]

Even if you could get them in that rating, I’d be hesitant to use SMD supercaps for a couple of reasons:
- They would be more heat sensitive than other components, so you’d need to be very careful soldering them.
- I don’t like using “large” SMD compoents that are only supported by a couple of small pads. The weight of the component can pull the pads off due to shock or vibration.

Well heat sensitivity, would just mean I'd have to solder by hand, and the small pad problem is solved by adding some vias to connect to another pad on the bottom. That prevents the pad from peeling up for a large part (technically it might make it harder to solder, but if the pad on the bottom is small, it wont matter).

And the above should raise the question that the OP needs to have answered up front ... what's the trade study for using SMD supercap vs good ole Li battery backup ? I can understand the desire to avoid the space and cost of the normal Li charging circuit but does that still hold up ?

Personally, I have space restrictions that prevent me from using a Li battery. I might be able to justify it if I use a backup battery for the entire system, but Li battery for just an RTC is not justifiable for my application. Really the choice is between a coin cell and a supercapacitor. While it is easier to just run with a coin cell battery, I'd prefer to design the device in such a way as to minimize the number of times it needs to be serviced. It needs to operate sometimes remotely on a regular basis. So thats why Im leaning towards the supercapacitor.

However, what if the temperature goes into both extremes, as with very hot summers, and very cold icy winters? Would a coin cell do better or would a supercapacitor? How much would the effect of temperature extremes have an effect on lifespan of the energy storage?

Another question, and this might sound a little redundant, but is there a Li battery or Supercapacitor in the shape of a coil cell that I can optionally remove from the circuit if it fails, rather than SMD the actual energy storage device? Granted this would increase the number of times that the device would need to be serviced, but, still I'd like to know.

[viskr]

Did you look at the ML621, which is available at Digikey. 0.27" coin size available in coin or SMT?

There are smaller rechargeable batteries out there, just Digikey doesn't carry them.

[MichaelN]

...the small pad problem is solved by adding some vias to connect to another pad on the bottom. That prevents the pad from peeling up for a large part...

Yep, this is quite effective and I do this myself when I need to use SMD connectors etc that might have some force applied to them. Surprising that manufacturers don't suggest this in their datasheets.

[UNTEngineer]

Did you look at the ML621, which is available at Digikey. 0.27" coin size available in coin or SMT?

There are smaller rechargeable batteries out there, just Digikey doesn't carry them.

But then, I would need an additional recharging circuit to have a rechargeable Li battery embedded in my circuit, would I not? That just means I have to pack more electronics in a small space, and Id rather not. Can you suggest a reasonably small circuit that could actually compete with the supercapacitor, in terms of cost and size, to convince me to invest in additional electronics just for an extra 4mAh?

[viskr]

All you need is a diode and a resistor from a 3.3V supply, both can be 0603 or even 0402 parts. It doesn't get much smaller than that. Mouser carries the 421 and 414 size parts which are a little smaller.

From the Digikey page they have a link to Panasonic's recommended charging circuit and the appropriate values.

[UNTEngineer]

All you need is a diode and a resistor from a 3.3V supply, both can be 0603 or even 0402 parts. It doesn't get much smaller than that. Mouser carries the 421 and 414 size parts which are a little smaller.

I found a couple of circuit diagrams, but Im not sure which to use. The top or the bottom, or does it really matter?



Im still leaning towards the supercapacitor though, because that can run straight off my 5v rail, with no extra components. With the battery, I'd have to route 3.3v over to this circuit, Id have to include a protection diode into my 3.3v rail and then I'd have to add the additional components to charge the circuit. :think:

[stevech]

what's the supercaps discharge due to leakage? Probably varies greatly among vendors/parts.

[viskr]

The schematic diagrams you copied above are for direct charging from 5V. What both do are to create a Thevinin equivalent circuit for a 3.3V source and series resistor. In that same spec sheet they show the following as the charging circuit when using a 3.3V source.



This is the circuit I use, and you do not need to use large traces for routing, the current into or out of the battery will be in the micro-amps, much less than most signals

[UNTEngineer]

The schematic diagrams you copied above are for direct charging from 5V. What both do are to create a Thevinin equivalent circuit for a 3.3V source and series resistor. In that same spec sheet they show the following as the charging circuit when using a 3.3V source.

Say I run with the Lithium Battery recharge idea for the moment, do you think I could use just a 850mAh battery from Sparkfun, or is that just way too much capacity? I think I'll have to get a few things to run an experiment or two to verify this.

[viskr]

NO, if you use an 850 mAh battery that one WILL require a more complex charging circuit, not just a diode/resistor. Using a diode and resistor to charge a large Li battery can result in fire.

The small coin Li rechargeable batteries have been designed for low current charge/discharge cycles which is what you need to keep an RTC alive. Those type batteries are available in 1-48 mAh ranges.

[UNTEngineer]

Oh ok. I guess Im stuck pondering this until after Thanksgiving. Thanks for the help.