I need to power a 12VDC device from an AGM battery, of which the voltage will vary from about 10.5 to 14.5V. The device's datasheet does not mention the tolerance on the supply voltage, I'll assume it is +/-10%.

I need the conversion to be extremely efficient, with an efficiency >= 0.9. The good thing is that the device should draw a constant power of 5W - perhaps this would allow some kind of tuning.

My current baseline is a non-isolated Buck-Boost converter from Tracopower, a TMDC 2412 or TSR. But its efficiency is only stated at 9V input voltage (0.91 and 0.94), and I suspect it is going to be below my requirement around 12V in.

Is there a particular technology, or a circuit add-on, which would yield the best efficiency for this particular application?

  • 1
    \$\begingroup\$ What device it is? Can you modify it? If it already consumes constant power at 10% tolerance in voltage, it should already have a regulator in it. \$\endgroup\$
    – Justme
    Commented May 4, 2021 at 9:32
  • 1
    \$\begingroup\$ +1 for @Justme: 12V devices are generally designed for lead-acid battery power so they happily accept anything between 10V and 15V. \$\endgroup\$
    – fraxinus
    Commented May 4, 2021 at 9:58
  • 1
    \$\begingroup\$ Best not to discharge an AGM below 12V anyway, because that indicates a 25% state-of-charge. Lead-acid (AGM) batteries are NOT perfect batteries, and are permanently degraded by deep discharge. Those who want thousands of cycles aim to keep their lead-acids between 70% and 100% SOC. (yes that's really that bad). The only reason people use these lousy batteries is a) they're cheap compared to the competition, and b) they happen to be good at starting engines which creates stupendous economies of scale, see "cheap". \$\endgroup\$ Commented May 5, 2021 at 0:31

1 Answer 1


The battery voltage will vary as a function of state of charge. So, if your load cannot accommodate for that range you need an interface dc/dc converter, and being the load 12Vdc the converter must have buck and boost function. Consider that while charging the battery the voltage should approach that 14.5V you mention.

The Tracopower isolated converters are not so efficient, if I remember well. Your TSR1 is non-isolated and advertised with 0.94 at Vin,min (9V) that is a pretty good figure, instead.

Isolated converters have lower efficiency: as a comparison Murata NCM6 gives 84-86%, and this above 40% load level (measured with respect to 6W). Recom REC6 are very close to that. It is more or less a good standard efficiency level to use as reference.

The non-isolated versions in principle are more efficient, but much depends on the amount of information the manufacturer provides. For example the Cincon EC5A does not go better than 87%, and I think in the best operating conditions, but no clue which ones. Your Traco does not show curves as a function of Vi and load % either.

Generally speaking in case of a custom design efficiency for a dc/dc converter of this kind may be improved by: reducing voltage drops across diodes (Schottky, correctly selected for low VD at the operating current), even using synchronous rectification at the secondary (the used MOS would reduce by far the voltage drop), low DC resistance of inductors and transformers. For a quick hands-on solution, instead, use commercial products.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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