Im helping out my nephew who is interested in electronics and came to me because I specialize in automotive electrical (but Im pretty green when it comes to the controller side of things as I deal with the harnesses and physical components external to the controller). He is a home schooled kid and as a summer project, he wants to make a programmable fuse that would theoretically replace a regular mechanical fuse. I explained that car manufacturers do just this nowdays and I'd be happy to help him out. I tried my hand at this years ago and wasnt really worth my time, but he is getting into cars and really wants to do this.

What Im looking for is some direction to the right components I should consider for the project beyond the microcontroller. A typical automotive load that a regular ATO style fuse would handle is 5-30A at our nominal 14v. My thinking was to use a shunt style current monitor circuit with an ADC back to the micro and an output from the micro to a MOSFET that would be able to PWM the current to its programmed limit or if its simpler, to just cut off the current flow if it is a dead short. Im not interested in dealing with motors and the inrush issues and hoping more to make something that would handle halogen or incandescent lights.

I've seen some components from TI and MAXIM that looked like they would work, but I hoped that I could get a bit of advise on what I should look at in particular. If you need any particulars, I will be happy to give them if necessary.

 

One of the key requirements would be a very low voltage drop across the fuse so using a Hall Effect current sensor which only adds a tiny voltage burden to the circuit might be a good idea. Something like the one shown at the link below could work, but that particular one is only rated at 5 amps.
https://www.sparkfun.com/products/8882

For the PWM to limit the current, you would probably want to use a buck converter topology. There is a clear explanation at this link http://www.learnabout-electronics.org/PSU/psu31.php

 

One of the key requirements would be a very low voltage drop across the fuse so using a Hall Effect current sensor which only adds a tiny voltage burden to the circuit might be a good idea. Something like the one shown at the link below could work, but that particular one is only rated at 5 amps.
https://www.sparkfun.com/products/8882

For the PWM to limit the current, you would probably want to use a buck converter topology. There is a clear explanation at this link http://www.learnabout-electronics.org/PSU/psu31.php

Didnt even think about hall effect, would you think this is what I should be looking at for current sensing? http://www.digikey.com/product-detail/en/ACS714LLCTR-30A-T/620-1229-1-ND/1824989


Im still looking at the PWM option over just cutting off current at a certain setting. It would be pretty easy to set up the circuit to open up at a given voltage input from the current sensor and act more like a programmable circuit breaker.

 

Excellent choice!

Yep, the circuit could be very simple.

 

It is not advisable to use a PWM method to control an automotive circuit unless you know the characteristics of the load. Wiring looms also introduce a fair amount of stray inductance and potential crosstalk which can interfere with modern digital comms between the ECU and outlying digital modules or subordinate ECUs.

Also, introducing multiple components to 'replace' a fuse can itself be adding potential failure points in the circuit.

Electronic components for automotive use require additional certifications and must be so rated.

 

Go to the Diodes, Inc web site, and check out the ZXCT1009 and all of its brothers...

 

That's an interesting part. The ACS713 has a much lower resistance and can handle rated currents with no heat sinking.

 

That's an interesting part. The ACS713 has a much lower resistance and can handle rated currents with no heat sinking.

The ZXCT series are designed to be used with an external shunt, like the OP's fuse...

 

Precision power shunts add cost and require cooling.

 

One of my airplanes has a typical +60A_0A_-60A zero-center ammeter, which has an internal 1mV per 1A shunt. I wanted to log the net battery charge/discharge current into a channel of a data logger so I put a bidirectional ZXCTxxxx across the ammeter terminals. That produces a ground-referenced voltage proportional to ammeter reading...

 

It is not advisable to use a PWM method to control an automotive circuit unless you know the characteristics of the load. Wiring looms also introduce a fair amount of stray inductance and potential crosstalk which can interfere with modern digital comms between the ECU and outlying digital modules or subordinate ECUs.

Also, introducing multiple components to 'replace' a fuse can itself be adding potential failure points in the circuit.

Electronic components for automotive use require additional certifications and must be so rated.

Have you read what the point of this project is for? Its for a 14 yr old that is interested in electronics as well as cars. I dont care if this turned into building a full on SoC if it helps him learn and he is successful.

Yes, replacing a 10 cent fuse with a uC based system is nowhere near simple or cheap, but that isn't the point. Hell, building an Arduino so you can blink an LED is pretty expensive as well, but again, that is not the point. The objective is to learn pcb layout, component selection, uC probramming and have a device that works and does what you set out to do.

To add, the components I was looking at are from Texas Instruments and Maxim, whom if you were unaware, make a whole lot of automotive grade components. Beyond that, other than IC's, most auto grade components are rated to +125c for use in engine compartments. I will guarantee you that the same diodes, resistors, capacitors and transistors found in a TV or other home electronic device can and are used in automotive applications. I didnt ask about micros or power regulation, I asked about current sensing and nothing more.

I'd ask to focus your replies on the goals of the project. If it costs me $50, I dont care as the goal is to teach my nephew and allow him to add that to his resume of capabilities for college. I'm sure if you keep that in mind, you'll be much more helpful here.

 

I will guarantee you that the same diodes, resistors, capacitors and transistors found in a TV or other home electronic device can and are used in automotive applications

If you can guarantee this then you clearly are ignorant of what you don't know.

I draw a simple reference. A capacitor must be rated for its application. Voltage, ESR, ripple, operation temperature and MTBF are a few of the key specs. A n average capacitor found in a consumer appliance is highly unlikely to survive an automotive environment for very long.

I have been designing and manufacturing automotive ESAs since 1990. Perhaps you have more experience.

Good luck with your project.

 

One of my airplanes has a typical +60A_0A_-60A zero-center ammeter, which has an internal 1mV per 1A shunt. I wanted to log the net battery charge/discharge current into a channel of a data logger so I put a bidirectional ZXCTxxxx across the ammeter terminals. That produces a ground-referenced voltage proportional to ammeter reading...

Sounds intriguing Mike.
I'd be interested in the schematic for the application as I am doing a project with ACS713s and your proven solution might be more elegant.
Perhaps a PM to not add noise to the TS's interest?