EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: soldar on November 29, 2024, 03:34:32 pm
-
I have a Hoover car hand-held vacuum cleaner which has 7.5 volt rechargeable batteries which are long dead.
So I bought this XL4016 buck step-down module with the intention of plugging the vacuum into the car 12 V.
https://www.ebay.com/itm/122922143476 (https://www.ebay.com/itm/122922143476)
The idea is that the vacuum uses (all very rough, round numbers) about 50W so, say 7 A @ 7 Volts or 4 A at 12 V or, taking conversion losses into account somewhat under 5 A @ 12 V
So I receive the module and do some tests and I am quite surprised the efficiency is so bad that it takes the same current in as it outputs.
IN: 6.4 A @ 12 V = 76 W
OUT: 6.4 A @ 8 V = 51 W
Efficiency: 67% (dismal)
Lost as heat: 25 W (Way too much, radiators get too hot)
For this kind of result I do not need any kind of fancy switching module. A resistor will do the same, much simpler and cheaper.
In fact, I can size the cable just right so that I drop the voltage in the cable itself.
Are these modules really that inefficient? Am I missing something?
I was hoping to plug this into the car lighter socket and keep the current under 5 A.
-
Possible starting surge from vacuum motor damaged it. Motor can draw a lot of current at start. Can also get back EMF surge when switched off too. Converter may not have been designed for inductive loads.
Mike
-
Possible starting surge from vacuum motor damaged it. Motor can draw a lot of current at start. Can also get back EMF surge when switched off too. Converter may not have been designed for inductive loads.
Mike
That does not make sense to me. It it supposed to be protected against over current. And if it had failed it would not work at all but it is working, it is switching, just with very low efficiency.
-
I've looked at the ebay page and cannot
find anything about over-current protection.
-
I've looked at the ebay page and cannot
find anything about over-current protection.
I might have seen it in some other implementation of the same IC.
In any case, the module is working, it has not failed, it switches, it reduces voltage.
-
https://www.xlsemi.com/datasheet/XL4016-EN.pdf (https://www.xlsemi.com/datasheet/XL4016-EN.pdf) (datasheet)
The module is behaving like a linear regulator (input current = output current). :-?
What are the two devices on the heatsinks? Is one a diode?
-
I had a similar problem a couple of years ago, and what I found is that the gate drive was so poor, that the Mosfet’s rise and fall times spent about 15% of the TOTAL switching period in linear mode.
Another possibility: the Mosfet is not fully saturated.
These are only my speculations. Time to take out the scope and probe the Vgs and Vds waveforms.
-
I once took a really nice battery powered production-quality torque limited screwdriver and made it corded from a AC-DC converter. The inrush due to the brush motor starting hit current limit on the power supply every time. I had to size up the power supply many times the run current just to get the screwdriver to spin.
The punch line here is that batteries are actually really good low impedances sources and making comparable DC DC converters is non-trivial.
-
I had a similar problem a couple of years ago, and what I found is that the gate drive was so poor, that the Mosfet’s rise and fall times spent about 15% of the TOTAL switching period in linear mode.
Another possibility: the Mosfet is not fully saturated.
These are only my speculations. Time to take out the scope and probe the Vgs and Vds waveforms.
I have done some tests and the efficiency is not too bad when demanding low power but falls when power demand increases.
With about 13 V input 8 V output and resistive load the efficiency is (roughly)
13W - 100%
26W - 97%
40W - 88%
50W - 86%
So not too bad with a purely resistive load. But with the Hoover vacuum, even though I have put some big capacitors in parallel, the efficiency drops dramatically.
I guess I could just put a resistor in series with the vacuum cleaner but I was trying to keep the current down because I do not want to blow a fuse in the car. I need to check what the fuse is rated for I guess if it is 10A and I just put a resistor in series then I am safe.
And the step down module I can just keep it for some other application.
-
Something I've found with these cheap modules is the capacitors are grossly undersized and the control loop goes unstable.
470uF is way too small for 5-8A and the high ripple current will cause them to heat up as well.
-
Something I've found with these cheap modules is the capacitors are grossly undersized and the control loop goes unstable.
470uF is way too small for 5-8A and the high ripple current will cause them to heat up as well.
470uF is too small for the output of a 12V to 7V DC DC Converter? Are you expecting a switching frequency of 10kHz or something?
-
I'm saying two things I guess.
You can't get 5A ripple current out of a 470-1,000uF cap for very long. Try it sometime.
OP's board those caps look low quality ChongX or Chenxing babies, at least the low ESR green ones I would expect. They likely expect you to add additional output capacitance which I would try. XL4016 is 180kHz, datasheet uses 1,000uF but to 12A. Also the motor ripple current is LF though.
I have seen unstablity (in other boards XL2596) make extra heat until the caps were reasonable.
It might even be a smaller XL4015 or dud XL4016 on the board. Or cheap high RDSon mosfets. I find these boards a PITA.
-
I agree that the capacitors are too small and I always, as a matter of course, add capacitors at input and output.
Still the module works correctly with a purely resistive load but not with the vacuum motor.
I might experiment to see if there is anything I can do, like some filtering, that would make it work correctly.
If not, then I will just put a resistor in series. Probably size the cable for the correct resistance.
-
I've looked at the ebay page and cannot
find anything about over-current protection.
Even if it was stated, that would have been the seller relying on the designer relying on the chip vendor. Clicking thru the eBay seller's feedback to the About, says the actual seller is located in China. Now, if the designer used some very inexpensive and very likely cloned AMS1117 regulators, they probably have dodgy overload thermal protection. So here we are, and Bob's your uncle.
-
The vacuum motor should have a disc cap(s) across the terminals to suppress arcing at the brushes.
That might be making the module go unstable or false trigger its current-limiting (LF 48kHz) mode.
Also vac's generate high level ESD so the motor housing/frame I would also have some path to circuit ground so it isn't arcing somewhere unexpected.
Another problem sometimes with these modules is a crap PCB layout. The ground pour has slices, feedback trace goes all over town, close to inductors etc.
This can make them fall apart under higher load and maybe the motor currents here flush that out.
-
I have done some further experimenting and I think it is not worth the trouble and it is better to just drop some voltage with a resistor.
The module seems to work OK with purely resistive loads so I will keep it for future use in some other project.