Power supplies die during blackouts

Thread Starter

StrongPenguin

Joined Jun 9, 2018
287
I have noticed from some of the blackouts I have encountered in ships, that ocasionally a power supply dies. Typically SMPS.

And if a device is hooked to an UPS, the UPS dies, not the device.

What is it in the power supply that does not like blackouts?

I am assuming what is killing these supplies is the voltage spike before shutdown.

This is onboard ships.
 

KeithWalker

Joined Jul 10, 2017
1,681
I have noticed from some of the blackouts I have encountered in ships, that ocasionally a power supply dies. Typically SMPS.

And if a device is hooked to an UPS, the UPS dies, not the device.

What is it in the power supply that does not like blackouts?

I am assuming what is killing these supplies is the voltage spike before shutdown.

This is onboard ships.
How do you know that it died before the power came back on? It could be a voltage spike that occurs when the power returns that is killing it or it could be a short brown-out condition when the power shuts down or returns. It is probably a characteristic of the generator.
 

Papabravo

Joined Feb 24, 2006
15,753
As I understand ship power systems, they use diesel generators to make electricity to power the motors that drive the shafts and the control fans. What I'm unsure about is if these generators are AC or DC.
 

KeithWalker

Joined Jul 10, 2017
1,681
As I understand ship power systems, they use diesel generators to make electricity to power the motors that drive the shafts and the control fans. What I'm unsure about is if these generators are AC or DC.
In my youth (many moons ago!) they all used to be 400Hz but that has probably all changed now to 110/220V 50/60Hz so that proprietry equipment is not required.
 

Hymie

Joined Mar 30, 2018
970
All SMPS use an oscillator circuit to generate the high frequency switching; power for this oscillator is often provided by an auxiliary transformer winding. But before the auxiliary winding is providing the required voltage (due to the operation of the oscillator), the initial start-up dc power is supplied via a high value resistor from the rectified mains voltage, charging an electrolytic capacitor of a few micro farads.
I believe that it is this capacitor failing that results in many SMPS failing to restart following loss of power. While the SMPS is operational, the failure of the capacitor has no effect – but it is required for start-up as explained above.
 

Thread Starter

StrongPenguin

Joined Jun 9, 2018
287
@KeithWalker I dont really know what comes first, the chiken or the egg. I just assumed that when the breakers to the generators popped, that this produced a voltage spike which killed our instruments.

@Papabravo Our generators are AC, 690V/60Hz, then transformed down to 440V and 230V. Some ships have a system called blue drive, where the generators are (if memory serves me correct) DC, feeding into a frequency drive.
 

Papabravo

Joined Feb 24, 2006
15,753
@KeithWalker I dont really know what comes first, the chiken or the egg. I just assumed that when the breakers to the generators popped, that this produced a voltage spike which killed our instruments.

@Papabravo Our generators are AC, 690V/60Hz, then transformed down to 440V and 230V. Some ships have a system called blue drive, where the generators are (if memory serves me correct) DC, feeding into a frequency drive.
That makes sense since transforming voltage levels with AC is much easier than DC. So If AC is being distributed on the ship to run DC power supplies, why is that different than the electric company doing it. Do they know something the ship builders don't know or do they just have more flexibility in how a shutdown is handled?
 

Thread Starter

StrongPenguin

Joined Jun 9, 2018
287
@Papabravo I am not sure I understand the question. Onboard our ship, almost all consumption is AC, like pumps (so many pumps..) and bow/stern thrusters. I am not so familiar with how the electric company does it.
 

Thread Starter

StrongPenguin

Joined Jun 9, 2018
287
That makes sense since transforming voltage levels with AC is much easier than DC. So If AC is being distributed on the ship to run DC power supplies, why is that different than the electric company doing it. Do they know something the ship builders don't know or do they just have more flexibility in how a shutdown is handled?
My question is why the shutdown mechanism on a ship should be different than the electric company.
Good question. Maybe the companies use some sort of filter on the grid? We don't have anything "fancy" protecting our equipment.
I also remember from back in the day, people would say laptops who spent most of their lives on ships tend to have shorter life, due to frequency oscilations. Don't know if it is just a myth though.
 

Thread Starter

StrongPenguin

Joined Jun 9, 2018
287
How do you know that it died before the power came back on? It could be a voltage spike that occurs when the power returns that is killing it or it could be a short brown-out condition when the power shuts down or returns. It is probably a characteristic of the generator.
Interesting thought. Could elaborate what you think could be the generator characteristic doing this? Maybe it overshoots during ramp up? The data on the generator is very sparse.
 

atferrari

Joined Jan 6, 2004
4,277
In an old SD14 ship, I recall my Timex computer I was using in my cabin, being reset in the rush of the start of derricks' motors (cargo runners precisely). Extremely noisy, easy to hear from inside the accomodation.
 
If anything, I'd expect spikes when the power comes back on, not when it goes off. Therefore, if the power has gone off and you're around and you have time to do so, it might be a good idea to unplug expensive, rare or sentimental equipment from the outlet, and wait until the lights are stable for a few minutes before plugging them back in.

A simple surge protector power strip (before the UPS if you're wanting to protect that) should help if you're dealing with spikes.

MOVs are usually installed between live and neutral, and then between live and ground. There must be a fuse upstream of the MOV for such a design to be safe. Smaller spikes and surges go through the MOV to ground: It conducts over a certain voltage. If there's a really large spike, or a sustained overvoltage (you plugged it into 220 instead of 110 :) ) a larger amount of current flows, which is enough to blow the fuse. Ditto if the MOV fails shorted after conducting multiple surges to ground.

However, rapid cycling on and off, or repetitive swinging between too-low and just-right voltage, can overstress the power supply in different ways. Basically, it has to rapidly compensate for the change in voltage input and try to keep the output steady. Exactly what failed and why, is hard to say without knowing the design of THAT power supply. This sort of thing may also be the "straw that breaks the camel's back" in a power supply that was already pretty old.
 
Top