Advice on USB-powered LED Lamp circuit

[AndyCW]

Hi,
My friend is a teacher and has his students build a simple project to teach them the basics of electronics. KIt looks like this:

Many thousands of these have been built over the years!

Just recently, he has had a report of on ‘new’ PC’s sometimes, despite the product working well, all soldering being apparently good and no short circuits, a USB power surge type warning will appear on PC screen resulting in mouse becoming inoperable and you have to knock PC off manually and restart it. This has never happened on older PC’s.

There was one report of an Apple Mac had its Magsafe board blow and this USB minilight circuit was blamed - although when they checked the board out later on a PC, it seemed to be operating perfectly.

Questions:
1.Why or under what circumstances, looking at the circuit, would a power surge warning appear?
2.Which component is likely to be causing the problem? (if one exists)
3.Would this product blow the Magsafe board of an Apple PC instantly as has been suggested to us and are there any differences between PC’s which run Windows compared to Apple computers?

Thanks.

 

[KrisBlueNZ]

Hi Andy and welcome to Electronics Point

If that capacitor, C1, really is 1F (one farad), then yes, it will draw a surge current when you first plug the device into the USB socket, as it charges up. I doubt it will do any damage, but the problem is easily fixed by insering a 47Ω 1 watt resistor in series with either one of the connections to the USB port power pins.



Edit:

This change may have an impact on the experiments that that board is used for, because it limits the amount of current that can be drawn from C1 without its voltage dropping, and it greatly slow down the initial charging of C1 when the circuit is plugged into a USB port.

That value, 47Ω, limits the current to 100 mA, which is the USB standard's maximum load current for a device that doesn't negotiate a higher current. It also slows down the charging of C1 - it will take several minutes to reach full charge.

For experimentation, rather than a commercial product, you can probably get away with drawing 500 mA from the port, using a 10Ω 1W or 2W resistor. This will charge the capacitor almost fully in under a minute, but it will get hotter than the 47Ω resistor during this charging time.

 

[garublador]

To elaborate a little, as KrisBlueNZ said, that circuit does not follow USB specifications, especially with that giant capacitor. Also, the host side implementation of the USB host voltage supply isn't super consistent from product to product. As long as they meet the minimum USB requirements they're OK to sell. Most have some margin in them, so they will outperform the minimum requirement to some extent. So any given USB host could do anything from just burn up once that device is plugged in to work fine with no hiccups and all would be acceptable. FWIW, neither of those would surprise me. It's dependent on the hardware design of the 5V output on each USB port, so the OS probably has little to nothing to do with how well it will perform. It may change how the overcurrent is reported, but it probably won't change how the hardware performs.

My guess is that the older PC's work better because the USB circuitry is just more robust. As USB hardware has matured, many more "less expensive" parts are available that meet the minimum requirements, but have less margin. That's just a wild guess, though. We have some anecdotal evidence from our customers that newer PC's don't seem as robust in terms of USB performance as well.

I'd also suggest double checking that capacitor value. 1F seems excessive.

 

[BobK]

Why is there any capacitor there at all? It should be unnecessary unless it is really a 1F, which I doubt, and is intended to keep the light on after unplugged from USB.

Bob

 

[Gryd3]

Do as @KrisBlueNZ has stated by adding a resistor in series with either power or ground from the USB plug.
I'm a little surprised that there was no consideration taken for the initial charging requirement of that capacitor.
If you still want to have a decent charging time, I think it would be best to advise the students not to use unapproved hardware on their computer and use a 3rd party USB charger like those for cell-phones and iPhones. Some of them have outputs capable of up to 2000mA, and should all be able to output 500mA with out any concern.

As far as being rugged... I've never had a problem with my USB ports, but with the advent of all of these cheap chinese USB fans, fridges, and toys... I think it's best to monitor and manage the current output to these ports, as a faulty USB device could cause some damage.

 

[AndyCW]

Thanks everyone. Very happy I joined this forum