Please help me to stop blowing up boost converters

[woody363]

So far I've blown up 4(!) boost converters with a fairly simple circuit. Obviously I'm doing something wrong, please educate me...

The converters I've tried are https://www.ebay.co.uk/itm/191827311442 and **broken link removed**

My power source is a 5C lipo battery approx 19V 16Ah

My load is a 200W LED around 42V, 5A

Eventually I want to replace the on-board current control pot with an external one for brightness control, however I'm not getting far enough to have issues with that yet


Method:
Get new converter, connect inputs to benchtop power supply and adjust voltage control until there's 42V on the output
Turn off and connect up LED, test with benchtop power supply and adjust output current to around 1A (the most my little power supply can handle)
Swap benchtop power supply for battery and re-adjust output current to 5A
Turn on and off a couple of times to test
Smell magic smoke
Order new boost converter from ebay
Return to step 1



Observations:
The converters always die as they are turning on, never during operation.
They always work fine on benchtop power supply and only die when connected to battery

I wandered if the order in which the battery/load are connected to the converter matter and I've tried both (by placing the switch on either side):

The switch was on the input for first three converters, but I swapped it to the output for the latest converter when we thought "maybe when the converter is turning on initially the current regulation hasn't fully kicked in and it sends too much current through to the LED, destroying itself". However the 4th converter died as it was being connected to the battery, with no load attached on the output.

The cheaper 250W converters lasted a few tests before blowing up, The 400W converters both died the first time they were connected to the battery (one with load, one without)

Yes I've got the +ve / -ve wires the right way round


Am I using the wrong type of converter? In what order should the battery/load be connected? Do I need something else in the circuit to protect the converter?

Thanks for any insights you may have!
Woody

 

[Pommie]

Try running at 3A and see if (and what) gets warm/hot? Measure the current and voltage of the output if possible.

Mike.

 

[ChrisP58]

Please note that a boost converter, or any power converter for that matter, does not increase power. The input power will always be equal or greater than the output power. Equal if you have 100% conversion efficiency. Greater if there is some inefficiency to the conversion process. One of the converters you link to states a maximum efficiency of 96%. Typical efficiency is probably less than that, but we'll use that number here.

Your output is 42 Volts at 5 Amps. Multiply those and we get an output power of 210 watts. The input power needed to make that output is that same 210 watts scaled by the efficiency, so that gives us about 220 watts.

If we divide the input power by the input voltage, that will tell us how much current is needed at the input. 220 Watts / 19 volts is about 11.5 amps. Considering that the efficiency is probably less than 96%, the input current is probably more than that. Also be aware that, as the battery voltage drops as it discharges, the input current will increase.

That's how much the converter is pulling from your battery. Reading the converters specs says that you need to increase the heat dissipation if it's above 8 amps. You're about 50% higher than that.

I expect that the unit your using just doesn't have the capacity to handle that much input current on it's own. You may be able to add a fan to cool it off.

Unit referenced in this discussion, **broken link removed**

 

[woody363]

Hi Mike. thanks for the reply.

I'm a bit apprehensive about buying another converter to test without understanding more about what I've been doing wrong.

Do you mean 3A input or output? The last two converters both died while adjusted to 1A output and blew before any measurable output was produced.

 

[woody363]

Thanks Chris,

It sounds like I need to use a different converter? Is there a different unit you would recommend, or something I should look for in the specs?

I had read the current limitation and have a fan connected up, but as the unit blows up instantly on power up I assume heat dissipation wasn't the problem? The unit didn't get chance to get warm.

 

[rjenkinsgb]

I'd guess it's poor design and they cannot stand the initial start-up surge current available from the lithium cells (possibly hundreds of amps).
The bench supply simply cannot provide enough to do any harm & presumably they test on something with limited capability.


The simplest fix I can think of is a "soft start" using a couple of relays and a resistor.

Resistor in series with the input, eg. 10 ohms.

The first relay coil connects across the output, eg. a 12V with a 30V zener with a zener so it only switches on when the PSU has reached around full voltage.

That relay shorts the input resistor. A second contact turns on the second relay coil via a series resistor plus a cap in parallel to the coil, so there is a slight delay for the PSU to settle with the input at full power.

The second relay contacts turn on the LED load.


Or, possibly, just an NTC thermistor in series with the input supply?
You can get those specifically designed for inrush current limiting, it just depends if there are any with appropriate ratings and low enough working resistance for the current involved.

Edit - they do exist, this one is 2 Ohms when cold and rated for up to 12A. It's resistance drops to something like 40 milliohms with 12A current, but it's also extremely hot at that.

B57127P0209M301 | EPCOS Thermistor, 2Ω Resistance, NTC Type, 31 x 7mm | RS

uk.rs-online.com

 

[woody363]

>Or, possibly, just an NTC thermistor in series with the input supply?

Thanks! that might just do it

I've found some which I hope may do the job. I'll report back when I've tested them. Hoping to avoid anything too complicated if possible.

 

[be80be]

Input current: 10A (MAX) exceeds 8A please enhance heat dissipation

I bet your way over that trying to boost 19 volts to get 42V, 5A

 

[dr pepper]

Sounds like the converter current limit operates on just the output, the actual converter circuit probably draws more than it can handle, as mentioned a design issue if so.
The thermistor is a good idea if you can get one with a low enough value, ones I have seen are meant for mains and tend to be fairly high homage.

 

[Nigel Goodwin]

Sounds like the converter current limit operates on just the output, the actual converter circuit probably draws more than it can handle, as mentioned a design issue if so.
The thermistor is a good idea if you can get one with a low enough value, ones I have seen are meant for mains and tend to be fairly high homage.

The RS one mentioned above looks pretty suitable?.