Possibly fried LTC3780

[vgamesx2]

Hello I recently received my LTC3780 boost-buck converter and I was using it to power a high powered delta fan when it suddenly stopped working and is simply giving me the light on the left saying input voltage is too low, as you can see here

then after leaving it connected for a while it started smoking slightly which you can see in this second pic that it seems to have melted the chip under the CV control pot a bit, 
does anyone have any suggestions on what could've caused this?
My power source was a 19V 3A laptop supply so it was within spec, the only odd thing I've noticed is when I plug it in, as expected that microcontroller gets warm however so does the underside of where the OFF/ON pins are, that aside there's nothing obvious wrong with it. Any ideas?

[singapol]

Circuit could have worked if it was driving resistive load but you are driving an inductive load (fan).

http://www.datasheetdir.com/LTC3780+Step-up-down

Don't know if there are protection provided for inductive loads to protect mosfets.
You can probably repair it by replacing LT3780 and making sure mosfets and capacitors are the required type as per datasheet. Ie. Logic input fets and low esr caps like Sanyo Oscon or equivalents.




Edit: It could be only mosfets are damaged then replace accordingly.LT3780 has builtin protection it could have survived.

Edit: The basic design has already diodes across mosfet  on board so need to check also if they are shorted.

http://www.aliexpress.com/store/product/LTC3780-DC-DC-CC-CV-Auto-Boost-Buck-12V-Converter-DC-Step-up-Down-Power-Supply/224263_32227074387.html

[TheElectricChicken]

......or a capacitive load, the power supply caps in the laptop.


For example., 300w 220v inverters can be completely destroyed by BBQ if they are connected to old fashioned 40w fluro light.

[vgamesx2]

Oh.. well I'm a bit of a newb so I didn't entirely realize how different loads were important, so to solve this issue I would only have to add a reverse bias diode in parallel with the fan correct? or.. would I also need a diode on the output of the LTC3780?
I suppose even if I don't need to I should probably do it anyway as a "better safe than sorry" protection though right?

Thanks for the help, I'll see about replacing the fets and see if it'll still work.

Edit: A 1n5819 should work well enough right? Actually I just realized a 1n5819 wouldn't be enough, it only supports up to 1A of continuous forward current and.. my fan uses 2.75A (max 3.3A) so.. I'll have to look a bit harder.

[c4757p]

If it's a reverse protection diode, it should only have to take current in pulses. Even though the peak current will be approximately equal to the fan's run current, the average as seen by that diode will not be.

ST's 1N5819 datasheet says:

I_F(AV): Average forward current: 1A
I_F(RMS): Forward rms current: 10A
I_FSM: Surge non-repetitive forward current: 25A

I imagine it will be perfectly fine.

[vgamesx2]

If it's a reverse protection diode, it should only have to take current in pulses. Even though the peak current will be approximately equal to the fan's run current, the average as seen by that diode will not be.

ST's 1N5819 datasheet says:

I_F(AV): Average forward current: 1A
I_F(RMS): Forward rms current: 10A
I_FSM: Surge non-repetitive forward current: 25A

I imagine it will be perfectly fine.

If you say so, I suppose I'll try it then.

[peter.mitchell]

is it just me or does the opamp near the shunt look like it's let out the smoke?
the top side output fet is probably cooked.

also, to improve efficiency put a small schottky (SS16, 1n5819 etc) in parallel with both the input bottom side fet and the output top side fet, this means in buck operation the body diode on the mosfet won't conduct when the driver hasn't fully turned on the mosfet, the body diode has a slow recovery time and will waste power, also, it means i boost operation your boost fet won't be subject to potentially high voltages if it turns on slowly. these modules are pretty horribly designed

[vgamesx2]

Surprise update, I've finally bothered to replace the mosfets on this thing and the good news is that it works, the bad news is it doesn't work properly...
It seems CV isn't working, it will drop the input voltage by a small amount but that's it, it won't boost it higher nor will it lower it further, so for example 8.4V from two batteries only gives me about 8V out, I haven't tested CC yet but I'd assume that doesn't work either...

So let me guess it's the fried opamp right? in that case does anyone know what opamp it was? my only guess was an MPC6002 but I have absolutely no clue because the middle numbers are melted and the only other chip I could find these things using was an LM358.