series resistance for relay control

I am using a SPDT relay with a diode to enable/disable the vehicle.

I am controlling the relay from a TTL transistor switch rated for 30V. My relay is normally closed, so there is a constant 12V on the switch. After a few days, my switch is failing - always on - relay triggered - car disabled. I swap out the transistor and it works again.

Would adding a series resistance before the switch help? If so, what resistance?

Thanks!
Cheryl

I seriously doubt the relay is harming the switch assuming you're feeding 85 or 86 from the relay, coil draw will be about 1/2 an amp.

-------------
Amateurs assume, don't test and have problems; pros test first. I am not a free install service.
Read the installation manual, do a search here or online for your vehicle wiring before posting.

Do you have the diode installed correctly? What is the part number for the transistor you are using? Just because it is rated at 30vdc doesn't meant it can handle the current.

-------------
Kevin Pierson

The diode is already attached to the Hella relay.

The transistor is a MOSFET N-channel rated at 60V/16A.

The relay control is connected directly to the drain of the TTL MOSFET on my PCB. I control the relay by applying voltage to its gate.

I'm trying to use a PCB without modifying it. I would have designed in a series resistor to take the load when the relay was "on".

The relay coil is a resistor that is already limiting the current through the mosfet. That is a very large transistor, there should be no load concerns at all. I run relays in commercial products through 200mA NPN transistors (with diodes across coils) for years and years with no issues.

Are you running a standard automotive relay (Bosch style)? They should pull about 160mA.

-------------
Kevin Pierson

Yes, I am using an automotive relay.

When I have the transistor off, my relay control signal stays at 12V. Since the transistor is rated at 60V, I assume it can live quite happily with 12V across it (no current).

When I turn the transistor on, I am drawing just over 100mA. However, there is still 3-4 volts across my transistor. I would think I would want to see no voltage across a transistor when it is on. So I was thinking that I am burning out the transistor by trying to dissipate power (100mA over 3-4V) through it.

So I thought I should add a resistor in series - before or after the transistor - to dissipate the power properly. If that makes sense, then what value resistor? Too low would still hurt my transistor. Too high might not provide enough current through the coil. ?

Thanks so much for your help!
Cheryl

The TTL (5V) does not have enough voltage to turn the MOSFET fully on, hence the heat.

Better off using a Darlington transistor, else a lower Vgs FET.

Why did I say 1/2 an amp? It's nearer 200mA.
We've discussed this before, this is the sort of thing I'm glad Oldie and K.P. have answered, I was right in fact but the theory side leaves me in the dust!

-------------
Amateurs assume, don't test and have problems; pros test first. I am not a free install service.
Read the installation manual, do a search here or online for your vehicle wiring before posting.

is the transistor getting hot to the touch (hot enough to burn you)? you may consider addind a heat sink if heat is a concern.

depending on the drive circuit you may be able to swap in an npn transistor and get away from the mosfet.

adding resistance in series will limit the current to the point the relay will no longer turn on. since you can not fully saturate your transistor you are most likely operating at the low end of the voltage range (hence the lower then normal current draw from an auto otive relay).

how is the drive circuit set up? what package is your mosfet? what is the part number of the mosfet?

-------------
Kevin Pierson

I'm using the Infineon MMBT3904LT1XT npn trasistor. (Sorry, my schematic says MOSFET.)

I'm not too worried about heat, since this relay should only trigger (or not trigger) when the starter wire is hot. So it is more a concern of burst currents. However, the transistor is hot, not quite enough to burn me.

I've blown two of these transistors after having them installed on a vehicle for less than a week. With no series load, I am seeing about 3V across the transistor with about 120 mA current. With 39ohm series, I see 0.6V across the transistor with about 100mA current.

Adding more resistance would lower the current through the relay coil. Is there a minimum current draw through the relay coil for it to work?

Thanks again!
Cheryl

Probably a formula but for an automotive relay with a coil of abut 80 ohms, in practical terms it's about 110mA.
Required voltage around 7.5 volts.

-------------
Amateurs assume, don't test and have problems; pros test first. I am not a free install service.
Read the installation manual, do a search here or online for your vehicle wiring before posting.

That transitor is only rated for 100mA.

Use a bigger or Darlington transistor. I'd suggest with at least 500mA or 1A (Ic = Collector) current.

The part number you posted is rated at 200mA. Is that correct? Above when you said it was a Mosfet you said it was rated at 16A.

A 200mA transistor should be enough to drive an automotive relay, especially in a low duty cycle application. I would add a second diode across the coil of the relay - the inductive spike created by the collapsing of the magnetic field in the relay when power drops off is far more destructive to a transistor then pulling the current you are pulling through it. Perhaps the diode is either defective or not connected across the coil the way it needs to be.

Again, I would also look at the drive circuit of the transistor. The base is most likely driven by a 5vdc signal - is there a resistor on the base? A base resistor will control the ability of the transistor to enter saturation and get rid of the voltage drop across the transistor that you are seeing.

If you want to add a resistor in line, which most likely won't help you, you will have to determine the minimal amount of current needed to turn the relay on. One way to do this is with a 12vdc power supply and a potentiometer - start with the potentiometer set to the highest resistance possible and slowly turn it down until the relay turns on. Measure the resistance of the potentiometer and start there. You will need to make sure you wattage rating is sufficient as well.

-------------
Kevin Pierson

I will replace my transistor with a Darlington. It needs to be the same footprint. (3 lead TO-252 D-PAK) Any suggestions?

Currently I have a 1K resistor in series to the transistor base which is driven by a uC running at 3.3V. Since the base will stay at 0.7 V, that means my Base current is 2.6mA. I don't seem to have any problems with my relay triggering, so that seems to be OK for the transistor I am currently using. Would I need to adjust it for a Darlington transistor?

(My schematic/BOM calls out a MOSFET n channel 60V, 16A TO-252 D-PAK part, but what I have on my board is the part noted above. Sorry for that confusion, I was not the original designer.)

I tested this, I had 12V on the relay input and toggled my switch every 3 seconds for 24 hours. No problems. But what I hear Kevin saying is that the relay input (the starter signal) going on and off is what is probably killing my transistor. Correct? So I should stress test my design by flipping the starter off and on. (The switch is either always on or always off.) Does that make sense?

Would the starter signal be floating or grounded when the key is out or in the ignition state?

Thanks SO much everyone!
Cheryl

The base current should be the output (Ic) divided by the gain - eg, design for 250mA (12V/.25 => a 60 Ohm relay coil as typical for many automotive relays) and if gain is 800, then 250/800 = .31mA.
That may mean 2.6mA/.31ma = 8.4x you present base resistor value depending on you base-ground resistor value (if fitted).    

But overdriving the base (current) should not matter unless its max base current spec is exceeded. However others are better to advise - I've forgotten my transistor design stuff.


The starter should be GND when not cranking as the starter solenoid else its relay's coil provides a path to GND.

A reverse biased diode across the relay coil (or collector to emitter) might be needed to prevent relay spikes damaging the transistor.


I don't know about equivalent footprints though they should exist. (The data I saw was for your transistor a dual-pak SMD footprint. But whether SMD or leaded, ensure the correct order for b, s & e (even if it means bending leads).)

I've tried a Darlington transistor to switch off the starter as was suggested.

Here is the sequence I am seeing...

1. Switch is off (gnd applied to transistor base)
2. Relay allows starter to start vehicle
3. Switch is turned on
4. Vehicle starts on first attempt.
5. Vehicle cannot start on all subsequent attempts.

#4 is what is unexpected and undesirable.

Could the Darlington transistor be causing an extra cycle of 12 Volts on the switch before it allows the relay to switch?

Thanks again,
Cheryl

What vehicle?
Wouldn't be a Prius by any chance?

-------------
Amateurs assume, don't test and have problems; pros test first. I am not a free install service.
Read the installation manual, do a search here or online for your vehicle wiring before posting.

No.

Actually, I don't know what vehicle my guy is testing it in.

Should it matter?

Hybrids such as the Prius have a different "ignition" cycle and this may give the result you just had.

-------------
Amateurs assume, don't test and have problems; pros test first. I am not a free install service.
Read the installation manual, do a search here or online for your vehicle wiring before posting.

Can you explain more on how this would happen?

My test car is a 2003 Acura TL Type S

Just wanted to know if it was a hybrid. On the Prius, the cut wire is constant, doesn't apply to normal gas/diesel cars.
Could you show the circuit? K.P. or Oldspark might be able to comment on what's happening.


-------------
Amateurs assume, don't test and have problems; pros test first. I am not a free install service.
Read the installation manual, do a search here or online for your vehicle wiring before posting.

A Darlington transistor is just two transistors in series with each other. The benefit with this setup is your Beta is much higher, allowing for higher amplification. Since they are two discreet components wired in series as one circuit as long as the components are spec'd correctly the entire thing should behave like a normal transistor, but with a much higher capability.

It would be beneficial to know more about the overall project and who is designing / making the parts being used. A simply switch controlled transistor should be able to instantly turn on an interrupt relay with no delays at all. What is the part number of the Darlington you used?

-------------
Kevin Pierson