Very high current power supply

[Braeden Hamson]

I am in the process of building a ~300A power supply with the intention of testing fuses at 300-70A over the course of 0.01 seconds to 60 seconds. So far I have an appropriate heatsink, 1 AWG conductors, and 300A capable contactor. I'm confident that all of my supporting elements are capable of handling the current. My issue is creating the variable current supply. I've tried using 5 mosfets that were each rated to 222A each. I was varying the gate voltage to attempt keep the mosfets in saturation and manage the current. However, it seems that at some point, perhaps when the fuse blows, the mosfets fail. I've found that there is no resistance between gate and source on any of the mosfets (after failure). My power source is a 12V car battery with 600 CCA.

My question is, is this the correct way to go about this circuit? Should I use BJTs, IGBTs, or something else? I'd like to have roughly 5A of resolution.

I'm familiar with mosfet current mirror circuits, however I'm unsure how to determine the length and width of the mosfet silicon in order to figure out the current gain of the system.

I understand this is a dangerous device and I'm taking appropriate precautions.

 

[Bluejets]

Sounds like a good way to loose your head or at least suffer severe burns....DC fault currents can cause ionisation which if handled incorrectly , become a problem, real quick.

 

[Braeden Hamson]

I'm aware, however I have to do this testing for our FSAE electric vehicle. It's not optional. I understand the risks.

 

[crutschow]

I'd like to have roughly 5A of resolution.

Over what current and voltage range?

You do realize that 300A at 12V is 3600 watts, which has to be dissipated somewhere.
That's probably what zapped the MOSFETs.
Using a 6V battery would cut the power in half.

 

[Harald Kapp]

My question is, is this the correct way to go about this circuit?

Imho no.

To test with high current at low resistance you need only a low voltage to minimize losses (as pointed out by @crutschow ).
Drop the voltage down to e.g 1 V or 2 V only using a step-down converter, then buffer this low voltage in a suitable storage medium (e.g. supercapacitor(s). Use the low voltage to force a high current with (comparatively) low power dissipation.

I was varying the gate voltage to attempt keep the mosfets in saturation and manage the current.

Meaning that almost all of the voltage from the battery dropped across Drain-Source of the MOSFETs. Even with a very large heatsink this is a problem, see above.
How are you "managing the curent"? The conventional way would be a current sense resistor in series with the source of the mOSFETs to GND, then measuring the voltage drop across this resistor and adjusting V_GS until the current matches the requirements.


Also your image lacks the return path for the variable Gate-Source voltage. You've probably connectd the circuit correctly and only forgot to show this connection in the image?

 

[Braeden Hamson]

Imho no.

To test with high current at low resistance you need only a low voltage to minimize losses (as pointed out by @crutschow ).
Drop the voltage down to e.g 1 V or 2 V only using a step-down converter, then buffer this low voltage in a suitable storage medium (e.g. supercapacitor(s). Use the low voltage to force a high current with (comparatively) low power dissipation.


Meaning that almost all of the voltage from the battery dropped across Drain-Source of the MOSFETs. Even with a very large heatsink this is a problem, see above.
How are you "managing the curent"? The conventional way would be a current sense resistor in series with the source of the mOSFETs to GND, then measuring the voltage drop across this resistor and adjusting V_GS until the current matches the requirements.


Also your image lacks the return path for the variable Gate-Source voltage. You've probably connectd the circuit correctly and only forgot to show this connection in the image?
View attachment 58986

I'm using a current shunt to measure the resistance, I omitted that from the drawing. It's above the FET though. And yes, the picture you show here above is correct in regards to the variable voltage. (With the red being negative of course)

Is your suggestion to build a step down converter that is capable of supplying this level of current? Then adjusting the set voltage to determine the amperage? If so, that sounds achievable. I think my mosfet idea was the fail fast fail expensive option (oof)

Thank you for the help

 

[Harald Kapp]

Is your suggestion to build a step down converter that is capable of supplying this level of current?

Exactly. Or use the step-down to charge a bank of NiMH cells or even 1.2 V single cell Lead-Acid cells. This way the step-down doesn't have to supply the test current for the fuse. The test current would come from the cells and the step-down is only needed to re-charge the cells.
You may want to look up spot welder circuits. These also use high currents at low voltage.

 

[kellys_eye]

with the intention of testing fuses at 300-70A over the course of 0.01 seconds to 60 seconds.

Far be it for me to tell you your business but don't fuse manufacturers provide such detailed information with their products? There are 'standards' that have to be met before they can retail them as their described function.

Unless, of course, you are making your own fuses?

 

[crutschow]

Can you us AC instead of DC to test the fuses?
If so, you could use a VARIAC to vary the voltage to a 115V transformer.
The transformer could be a modified microwave or standard power toroidal transformer with just a few secondary turns to give a high current, low voltage output.
You would experimentally need to determine how many output turns give the desired output voltage and current.
It would likely be no more than several turns of large wire.

An alternate interesting source of a high AC current would be a soldering gun, which has its tip connected to the 1-turn secondary of the gun transformer, which generates several hundred amps through the copper tip to heat it to soldering temperature in a few seconds.
Its current could also be controlled by connecting the gun to a VARIAC.