Current limiters/overcurrent protection advice needed 1

[EVConcepts]

I am designing a portion of an automotive circuit which will provide a 12vdc output when activated. This circuit in particuliar is a dimmer control circuit using the PWM supplied by the vehicle.

With the following requested specs in mind, how does one design a circuit properly?

1. Automotive 12vdc circit is usually 13.5 average
2. Circuit design is currently a Power MOSFET of 10a to allow up to 1a output.
3. PWM input is only rated at 20ma, 12vdc.
4. If output exceeds 1a, circuit output trips off.
5. Another circuit on the board can provide stable 12vdc if needed.

In a nutshell, I've got a 20ma, 12vdc that provides a PWM. I need to "amp up" the current to drive the same PWM to other automotive circuits with backlighting. The requirement that I have is to have the circuit shut off cold if current is above 1a (or in that ballpark) and reset automatically when it falls under 1a.

If I don't use this circuit, I believe I'd overload/overheat the MOSFET if the current gets too high. It can also serve as a short protection.

I also have 2 other similiar circuits with the same setup but for different functions. When a solution is found, is it necessary to use 3 seperate circuit protection devices or can one circuit protection device be designed to protect all three?

 

[itsmoked]

I'm confused... Why do you think the vehicle is going to provide PWM?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[EVConcepts]

This particuliar vehicle is providing PWM's for domelight dimmer. I plan to tap the circuit which is 20ma 12vdc and use it to drive the MOSFET for other aftermarket components to use. I suppose the newer vehicles are getting more technical. It does makes sense to use PWM technology for power effeciency purposes anyway.

 

[zeitghost]

The battery supply is about 13.8V.

With enormous spikes from the alternator of 100V or more...

Your circuit will probably have to survive reverse polarity battery connection too...

There are MOSFETs available that have built in overcurrent protection.

Automotive design is a joy.

 

[itsmoked]

Still having problems here.

Dome-lights don't have a constant dimmer PWM. Dome lights are on, off, or fading in a rapid PWM from full to zero. How can you harness that signal for something useful?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[EVConcepts]

My mistake, I meant to say dashlight dimmer as opposed to domelight dimmer. Sorry for the confusion!

 

[MacGyverS2000]

Is this a professional product?

Dan - Owner

http://www.Hi-TecDesigns.com

 

[Comcokid]

I've in the past designed many things for vehicle systems, load dump (vehicle transients), and actually PWM dimmers for vehicle instrumentation.

Check/reverse engineer the dash PWM signal you plan to use. Is this system switching ground to the dashlights (N-channel MOSFETS) or power (P-channel MOSFETS)? This will probably determine if you need to use N or P MOSFETS in your design as the PWM signal may be reference to one rail or the other. Probably using N-channels for cost reasons.

Reverse voltage may not be a problem. A MOSFET has a back diode. In series with a lamp load under reverse voltage, it lights the lamp.

Load dump can be somewhat handled by using a higher voltage MOSFET. May have to look somewhat at protecting the gate as it is usually rated for 20V max - it may be that the PWM signal is already limited - you need to check. A zener of less than 20 volts can be used. A zener circuit can also be used to turn-on the MOSFET in overvoltage to protect the MOSFET if you want - However, this puts the transient on the bulb (or LEDs).

In a PWM situation, you typically don't need to heatsink the MOSFETS since Rd-on is so low, they don't generate heat. If you want some heatshinking, usually a little PCB copper will suffice.

For bulbs, cold filament inrush is typically 10X the bulb current. Size your MOSFETS current by approximately 10X. Also look at the datasheets for pulse load ratings. They are much higher that the average current load. MOSFETS are relative inexpensive compared to PTC self-resetting fuses and other current protection methods. Sometimes it's easier to just size the MOSFET current up to make sure under a short condition you blow the accessory fuse rather than try to select a MOSFET to just barely meet the load needs. Just because you may decide to use a 30 Amp MOSFET, you don't need to put a huge heatsink - remember, the higher the current of the MOSFET, the lower the Rds-on, and the lower the heating of the MOSFET. Even when blowing a 30 AMP fuse, this takes only a second or two, and the thermal mass of the MOSFET is enough to absorb any transient heat from instantaneous power dissipated without any significant heating.

If the equivalent resistance to the gate (all resistance of your protection, other circuits, etc) is on the order of 4K to 12K Ohms, you slow down the turn-on of the MOSFET. This trades a slight amount of heating for reduced possibility of generating RFI - slows down the switching edges. Resistance value somewhat determined by PWM frequency.

Need to watch the PWM frequency. Some larger lamps can begin to have the filament loads "sing" with the PWM frequency - you hit the resonant frequency of the filament. It can be slightly annoying.

Check your accessories. If you are dimming lamps in switches, they are typically wired for power-switching along with the load the switch controls. If you PWM is switching ground, it will not control the lamp in the desired manor. Also, when the accessory switch is off, the switch lamp sees a ground through the device the switch is controlling. A PWM lamp dimmer can make the switch lamp still light even if the switch is off. Depending upon your lamp loads, and their power or ground reference, you many need to have two PWM MOSFETS - one P and one N to control all of your various lamps. Otherwise you can get into the situation where the dimmer control is opposite on some lamps.

I've rambled on her a bit in a sort of memory dump. Hopefully somewhere I've provided some information for your problem.

 

[EVConcepts]

The vehicle is supplying a positive 12vdc, 20ma PWM. This info was from the mfg's upfitter guide.

The aftermarket equipments have dedicated dimmer circuits where normally in other vehicles, we connect direct to the closest dimmer circuit such as the ash tray lamps that are often on the same dash light dimmers. While I could tap into one for this vehicle, I'd rather leave it alone since I have more than one vehicle to work with and want to simplify wiring connections.

I never heard of "load dumping", how does it work? I've only started working with electronics for the last year and it is really fun to learn & try many circuits!

 

[itsmoked]

Load dumps. When an alternator is putting out a large current and the battery is breifly disconnected the large current translates into a VERY large voltage spike. 2-300V. There is little energy involved but no energy is required to toast a lot of electronics.

You need to research "load dump" on google. If you don't understand it you shouldn't be designing for automotive apps as you will be caught out(badly).

Go to National's site too and reasearch it. They may have an appnote on it. They actually have automotive voltage regulators that are designed for reversal,(idiots jump starting), and load dumps.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

The more severe load dump is when the alternator is producing lots of current and has a correspondingly high excitation. When load is "dumped" (battery disconnected, for instance), it takes the voltage controller some time (tens or hundreds of milliseconds) to reduce excitation correspondingly. During that time, a 12 V system can have 50 - 70 V DC output from the generator. This is different from the inductive kick-back. It is quite powerful and should be avoided.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[EVConcepts]

Gotcha,

I called them "spike backs". Now I know what load dumps are... I do have my circits with 1N4002's across the relay coils and I have an input power supply circuit with capacitators & zener diodes designed for the automotive environment. I don't have any CMOS or other sensitive components in my circuits. Just MOSFETS, transistors, resistors, relays, diodes, and leds to take a cluster of different circuits in one module. Are my transient protection options pretty broad already?

I finally found a solution for current limiting applications... PPTC's. Polyfuses & Polyswitches were the two I garned much information on and settled on a few for different current ratings, all at 16v. I plan to ensure that my MOSFETS are twice the size of the rated clamping voltage of the PPTC's anyhow. Sounds about right?

 

[MacGyverS2000]

ROFLMAO!
"I don't have any CMOS or other sensitive components in my circuits. Just MOSFETS, transistors..."


Dan - Owner

http://www.Hi-TecDesigns.com

 

[Comcokid]

Uhmm . . . Ahaaa . . Load dump does not refer to the kick-back from a relay. Load dump as a general term refers to a variety of voltage spikes that arise in a vehicle or other alternator/battery systems. Some of these spikes arise from inductive V=L di/dt sources like a relay coil but are much more powerful.

Starter motors, seat motors, fan motors are a lot larger than relay coils and store more energy. Wires run parallel to each other in automotive harnesses allowing spikes to appear on other circuits (co-induction). An alternator charging a dead battery with, say, 45 Amps at 12 volts when the corroded battery terminal makes intermittent contact can create a whopper of a voltage spike (this is the specific defination of load dump). In the cold of winter, cars can be jumped from a 24 volt battery system in a tow truck. See the following old threads.

Thread67-98147 Thread240-100040 Thread240-98170

One of the old thread give a SAE spec for load dump. I have tested devices on load dump simulators built to the SAE load dump spec. Makes for nice fireworks. Transistors can dissappear in a bang with smoke, flame, and plastic shrapnel.

PTC are one of many devices or combination of devices you can use for protection.

 

[itsmoked]

Well now just a minute there.. You realize that PWM means an on/off signal that varies its on time verse its off time. More off = dimmer. I have never seen dome lights that let you adjust their brightness! They all dim slowly automatically on a sliding PWM you have no control over. Is this the signal you plan to use?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[geekEE]

Keith, the OP clarified in an earlier post that he meant dashlight dimmer rather than domelight dimmer.

 

[itsmoked]

Agreed... Strange. Oh, I know what happened. Eng-tips went down and I saved my response then sent it from a different comp after-the-fact. Answering twice.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[EVConcepts]

Well, there we go, proof that I still have a long way to go to fully grasp basic electronics. I made the mistake of assuming that there are no sensitive components. I'm glad I'm deaf. Literally!

Looking at other older posts regarding various surge, spikes, and voltage clamps as well as thru other sources, I'm still not certain what is the "best" solution. The entire circuit will not draw more than 3 amps and does require 12vdc. I found one schematics regarding some type of protection involving 2 10uf 25v caps, a zener diode (1n4742) and a 390 ohm resistor (didn't say what watt). It has a 12vdc battery input and a simple 12vdc output. Am I on to something? Should the resistor be 1/2 watt or?

 

[Skogsgurra]

A 390 ohms resistor will not let any 3 amps through at 12 volts. Hint: U=IxR.

And if it did - which it doesn't - the power would be a lot more than 1/2 watt. Another hint: P=RxI^2 (390 x 3^2 = 390 x 9 = 3510 watts.

Your best bet is to design with components that can take the surges that can be anticipated in your application. There are transistors that easily handle hundreds of volts. So that part shouldn't be a problem. Protecting the low-power electronics is sometime done the way you described.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...