diy solar

diy solar

DC/DC converter module efficiency testing

Sverige

A Brit in Sweden
Joined
Oct 8, 2020
Messages
742
Location
59.5N, 15.5E
I ordered a few different types of DC/DC converter modules from amazon all of which had vague specs like “up to 95% efficient“, so I started to wonder how efficient they really were in the application I wanted to use them in.

All work in a similar way - you connect a DC voltage at the input then adjust a potentiometer on the board to set the output voltage to your desired level. Step up and step down modules are available - all of these were step down modules where the output voltage must always be at least half a volt lower than the input. Output voltage then remains fixed even when the input voltage varies, which is a really great feature.

These are great for taking DC straight from your house bank battery DC load centre to all those devices which normally use a ”wall wart” AC adapter (which can then be retired from use, reducing your overall energy consumption).

These modules are so cheap! The mini 360s were €7.99 for a ten pack, the LM2596S about €1.36 each and the more powerful XL4016 units €6 each. I’m attaching some photos of the units with summary specs and below are the test results I obtained when comparing their performance converting down from 14V to 9V, 7.5V and 5V.

Clearly the overall conversion efficiency is highest when the volt drop from input to output is minimised and also as output current grows, so does the efficiency. This makes sense if you consider that each module has a self-consumption current which gets taken from “the bottom line”.

I didn’t expect to see an efficiency as high as 94% in a real world test, but the XL4006 achieved this with output of 9V and 1.7A. This was not an exhaustive test and I didn’t adjust test parameters in search of the highest efficiency numbers, I simply tested them with the output voltages and currents I expect to use in my home.

Despite the XL4006 being the winner, I think the most impressive device was the mini 360, which is the size of a postage stamp, very cheap and performed almost as well as the larger module. The only drawback is the tiny pot on the mini 360 is quite hard to set accurately. These units must have a ton of applications in distributing DC around an off grid home - hope the test gives some useful info, even though it’s a partial one.

54367DC1-B3E2-4EF6-91FC-3AF6B3EC3CFD.jpegE878CB32-4AE2-431E-865E-7D00DC7B9862.jpeg
CCED4074-AFA4-47FD-B3DC-A6A3385534A7.jpeg

20713583-2E48-4955-9A05-BFF6A03C1449.jpeg
 
I have rejected the mini360 modules. They have a far too high quiescent consumption, some of them get hot without any load.
The MP23070 chip, on them they base, is not produced any more.
For a reason; probably because their synchronous operation is buggy and the two FET's might overlap?

The alternative heretofore is the D-Sun-Mini.
Here are my findings about buck converters:
Choice of buck converters.pdf

The quiescent current consumption is important, since it just will drain your battery from nothing, especially if you have many of these modules running in parallel.
 
Last edited:
Thanks @rin67630 - looks like you’ve done more extensive testing than my short test. I hadn’t heard of the D-sun mini. Where did you buy it and do they go by any other name, as I don’t see it on amazon?

I tried to measure quiescent current as part of my testing, but the meter I was using was not reliable at very small currents. I guess the variation you see in mini 360 performance may be due to different hardware producers making modules of various quality levels. Even if it looks like a single product to us, there will surely be a mass of different producers making better or worse copies of the original design. Pure luck whether you get a good or bad one I suppose, so even more reason for testing these things as you have done.

I hope your PDF file remains accessible as your test results add to what I’ve posted. Sometimes with forums, attachments get lost in software upgrades, storage maintenance, etc. If it’s easy for you to do so, consider adding some highlights or summary of your test results directly in this thread for the benefit of users who find it thru google or whatever in the future.

I searched for, but didn’t find good info on comparative efficiency tests of these modules, which is why I started doing my own testing.
 
D-SUN converters at eBay

Maybe I can find better ones.
They perform well and efficient up to 10W (19V to 13V), but then even with a heat-sink at the back they begin to throttle and lose efficiency after a couple of minutes.
Thy might be better at 13V->5V...

you did not mention these modules any more:?
1609345337963.png
I thought you have some of them to hand, haven't you?
I have not finished the tests, but they seem to warm less than the D-SUN and be more efficient.
Beside that, they have an enable pin and the control adjustment is easier to tap.
 
D-SUN converters at eBay

Maybe I can find better ones.
They perform well and efficient up to 10W (19V to 13V), but then even with a heat-sink at the back they begin to throttle and lose efficiency after a couple of minutes.
Thy might be better at 13V->5V...

you did not mention these modules any more:?
View attachment 31640
I thought you have some of them to hand, haven't you?
I have not finished the tests, but they seem to warm less than the D-SUN and be more efficient.
Beside that, they have an enable pin and the control adjustment is easier to tap.
No, I didn’t buy any of those particular ones. I was initially interested when I thought they were outputting multiple voltages simultaneously, but I realised I was mistaken so I didn’t bother with them. I’ll try to get some D-Suns if I can find them within Europe. Thanks for your link, but ordering from Asia attracts extra tax and collection fees here, which would effectively make the price on a low cost item like that 4x or 5x the ebay price.
 
No, I didn’t buy any of those particular ones. I was initially interested when I thought they were outputting multiple voltages simultaneously, but I realised I was mistaken so I didn’t bother with them. I’ll try to get some D-Suns if I can find them within Europe. Thanks for your link, but ordering from Asia attracts extra tax and collection fees here, which would effectively make the price on a low cost item like that 4x or 5x the ebay price.
So just search for "mini buck converters" on your favourite dealer shop you should be successful in the results.
 
I just found out that the green mini module with the various voltage presets stages a MP2315 chip, which is really the last synchronous switching technology.
It delivers 15W from 19v to 13,8V without becoming excessively warm.
 
I have now tested more intensively the D-SUN converting 18V to 12V.
With a heat sink and 22°C around, it manages 18W permanently with 93% efficiency.
1609362147552.png

At 22W, the efficiency decreases to 86% and the temperature increases rapidly, after 10 minutes the D-SUN will then stop working for 1-2 seconds and resumes, stopping again a minute later...
That remains safe, the chip has a temperature protection, is however clearly over it's power capacity.
I would rate the D-SUN solution to 15W to have a bit reserve...

Else, one would need a bigger heat sink.
 
I have now tested the green mini module with the various voltage presets staging a MP2315 chip.
With the same heat sink and 22°C around, it feels perfectly cosy managing permanently 22W with 95,96% efficiency.
1609365819664.png
That module is definitely superior!
It seems to run >24W without any difficulty:
Tomorrow I can test more, I need another bulb, that one is an 20W bulb, i cannot put more...
1609366342315.png
 
Do these units limit current themselves or do you have to size the load appropriately? Can they be used to charge one battery from another? My house battery is a large but aging lead acid (1600ah 24v) and could do with a boost some nights through winter. At the same time i have 110ah 12v lifepo4 battery sitting there doing nothing most of the year. Would be handy to transfer power across.
 
I have now tested the green mini module with the various voltage presets staging a MP2315 chip.
With the same heat sink and 22°C around, it feels perfectly cosy managing permanently 22W with 95,96% efficiency.
View attachment 31675
That module is definitely superior!
It seems to run >24W without any difficulty:
Tomorrow I can test more, I need another bulb, that one is an 20W bulb, i cannot put more...
After half a year experience, using them to solar charge, i can't recommend them any more: they are too touchy in the filed and burns through to easily. They are only good inside a device.
There is nothing like converters with discrete FETs on a good heat sink...
 
I often add ntc thermistor on cheap converter to add current regulation and keep working at good efficiency but the drawback is weak precision on voltage output setting; NTC 47kohms is parallel with FB divider high side resistor, when device is heating, the output voltage is going down and stabilize at a desired max temperature.
It's working really fine for charging battery
 
Back
Top