What's inside my LED driver?

I have a strip of RGB LEDs in my bedroom and have decided I want to use an Arduino to control them. Mainly due to the PWM frequency of the supplied driver. Too low.

I opened up the driver/controller and was surprised how teeny the transistors (or are they FETs?) are. Will these handle the 2A they’re supposed to. Sticker says 6A max for the controller so 2A per channel. I found a guide to hacking the controller on another site and the ones I’ve seen have much bigger parts for each colour.

Yes those are FETs, you can read all about them here:-
http://www.onsemi.com/PowerSolutions/product.do?id=MGSF2N02EL

But yes they will handle 2.8A according to the headline data sheet figure.

Excellent, thanks for the info. I did try Googling the VS30 on top but didn’t come up with much. How do you know what type they are?

I did try Googling the VS30 on top but didn't come up with much.

Well try "VS30 FET"

How do you know what type they are?

Not sure what you mean. Did you down load the data sheet.

I didn't think to try VS30 FET. Although I just tried it now and the top result was what you linked to. But I don't see VS30 mentioned there anywhere so wouldn't have known if that was the same thing or a completely different FET.

It must be in the meta data for the page. I have seen this sort of thing before for small surface mount components.

The FETs are indeed tiny, but notice the heatsinks they provide on the PCB. That is the critical part.

I did notice the large pads and was wondering why the tiny fet was in the middle of it. Thought it might have been for heat or a bigger part that they’d decided not to put on.

I tested a section of the LED strips with a BC547. It shows the emitter on the left side with the flat section at the bottom. If I try it that way the lights stay on. If I turn the transistor around so the emitter and collector are reversed it works.

I’m not sure why. Is my circuit correct?

2015-01-28-07-41-35~01.png

A BC547 has a maximum current rating of 100mA so that will not light up many LEDs. So if you have 12V with 3 LEDs in series with each set of three pulling 20mA, if you have 15 LEDs then that is at the limit and therefore too much.

The pinout you drew is not very clear, here is the pin out from the data sheet.

BC547.png

Ah, now that pin order is the opposite of what is shown in the eBay listing for what I bought. That would then make sense if it's meant to be that way.

I was just testing different PWM frequencies with a strip of 3 lights so I'm not overloading the BC547. I'll use the FETs in the controller when I light up lots of them.

Thanks for your help.

This is what eBay shows for it.

Screenshots_2015-01-28-17-24-14.png

Sometimes there are different variants of a transistor with different pin outs, however these are normally signaled by a suffix letter after the transistor number.

I have had a quick look and I can't find any variant of the BC547 with the pin out you show, so I guess the eBay seller got it wrong.

Yeah, I think you're right, the listing was wrong. I typed BC547 into a Google image search and all the images show the same layout as you posted.

At least I know now. I'll have to swap one of the other circuits I've built over now. Strangely that works fine.

Well not so strange. An NPN transistor has a degree of symetry, if you swap the emitter and collector over it still works as a transistor but at a very much reduced gain.
More than one student of mine has told be that knowing that got them through a job interview. It is not the sort of thing they normally teach but you tend to know if you have experience.

I read something about reverse active. Is that how I'd accidentally connected it?

Why would the LCD I used the same setup dim correctly but the LED strip just stay on when they were both reversed? The 12V at the emitter instead of 5V change things?

Thanks for the info. Interesting.

Why would the LCD I used the same setup dim correctly but the LED strip just stay on when they were both reversed?

Because of the amount of current each takes. The LCD takes bugger all and so the gain being low doesn't matter. With the LED the gain needs to be higher to work.

Right, I think.

The gain determines how much current flows from C to E depending on the B current? Is that right?

I still don't quite understand why the LED is on when the base is low when reversed.

Yes that is what gain is.

It is not helpful to speculat why something happens when a component OS wired up wrong, then all bets are off.