5v over voltage protection

Thread Starter

Georgejr66

Joined Oct 12, 2021
7
I have a fairly simple pcb board I designed.
12v power goes thru a diode for reverse polarity protection.
Then the ground and 12v go into a buck converter to drop down to the second voltage or 5.0v.
From here goes to a 470 microfarid capacitor
Then to a microprocessor,lcd and a small hobby style servo.
Today on some units the microprocessor is damaged if they hook up 12v to the 5v wire and some I believe might be getting higher voltage as a guess by moving the servo when off. Either way I need to at least blow the fuse or protect the microprocessor/lcd from a over voltage.
This condition doesn’t happen often so want to keep the design minimal as my project has a small footprint. Can a simple zener diode be put across the same spot as the capacitor with 5.1v so if it exceeds would dump to ground. Would I need a additional resistor or sizing calculation to ensure the fuse blows when the condition is met.
should I redesign the board with something better. Trying to keep cost down and size down but want it to avoid hitting the upper side of 5.5v to keep microprocessor from cooking.
Thanks. I’ll see how to get a schematic uploaded as well.
 

dl324

Joined Mar 30, 2015
16,846
Welcome to AAC!
Today on some units the microprocessor is damaged if they hook up 12v to the 5v wire and some I believe might be getting higher voltage as a guess by moving the servo when off. Either way I need to at least blow the fuse or protect the microprocessor/lcd from a over voltage.
Why are they connecting things wrong? If users can't be expected to be able to follow instructions, the connections should be made foolproof.
 

Thread Starter

Georgejr66

Joined Oct 12, 2021
7
I haven’t been able to repeat but have had a shop that hooked everything up right and multiple parts are showing signs of over voltage smoked/burnt smell. Once parts are replaced works as expected. I’ve started shipping in ESD bags in case static is causing this but not sure that is the problem. Was looking at Zener diode as a failsafe option with minimal cost and design change. I however I’m not sure if that is the right approach. Basic need is to try to limit below 5.5 v to avoid a smoked microprocessor and lcd which both feed off the buck converter which I adj and seal the pot to keep it from changing. Same buck converter is used after replaced parts and no repeat issues. I’m not seeing a high number of these issues but is a delay and confidence issue with customers. So far no one has had a problem but want to solution this so it doesn’t occur.
 

crutschow

Joined Mar 14, 2008
34,285
Keeping the voltage below 5.5V requires a reasonably precision circuit, which is more than a Zener and a resistor or fuse can provide.
You need something like a precision voltage reference and an SCR or series MOSFET.
 

Thread Starter

Georgejr66

Joined Oct 12, 2021
7
Keeping the voltage below 5.5V requires a reasonably precision circuit, which is more than a Zener and a resistor or fuse can provide.
You need something like a precision voltage reference and an SCR or series MOSFET.
Normal operation the buck converter would keep it at or just slightly below 5v.
SCR looks to be part of a crowbar circuit which was something I did look at. I would need to scrap by current boards and redesign the board which I can do. I struggle with this as it is a very small outlier problem but is more then a few. I’m trying to figure out what is causing it but so far I’ve drawn a blank. Work mostly in IT but branching out on a side business. Having protected circuit can’t be a bad thing but does add more points of failure.
 

Thread Starter

Georgejr66

Joined Oct 12, 2021
7
On my 5V microcontroller circuits I use LM2931AZ-5.0 low dropout voltage regulators.
That looks interesting. The servo can draw some amps. On the 12v side I’ve only seen 0.625 amps at full stall or the servo. Servo can be switched to another power source. Just need ground and signal and could drop the 12v at the servo as it has a wide range of voltages it can use.
 

Ian0

Joined Aug 7, 2020
9,671
Use a 5.1V 3W zener and a PPTC fuse or PTC thermistor. The zener will withstand the extra current for the time it takes for the PTC thermistor to change state.
https://uk.farnell.com/epcos/b59985c0120a070/thermistor-ptc/dp/9753095?st=ptc thermitor
With the right choice of components, it can be protected against accidental connection to 230V AC mains.
A crowbar would probably damage other components when protecting the microcontroller. If anything is damaged then the board has to be repaired. Microcontrollers are cheap (although pretty rare at the moment) compared to labour and postage.

The PTC circuit resets itself when the power is removed, and lives to fight another day.
 

crutschow

Joined Mar 14, 2008
34,285
Use a 5.1V 3W zener and a PPTC fuse or PTC thermistor. The zener will withstand the extra current for the time it takes for the PTC thermistor to change state.
The Zener may withstand the current, but will likely go above the 5.5V maximum of the micro before the PTC can react.
 

Ian0

Joined Aug 7, 2020
9,671
The Zener may withstand the current, but will likely go above the 5.5V maximum of the micro before the PTC can react.
PTCs are a bit tricky to simulate, so it might have to be tested with real components!
Putting an active shunt regulator (TL431+power transistor) or a crowbar after the PTC might also work.
 

Ian0

Joined Aug 7, 2020
9,671
I wouldn't rule out the simple zener solution before I tried it - a 1N5338 has a slope resistance of only 1.5Ω. How well it might work depends on the prospective fault current, and that depends on the PTC device. A PPTC fuse would have lower resistance than a Barium Titanate PTC thermistor.
 

Reloadron

Joined Jan 15, 2015
7,501
Welcome to AAC!
Why are they connecting things wrong? If users can't be expected to be able to follow instructions, the connections should be made foolproof.
Well I totally agree with you. Years ago my department head would task me with a project and his closing words were "Ron you need to design this foolproof". My response was always the same, "Mike, that's impossible because fools are very ingenious people".

Ron
 

Thread Starter

Georgejr66

Joined Oct 12, 2021
7
Thanks for all the help… So several ideas here… I’m solving for a unknown reason the lcd and nano board are burnt up and a few cases where they hook up and change wiring to send 12v on the 5 volt wire. I work in IT so I know how you can’t code everything out from people making mistakes.

Could some static discharge cause a nano board and lcd to be damaged? If so then assuming my change to ESD bags will resolve that issue. Only has occurred on shipped units and only a very small percentage of units. This is my biggest thing I want to solve. They from what I can tell have done nothing wrong but a working unit tested doesn’t work when they install it.

For a 12v run down the 5v line…. I’m ok with blowing the fuse which is a 2amp on the 12v side currently. I could supply with a 1amp fuse. Could also put a diode so it would block Voltage coming from the external wire instead of power from the buck converter.

Could someone rotating a hobby servo act like an alternator when the unit is off? It has a higher max voltage (4.8-7.4v dc) and uses the 5v wire connected to the nano. If this is the cause a low voltage drop diode could resolve this. Testing this theory right now.

Kicker for me is once fixed these units don’t have any more issues and are rock solid. Wish I could repeat and have asked many questions trying to get steps done to cause this to no smoking gun.

Going to build some of the suggested solutions and see how I can incorporate easily into the current design.
 

Ian0

Joined Aug 7, 2020
9,671
“Burned up” isn’t an ESD problem! If it dies due to ESD then there is rarely any smoke!
Burned up indicates that there was some large amount of power available when the fault happened.
 

Reloadron

Joined Jan 15, 2015
7,501
Below is what an Arduino Uno board looks like and if I am not mistaken the Arduino Nano power looks about the same.

Arduino UNO power dist and Reg.png

The DC Barrel Connector is designed around a DC 7 to 12 Volt input. Yes, it can exceed 12 VDC but when 12 VDC input is exceeded the voltage regulator tends to run a little hot. The Vin pin is also designed for a 7 to 12 vdc Input. The 5 Volt pin is designed for 5 Volt use including 5 Volts out when using the barrel connector or Vin pin. The 5 Vin pin can also serve as a 5 Volt input. All of the pins for any Arduino board are well explained on the Arduino website.

I have seen boards destroyed because end users try and do things the board(s) were never designed to do. For example the 5 Volt pin when used as a 5 Volt supply out is not an infinite current source. These boards have limits which many people fail to recognize. As I mentioned nothing is foolproof and fools are very ingenious people. When any circuit board is abused beyond its limits bad things can and often will happen.
I’m solving for a unknown reason the lcd and nano board are burnt up and a few cases where they hook up and change wiring to send 12v on the 5 volt wire.
You overvoltage something you can expect overcurrent and the end result is something is likely going to smoke. What else could possibly be done on the above circuit and still keep the bill of materials down? Foolproof can only be taken so far.

Ron
 

Thread Starter

Georgejr66

Joined Oct 12, 2021
7
Below is what an Arduino Uno board looks like and if I am not mistaken the Arduino Nano power looks about the same

The DC Barrel Connector is designed around a DC 7 to 12 Volt input. Yes, it can exceed 12 VDC but when 12 VDC input is exceeded the voltage regulator tends to run a little hot. The Vin pin is also designed for a 7 to 12 vdc Input. The 5 Volt pin is designed for 5 Volt use including 5 Volts out when using the barrel connector or Vin pin. The 5 Vin pin can also serve as a 5 Volt input. All of the pins for any Arduino board are well explained on the Arduino website.

I have seen boards destroyed because end users try and do things the board(s) were never designed to do. For example the 5 Volt pin when used as a 5 Volt supply out is not an infinite current source. These boards have limits which many people fail to recognize. As I mentioned nothing is foolproof and fools are very ingenious people. When any circuit board is abused beyond its limits bad things can and often will happen.

You overvoltage something you can expect overcurrent and the end result is something is likely going to smoke. What else could possibly be done on the above circuit and still keep the bill of materials down? Foolproof can only be taken so far.

Ron
I have the servo/lcd/nano all feed off of the buck converter output with a 470uF capacitor to buffer the servo. The Nano I go direct to the 5v pin so the internal voltage regulator is not used. Servo movement is slowed in code so it never rotate direct to a specific spot but makes a
small movement and has a small ms wait to smoothly rotate and loops thru till at the desired position.

Because the lcd and nano both are burnt when getting parts back I’ve been trying to figure out how. And lack of knowing that I go to well if it occurs how to protect it.

Leaning towards a crowbar circuit but saw the concern noted as well. Figure I’m not the only one wanting to do this but what is the simplest method to protect 5v low amp circuits.
 
Top