2

The datasheet of MCP73831 (a small LiIon charger device) recommends a transzorb on the USB power input (VDD) for input over voltage protection. See Section 6.1.1.2.

Absolute maximum ratings for VDD is listed as 7.0V plus 4kV human body ESD and 400V machine model ESD.

Most transzorb/TVS/varistor devices that allows 5.1V, can be at 7V before the shunt current reaches 1mA and will specify a clamping voltage of 15-35V for higher currents.

Looking at some hobby devices using the same chip does not help. All these Sparkfun product designers clearly ignored the datasheet recommendation.

So what transzorb device did the datasheet author have in mind?

ADDED Apr 1, 16:

As an example of how the voltage waveform may look without any transzorb device during the USB plug-in event is shown below.

Vdd peaking just above 7V when plugging in a 5m USB cable

This is using a 5m USB cable to a standard USB wall adapter.

Rolf Ostergaard
  • 4,448
  • 18
  • 24
  • Looks like this is a registered trademark for a snubber transient diode. You should be able to just choose an appropriate diode. Given the application, probably a Schottky (but this is a guess, not an answer.) –  Mar 30 '16 at 15:03
  • @jdv A diode may suppress a too negative voltage. Not bad, but that's not what we are after here. The plug-in event will create a too high voltage. – Rolf Ostergaard Mar 31 '16 at 05:29
  • Even Vishay, the holders of this trademark, call it one of these: https://en.m.wikipedia.org/wiki/Transient-voltage-suppression_diode basically, two fast diodes back to back. –  Mar 31 '16 at 12:24
  • Ahh I see what you mean. And the question is what type of TVS can actually save this charger chip? Can you actually find a TVS that would work? – Rolf Ostergaard Mar 31 '16 at 17:11
  • Yeah, that's the part I'm not so great at. I'm not sure what the key items in the TVS datasheet you'd need to meet or exceed for your charger. But, I'm smart at searching SE, and I found this: http://electronics.stackexchange.com/questions/22489/choosing-tvs-diodes Maybe this will help? –  Mar 31 '16 at 17:25
  • I thought I'd be smarter and read the MCP73831EV app notes, and there is a reference to transient protection for the regulated car supply (AN947), but no further details. It might just be two schottky diodes back to back. But the actual USB charging app note, AN971, doesn't show any such protection. other than single Schottkys for input protection. Maybe the link from the Wikipedia article might help choose the device? http://www.littelfuse.com/products/tvs-diodes.aspx –  Mar 31 '16 at 17:54
  • @jdv Thanks for the effort. The key issue here is that no transzorbs seems to clamp to below 7V, yet allow 5.1V. So how can you _know_ whatever you do ends up doing actually protects the device? And what did the datasheet author have in mind? – Rolf Ostergaard Apr 01 '16 at 13:20
  • Yeah, I saw that, too. I have no idea how to design around one of these things. Maybe this is why some of these devices are very complicated. The bidirectional ones have upwards of 5-6 diodes in their block diagrams! My feelings are that two generic avalanche diodes back to back would give you lots of protection. Again, not an answer because I can't prove this. –  Apr 01 '16 at 13:45
  • A DZ RC snubber (as found in flyback converters) might do the trick. – Peter Smith Apr 01 '16 at 15:27
  • @PeterSmith A 5.6V zener across the input does indeed limit the input voltage to about 6V. Was that what you were suggesting? But would you call a zener diode for a "transzorb"? – Rolf Ostergaard Apr 02 '16 at 09:41
  • @Rolf For low transient currents a zener should work. No high current protection though. Transorbs in SMC formats tend to have slow clamping (about 500nsec. The devices from Littelfuse are very fast action though). A zener with a small resistor in series to limit the current in a large current fast transient scenario could be a potential solution, particularly if there is an RC snubber in parallel. – Peter Smith Apr 02 '16 at 12:53

0 Answers0