Protecting an AC to DC Power Supply - with a fuse

Thread Starter

Lumenosity

Joined Mar 1, 2017
614

24v - 20A Switching Power Supply
Is there a reason you could not put a fuse on the V+ line coming from this power supply to protect it from over current?
Specifically, an inline ATO 20Amp fuse

Condition 1).
You accidentally connect this power supply to a 36v battery. Would a fuse on the positive line(s) protect the Power Supply?

Condition 2).
You accidentally short the positive and negative leads.

Condition 3).
Any possible current overload beyond 20Amps

Thanks
 

Thread Starter

Lumenosity

Joined Mar 1, 2017
614
Except for the purejunk supplies, they all have an incoming power fuse on the PCB. Some are overload protected and some are not. Anoutput fuse might protect an outside load and wiring, and thus bee a good idea. Only an external series diode will protect it from damage by connectingto an external higher voltage source.
Thank you.
Connecting to an external higher voltage source needs a diode for protecting the PSU.
This is exactly what I needed to know.

Is there a formula to determine the size diode needed or will any diode do?
And I assume the diode will need to oriented with the cathode facing the external voltage source (on the Positive line) ?
 
Last edited:

Tonyr1084

Joined Sep 24, 2015
7,853
The diode you choose must be capable of handling the full 20 amps the supply is capable of putting out. Otherwise you'll blow the diode every time you draw more amperage than the diode can handle. In your case I'd recommend overkill. That is to say - use a diode capable of handling 1 1/2 times to 2 times the 20 amp rating. (30 to 40 amps). This way you'll never run your diode hot. Even if you're not using the full 20 amps.
 

Tonyr1084

Joined Sep 24, 2015
7,853
Well, if 24V is critical, yes, a crowbar circuit. But 23.4V is pretty close to 24V, so I don't think a diode will cause any issues.
 

Tonyr1084

Joined Sep 24, 2015
7,853
1621442037879.pngOK, a trim pot. Not what I'd normally consider an adjustable PS. This doesn't answer what I was thinking of - is this one of those adjustable supplies with panel meters and such that are designed for bench top adjustability. So - I would assume - and we all know how to spell ass-you-&-me - it probably can be adjusted at least 0.6V above 24VDC. But that's just an assumption on my part.

For my money (what little there is of that) I'd opt for a 30 amp diode as opposed to a crowbar circuit. But that's just one opinion; and opinions are like arse holes - - - everybody has at least one, and some of them probably stink.
 

Thread Starter

Lumenosity

Joined Mar 1, 2017
614
Yes, we're talking cheap Chinese PSUs, where many YouTube reviewers have opened them and found many used parts inside,
crushed capacitors, Mosfets not making contact with the heat sink etc....roll of the dice.
They are readily available on Amazon and ebay. I do have a few better PSU's on hand such as Mean well. They really are better.

The diode solution is just so much easier to implement. I'm not concerned about less than a volt drop.
Just something to help guard against connecting to a higher voltage rating than the PSU.
I haave a bench full of battery banks some in parallel, some in series.

But one question: Wouldn't I need to get not only a 30A diode, but also one with the correct voltage rating?
How will the 30A diode know to allow the 24v to pass.....but block anything above it? Just buy a 24v 30A diode ?

I looked and couldn't find such a beast.....DigiKey
 

MrChips

Joined Oct 2, 2009
30,712
Well, if 24V is critical, yes, a crowbar circuit. But 23.4V is pretty close to 24V, so I don't think a diode will cause any issues.
At 20A the diode forward voltage is not going to be 0.7V.

I was thinking more about the dissipation caused by the drop. The power supply can be adjusted to compensate for the drop.
At 20A the diode forward voltage could be 3V.
That's 60W power dissipation. Yikes!
 

Tonyr1084

Joined Sep 24, 2015
7,853
Wouldn't I need to get not only a 30A diode, but also one with the correct voltage rating?
How will the 30A diode know to allow the 24v to pass.....but block anything above it? Just buy a 24v 30A diode ?
I confess to not knowing as much about electronics as MOST experts on this site. Diodes are rated in max voltages they can handle before they break down. (see last paragraph)
At 20A the diode forward voltage could be 3V.
That's 60W power dissipation. Yikes!
I must bow to your superior experience and knowledge. I've never heard that before. But that also explains why the 40A diode I bought last year is heat sink mountable (and probably recommended).

I suggested a 30A diode because it's more than the PS can deliver. If the TS wants to use the PS for low power applications such as 24V @ 2A, then a fuse in the circuit would protect against blowing the diode out (if someone used a 1N400x diode). At AutoZone I bought a self resetting 20A fuse for testing on my car. It was blowing the 20A fuse that ran the fuel pump and the ECU. The pump was bogging down and blowing the fuse. Without harming anything else, I was able to determine the failure and fix it. So self-resetting fuses can be an option as well.

I'm not opposed to the crowbar solution, but I'm a simple guy who thinks in simple terms. Sometimes that proves dissatisfying, but I live with it.

Diodes are basically one way gates. Current flows in one direction only. In the reverse direction they block current. They all have their own characteristics - such as "Break Down Voltage". The point where the diode can no longer block the current in the opposite direction when the voltage exceeds the BDV. Zener diodes are constructed with precise and predictable BDV's, meaning they conduct in the wrong direction when voltages rise above a set design limit. You wouldn't want a zener in this application. And you don't need a diode that can handle more current than you're going to draw. So even though your PS can deliver 20A, you may be only using 2, 3, maybe 5 amps at a time. But that decision is up to you. It's a part of engineering even I understand; you have to have components that can work at the levels of voltages and currents you are planning on working with. So you don't need a 24 volt diode, you need a diode that can (likely) handle 500 volts (for example) and the current you're going to be drawing.
 

djsfantasi

Joined Apr 11, 2010
9,156
I would go higher since TS states he has many series batteries.
That’s a good recommendation.

I had a sneaky suspicion that people haven’t been responding to the issue of just finding a diode. And the TS was searching for very specific parameters.

His success would be improved by searching for a range rather than a specific number.

I wanted to put it out there that a higher voltage rating was fine and 30A diodes are available at higher voltages.
 

Thread Starter

Lumenosity

Joined Mar 1, 2017
614
The link I posted in #17 trims all that an shows only stocked items but obviously TS didn't click on it.
Obviously ? I did.

I was looking for the diode I mentioned as stated above, I do not understand how my goal would be accomplished with a diode voltage rating higher than I am trying to protect against?
Admittedly, My ignorance and lack of knowledge.

My goal is protecting from a voltage difference....not current surge or reverse voltage (I think)

24V power supply connected to a 36v battery.
 

Tonyr1084

Joined Sep 24, 2015
7,853
I do not understand how my goal would be accomplished with a diode voltage rating higher than I am trying to protect against?
This is just how diodes are manufactured. Here's a data sheet for the 1N400X series of diodes. IF the link worked you can see that the 4001 has a max recurrent peak reverse voltage rating of 50 volts. That means it can block maximum current at a peak reverse voltage of 50 volts. The 4002 can withstand 100 volts. The 4007 can withstand 800 volts max recurrent peak reverse voltage. That's how they're made.
 

Tonyr1084

Joined Sep 24, 2015
7,853
Diodes are like capacitors. A 100µF cap at 16V is good UP-TO 16 volts. Whether you use it for a 5V circuit or higher, you can't exceed the 16 volts without risking damage to the cap. The diode is the same way. They're rated UP-TO a given voltage. It doesn't mean you need a 16 volt diode on a 16 volt circuit, though that would be OK. Diodes have a maximum capability. Getting one that can handle 200 volts is the same as using a cap rated for 200 volts. You don't have to use the total voltage the device is rated for - you just have to use a device that is rated suitably or higher. Going higher gives you extra security that you're not likely to exceed its rating. Using a 1N4001 (rated at 50V max) or using a 1N4002 (rated at 100V), both will give you the same performance. If you're going to be getting close to 50V then go with the 4002 diode. You have the margin of safety that you're not going to blow it up if you get a spike in voltage, which inductors can cause.

Why use a 200V rated diode as opposed to a 50V? There is no specific reason, but both cost the same and both do the same job. Just the 200 has the ability to withstand higher voltages. If that doesn't explain it - then I can add no more.
 
Top