Building your first circuit is empowering, but make sure it’s not too empowering. [Jon] sent in a great tutorial of power protection circuits to make sure you don’t release the mystical blue smoke that make electronics work.
There’s an in-depth tutorial of the classic series diode that’s the simplest of all power protection circuits. There’s not much to it – just a diode that provides reverse polarity protection. A fuse and parallel diode doesn’t have the voltage drop a series diode has, but doesn’t do anything for an overvoltage. A P-channel MOSFET gets around the problem of voltage drop, and [Jon] gives us some really nice empirical data to demonstrate his testing setup.
There’s a ton of nice write ups on [Jon]’s site that are perfect for getting ideas for projects like ten switches on one pin and some strange stuff [Jon] picked up at his Goodwill. If you’ve got any tutorials on general electronics, be sure to send them in on our tip line.
Is it just me, or is the site down?
Link is down for me too
Crap. When the site gets a post here, the mad rush of hits slows it way down! Sorry to both users of the site and to readers here.
Jon
The site is sporadically working, i guess it may be slightly overloaded…
Couldn’t you also use a npn mosfet in the 0V-line with its gate tied to V+? Heard that they were cheaper and (generally) have lower on-resistance..?
yep, also down……
I like the LTC4365 circuit with one caveat. He says -40 to 65 VDC is enough to protect a circuit in an automotive environment. This is not complete. Load dump occurs when the battery is disconnected from a vehicle while it is being charged by the alternator (i.e. when the engine is running). This can cause voltage spikes upwards of 100v for hundreds of milliseconds. I would put some transient voltage protection on the input of the LTC4365 for a complete protection solution if it were to be connected to an automotive electrical system.
Yes, this will work just as well with a NFET on the negative line, however it is customary to leave the negative connected as a grounding line.
If an AC/DC power adapter has ground connected through it, it is almost always connected to the negative output.
Here is another scheme using a pnp transistor if anyone is lacking in p-channel mosfets: http://www.geofex.com/Article_Folders/cheapgoodprot.htm
That page says Vout = Vin for positive Vin, but there will be the Vce(sat) drop for the PNP transistor, so Vout will be about 0.1 V to 0.3 V less than Vin, depending on the exact transistor used and the load current.
Also, look how complex the PNP power protection circuit is compared to the p-channel MOSFET version! The PNP circuit requires 7 components, whereas the MOSFET circuit requires only one. That’s a huge difference in circuit complexity, board space, and assembly effort. Not to mention that a $0.12 MOSFET will provide dramatically lower voltage drop (and thus less wasted power and generated heat).
The graphic for the Linear Technology chip is wrong. At the top it says “LTC-3456” when in fact the chip part number is LT4356-1. There is no ‘C’ in the number, and no dash/hyphen.
LT4356IMS-1#PBF, the SOIC-10 version of the chip, is in stock at Arrow for $3.39 (USA).
I also agree with [tony] that a TVS should also be considered essential in most circuit power supply designs. You just never know when lightning will strike, so to speak.
For most designs the LT IC is just to expensive. The guts of many electronics are $4 in total, no way to spend another 4 on a single “super diode”
Sorry for typos on the Linear Technology part number. It’s actually LTC4365. I guess my dyslexia was acting up that day. The links to the data sheet and ap notes do go to the correct pages.
It’s in stock at Digikey for $3.67 in single quantities.
Nothing I build costs as little as $4 in parts even with cheap Chinese circuit boards. The P-Channel MOSFET is a good method of reverse polarity protection that adds little cost to the BOM but doesn’t handle the under/over voltage conditions that the LTC4365 does. It’s far more than a “super diode.”
Is an additional cost of $4 (dropping to less than $2 in quantity) worth it to your device? Depends on the application and what happens when the device fails. If you have to supply a new one, paying for next-day shipping and somebody’s time to replace it, $4 is cheap insurance. If your thing blows up when you’re trying to impress your buddies, $4 may or may not be a big deal.
Jon
You can improve the crowbar diode by using a zenner diode. This will protect to both reverse and over voltage.
The advantage(or maybe disadvantage) of the zenner is that if a high current will pass through it, after the fuse is burn the diode will melt and turn into a short and thus it will have to be replaced.
Interesting. Can you elaborate on how the Zener diode would fail? Wouldn’t the current quickly blow the fuse first (if the fuse value is properly selected) and then current flow would cease, meaning the diode would be safe? Why would the Zener crowbar diode fail when a silicon diode apparently would not?
Man, really wish this site was still up
The site is still up. The opening picture is not working at the moment – a recent upgrade screwed up some links.
There are a couple related articles on Digital-DIY.com (no www in front)