Trying to figure out why my cicuit blows up transistors

[iso9001]

Hi,

I have a pic driving the base of an NPN through a 1k resistor.

One end of a relay coil going to 12V and the other end going to the collector of the NPN. These are diode relays, but I'll get to that.

The emiter of course goes to ground.

Anyway, I turn the pic output high and the relay siwtches.

Now, my problem is that someone hooked the relay up in reverse so that the diode inside was going the wrong way. Now, what I "think" was happening was that when I triggered the relay, 12V went down through the diode and not the coil, into my transistor where it was connected with ground. Shorted.

I wasn't around but I beleive the relay never fired, so it was not a EMF problem.

Would that blow an NPN?

Obvious choice would be to put a resistor inline on the collector to prevent this from happening again, but I'm not sure about what size, too small and it will just blow the resistor (i hate fuses) and too big will stop the reay from working no ?

Thanks

 

[audioguru]

Hi Iso,
If the transistor has an hFE of 200, it tried to amplify the 4.3mA of base current to 860mA. Its absolute maximum current rating is probably only 100mA, any more will burn it out. It probably couldn't pass 860mA so it fried instead. The reversed diode was forward-biased and became a dead-short load to the transistor.

How to prevent it from happening again?
1) Fire the person who hooked it up backwards. Was it you? He, he :lol:
2) Use a poly-switch-fuse-thingy and a huge transistor. When the poly-switch heats from too much current, its resistance rises to reduce the current. The transistor must withstand excessive current until the poly-switch heats-up. The poly-switch returns to normal when it cools.

 

[iso9001]

Just what I was thinking.

Never seen a poly-switch-fuse-thingy, I'll look that up.

I'm using a pretty big tranistor as it is. Fairchild BCP54, 1.5Amp Continous, 1.5W Disipation,

I have a sim that shows that with a 40ohm resistor the transistor survives and the relay still clicks, the sim is pretty simple and I'm not sure how much thier resistor is going to disipate compared to my 1206 chip resistor.

It would be very very easy just to put a resistor around ~40 in, I'm just not sure if it should be higher

 

[audioguru]

Hi Iso,
You must have a huge relay to be using such a big powerful transistor.
Don't you understand that with the diode wired backwards it is a short across the relay coil? The relay won't do anything with only the diode's forward voltage of only 0.7V across the coil.
A simple calculation will show that a 40 ohm resistor in series with the circuit will dissipate about 3W when the diode is backwards. Another simple calculation will show the voltage-divider action of the 40 ohm resistor with the relay coil resistance and show if the resulting reduced voltage across the relay coil will stop it from activating.

 

[iso9001]

I get the reversed diode part. Its not a big relay at all, general 30amp, coil has about 75ohm on it, 12V feed, nothing massive by any means. BUT.... I want it to last FOREVER (you know what I mean)

What I'm wondering is, say this were to happen again, but maybe not the diode being reveresed, maybe the coil fails, etc. Will a small resistor help me at all as far as protection ?

I dont really get how the voltage divider is working there since the relay is 75ohm, if I put a 2k just infront of the resistor I get 11.6V in between. But obviously 2k wont allow enough current to trigger the relay.

Problem is I'm not sure how much voltage the relay needs to activate.

 

[audioguru]

He, he. Your 12V relay has 30A contacts and a 75 ohm coil. My 12V relay has 1A contacts and a 2350 ohm coil. Big difference.

Most relays work with no less than 75% of their coil's voltage rating which is 9V. 9V across the 75 ohm coil is 120mA. Your transistor might conduct with 0.5V across it so the difference from 12V is 2.5V. 2.5V divided by 120mA is 20.8 ohms. Therefore the relay will barely work with a 20 ohm resistor in series with it.
If the diode is backwards it will have about 0.7V across it when the transistor is turned on. Therefore the 20 ohm resistor will have about 10.8V across it and dissipate 5.8W. If the transistor conducts well enough to have only 0.5V across it, it will disipate 0.27W.

 

[Sebi]

A safe solution..., but more components.

 

[eblc1388]

To prevent the relay being connected wrongly in the future, you can either:

1. put in series a rectifier diode with one of the relay coil connection wires, or

2. fit the back emf protection diode permantly to the collector of the NPN transistor and positive rail and then use a relay without diode across its coil.

In addition, very few circuit design would take care of the fact that the relay coil becomes short-circuited as this seldom happens. More likely is the coil going open-circuited.

Regards

L.Chung

 

[iso9001]

I get it.

There is no way my 1206 20ohm resistor is going to disipate 5.8W :shock:

And actually this was a design one revision down, the new ones all have the diode attached right to the collector so I don't "need" those fancy pants relays anymore. It just happened to be I was testing an old board.

Thanks Sebi, although that is more parts then I have space for.

I know it probably isn't likely I'll have this problem again. I just wanted to check and see if there was some trick protection solution.

The only relay I've seen go bad did stick closed, cause it kept blowing fuses. Oddly enough the transistor was fine then. Hmmm ? Again, now that I've id'ed this problem I am destined to never face it again. :?