I'm designing a custom audio system for my electric motorized pedicab. The motor of my cab runs on a 48 V 33 Ah battery, and I have planned to use a 48 V to 12 V step-down converter to simultaneously also run my lights and audio system. I am new at all of this and have been learning a lot, but am nervous I am going to either:

  1. overload something (like my battery) or,
  2. lose a lot through the inefficiency of my system.

My batteries are rated for a continuous draw of 50 A. I have been told to set aside 30 A for my motor, so that leaves me with 20 A for lights and audio.

The cab comes with a 20 A rated 48 V to 12 V step-down converter installed. I was worried about peak draw from my amplifier/speaker system and purchased a 60 A converter to use instead. Someone on an audio forum pointed out that, while yes this would protect me from any peaks, it would be much less efficient as these step-down converters are more efficient as you get closer to matching the actual current that runs through them with their capacity.

Is that true? Will my 60 A rated 48 V to 12 V converter be less efficient than sticking with the installed 20 A one? My amplifier puts out 45 watts RMS per channel and I was hoping to bridge my speakers, so that's 90 watts RMS per channel for a total of 180 watts. 180 watts at 12 V is 15 amps. Folks on the audio site also said I'll never ever get close to this and will most likely only ever draw 2-3 amps at a time. That makes sense because the amplifier has a 15 A fuse on it, so I'd like to think it would never get close.

Since I'm running a mobile system, it's important to me for it to be efficient. I misread the advice I was given and bought this 60 A step-down converter, but have now realized that perhaps it will be more inefficient than the installed one. I'm fine sticking with the 20 A installed one IF it's not going to hit its upper limits. I'll probably be using just over 5 A for lights as well, so really I only have 15 A to work with for audio.

As I stated, perhaps I'm overly worried and my audio system will hardly ever draw anything at all. I'm using a class D amplifier and two 6x9 3 way speakers, rated at 150 watts RMS with a sensitivity of 93dB.

Thanks for any and all help! I've got to figure this out in the next week.

  • 1
    \$\begingroup\$ 20 A for lights and audio gives 20 x 48 = 1000 W. Cars typically use 2 x 55 W headlights 4 x 6 W side/tail lights and occasional brake and turn signal lights so 200 W is plenty for those before you even consider LED which should reduce that to about 50 W or so. Try your audio on the weaker PSU and let your ears tell you if there's a problem. \$\endgroup\$
    – Transistor
    Commented Mar 31, 2019 at 21:05
  • 1
    \$\begingroup\$ I imagine there is no way you'll be using \$150\:\text{W}\$. For typical music, you want an amplifier rated for about 5 or 6 times what you actually will be delivering to the speaker system. So you probably have the right size. Just be aware that it won't be consuming it's full rating. Perhaps a half, or a third of it, in actual use. If that much. You can't precisely predict all this, so don't over worry about efficiency. You'll probably only need a couple amps at 48 V, worst case, with a DC to DC switcher. \$\endgroup\$
    – jonk
    Commented Mar 31, 2019 at 21:10
  • \$\begingroup\$ Thanks for your insights. A couple amps at 48 V translates to 8a at 12V which is getting a bit high. The LED light strip is 16ft long, has 300 leds and uses 5a. Another amp for the fairy light strip and whatever my front and tail lights are (which you are saying is about 50W or so, so at 12V that's 4 more amps and I'm at 18a (8+5+1+4). Getting close to the cutoff point. So the other question is, what if my system is asking for more than the 20a the step down converter can handle? Will the lights dim a bit, or the audio fade, or would something give out/burn out on the converter? \$\endgroup\$ Commented Apr 1, 2019 at 0:29

3 Answers 3


It's impossible to make any general statements about the relative efficiency of the two converters at different power levels. It depends very much on their actual design.

You would have to actually test both converters with dummy loads, measuring both input and output power.

  • \$\begingroup\$ It would be nice to test everything and find out exactly what is best to use. Unfortunately, I'll be getting my newly manufactured pedicab delivered only days before I need to leave to work a very big festival. So I need to decide whether or not to bring the manufacturer the 60a converter or be okay with them installing the 20a that usually comes with the bikes. \$\endgroup\$ Commented Apr 1, 2019 at 0:32

Although this would be very design dependent, and there are many optimizations that can be made on how to react with variations in load, if you look at the efficiency curve of most converters you will see a clear pattern emerge:

Converters reach a peak efficiency at a high percentage of the rated load (~ 70% - 90%), this a very likely design decision as it is the most likely operational point that system designers would choose.

Likewise, if not addressed, their efficiency would drop quite dramatically at low loads (it takes more power to handle large components for large loads, all of this is wasted if your load is close to zero). That's partially why many converters incorporate some sort of "hiccup mode" in which the converter mostly shuts down for low loads. This design decision could cause problems (like audible noise) in some systems.

  • \$\begingroup\$ Thank you for helping me understand how these work better. The following specs are found on the product description of the 60a converter: * Input Rated Voltage: DC 36V 48V 30-50V * Output voltage: DC 12V 60A 720W * Voltage Regulation: Less Than 1% * Load Regulation: Less Than 2% * Efficiency: 97% (half load); 95% (full load) I'm curious why it says the efficiency is 97% at half load, but only 95% at full load. Of course, I've noticed that often Chinese parts throw up all sorts of numbers and they are not always to be trust. So I don't know. \$\endgroup\$ Commented Apr 1, 2019 at 0:34

To take a different tangent from the answers you have, a few things:

Note that better quality voltage converters will typically provide an efficiency vs load graph, and that is a good reason to pay a bit more for high quality, properly documented product. Also, if you already have both voltage converters in your possession, don't forget that you can simply test both of their input power at a matching load if you like.

Another thing to note is that you're specifically stepping voltage down, which means your converter could well be a buck converter, and that buck converters tend to be PWM. If your converter has no graphs, you may be able to look up the controller chip it uses and figure out what type of converter it is.

Switching voltage converters suffer from two types of loss, switching loss and conduction loss. Which loss dominates will depend on the type of converter. If conduction losses dominate, the converter tends to be efficient at lower power output, and where switching losses dominate, the converter tends to be efficient at higher power levels.

There are two main types of switching control, Pulse Width Modulation(PWM) and Pulse Frequency Modulation(PFM)

In PWM mode, pulses of varying length are provided at a constant frequency. The loss per switching is constant, so when the ON time is very low, the switching loss per pulse is increased, and when pulses are long, the switching loss becomes tiny by comparison. This results in a converter that achieves best efficiency at close to its rated load.

In PFM mode, pulses of constant length are provided at a varying frequency. The switch is chosen to have insignificant losses at the given pulse length, so conduction losses dominate, resulting in a fairly flat efficiency graph(compared to PWM) with lower efficiency as output power increases.

Non-isolated Buck converters IIRC are usually PWM, and I believe PFM is used in applications where the field in an inductor core is not to be exceeded, so whereas PWM would only work up to the point where it saturated the core, PFM would use a pulse length that never results in core saturation, allowing an optimally small core.

Finding a graph for a PFM converter at low voltages was surprisingly difficult(last time I looked it was the first one I clicked), so it is possible that PWM and therefore higher efficiencies at close to full load is much more common, at least in low wattage converters.

Efficiency vs load graph for CUI P78A-1000 converter, 2 different topologies are shown.  Right graph scaled to match left.

These are graphs from a through hole voltage regulator. You can see that the 3.3V model has a much different efficiency curve. We can see it has a worst case efficiency of 84% and a best case of about 92%. The converter on the right varies much more, but does achieve higher efficiency at close input voltages and high output current.

  • \$\begingroup\$ Thanks for helping me understand this all more. So it seems to me I need to determine whether this is a PWM or a PFM model. I received no manual, and as far as I can tell, there is none of that information on it's page on Amazon: amazon.com/gp/product/B07D27KNFD/… \$\endgroup\$ Commented Apr 1, 2019 at 1:04
  • \$\begingroup\$ If you're not an engineer it may take you a tremendous amount of time to determine that without spec sheets. Looking at your link, they are claiming 97% and 95% efficiency at half and full load. Very respectable if true. Online product specs from off brand companies can be cherry picked or exaggerated though, so take everything with a unit of sodium chloride. \$\endgroup\$
    – K H
    Commented Apr 1, 2019 at 1:27
  • \$\begingroup\$ Good point. So the question remains... do I go with the 60a converter... playing it safe on load but potentially having trouble with low load and efficiency, or do I go with the 20a converter, assuming as many have told me, my audio setup won't need many amps? \$\endgroup\$ Commented Apr 1, 2019 at 1:55
  • \$\begingroup\$ I would attempt to test the one you have before ordering another. Measure input properties under load. If efficiency is unacceptable, then worry about another one. If you're buying cheap undocumented converters, you'll just have to test. They don't all suck or anything, they're just an unknown. If you're buying them from a higher quality source, part of what you're paying for is predictability, reliability and documentation. \$\endgroup\$
    – K H
    Commented Apr 1, 2019 at 2:02
  • \$\begingroup\$ Appreciate everyone's help on this. Went with the 60 amp converter. Not noticing any problems with drain on my battery. May eventually figure out how to test it's efficiency. Thanks again! \$\endgroup\$ Commented Apr 17, 2019 at 3:02

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.