Power transistor overheating

I am building an electric fence using a 12 volt battery to energise a NE555 timer to drive a TIP41C transistor into a car ignition coil . Coil resistance is 1.7 ohms as measured. Drive pulses to TIP41 are about 30ms duration at one pulse per second. Problem is that TIP41 gets hot enough to destroy itself after about 30 pulses . Difficult to measure current as spark upsets multimeter. Using a lamp as load I measure 1.6 amps and TIP41 still gets hot enough to burn. What am I doing wrong !!! Should anyone volunteer to help me I can email the circuit which I got from the net but with no contact address.

A 30 ms pulse is way too long. The coil has reached a steady state maximum current, long before that. Think of a coil driving the sparks for an 8 cylinder engine turning 3000 RPM. That is 3000/2 * 8 sparks per minute or 200 per second. Even if the charge time was the whole time between sparks (and it is actually more like half that), that would mean that the coil had 12 volts applied about 1/200 s = 5 ms. And it still makes sparks just fine.

The second thing to be sure of is that the transistor has enough base drive current that it is well saturated at 12 volts / 1.7 ohms = 7.1 collector amperes. That may be hard to accomplish with a TIP41.

According to its data sheet:

the gain falls off dramatically for collector current above 3 or 4 amperes, especially if you expect it to stay in saturation.

Another potential (pardon the pun) problem, especially if you are using a TIP41 (not suffix B or C) is that it may stand only a little more 40 volts before breaking down right after turn off, when the coil puts a large voltage stress on it. This puts a lot of power into the die.

You may email me the schematic, if you wish, and I will critique it and make suggestions for better performance.

The TIP41 is a power transistor. It can handle the volts and current but it needs the help of a *heatsink* to get the resulting heat away.

Used in free air with no heatsink it'll overheat if much over 3 watts is dissipated in it typically.

If you don't understand what I'm talking about you need to go learn some.

Graham

I'm with John Popelish on this one. The 555 isn't providing enough base drive. I'd try to find a darlington that's rated for a much higher voltage, and put a transzorb that's rated at the transistor's BVCEO across it, to snub the spike at whatever the breakdown voltage of the transistor is.

Or, you could do it completely different - lose the transistor, and buy or build a little DC-DC converter to give you a few hundred volts, charge a cap with it, and periodically fire an SCR to discharge the cap through the coil.

Good Luck! Rich

Use a BU941ZT on a nice heatsink. It's a darlington made for driving ignition coils. The datasheet is here:

Ed

If you have the on/off the wrong way around,it gets very hot, else if the transistor is not saturated(enough base current), it dissipates more ,and last ,30 msec might be a bit long,and if so, the high voltage might become dangerous.(And drain your battery faster then you would like).So make the pulse time as short as possible,for the output voltage to shock in stead of kill.It also saves the battery.

All excellent points. With the transistor I mentioned, he'll have 1.8 V C-E (from the datasheet) at 1.6 amps (from his measurement), for 2.88 watts. He needs ~100 mA base drive (from the datasheet), and the 555 can source or sink 200 mA (from the datasheet), so the BU941ZT will saturate easily. It can handle 150 watts (from the datasheet), so with a nice heatsink 2.88 watts won't be a problem. So it should be able to handle the op's requirements, easily. I wouldn't want to run the base drive the wrong way round either! However, more to the point, the 30 mS figure is not necessarily the best choice. If it drives the ignition coil into saturation or if the voltage produces is too high that needs to be reduced. I think 30 mS is too long - I'd think 1 mS would be a good target figure.

Ed

Grumble. Left off a sentence: He can work up from there, increasing the pulse length until he gets the desired voltage, without overheating or saturating the coil.

Ed

Gentlemen, Thankyou for your comments and suggestions. I have tried some with modest success. However I have triumphed. My circuit is now fully functional and scaring the hell out of the possums which attack my vegetable garden. The problem was that the TIP41 was conducting for the 970 ms and off for 30ms. I did not know how to find this out. The LED flashed, that is all. I invested in a PC sound card oscilloscope and while difficult to interpret because of the AC coupling, it enabled me to see the pulses and set the pulse length. It now conducts for 2ms and rests for 998. It does not overheat, in fact it does not even heat at all!!

2ms is the optimum, shorter weakens the spark, longer does not improve it. Now to more exciting projects for which I will no doubt again solicit your advice !!!