Smoking Controller

[JohnC]

Looking for some help trouble shooting this darn diversion controller that let out some of that magic smoke.

This controller is only for a fan when the voltage rises above 14 Volts

I accidently shot 26 volts in this controller I'm playing with to set up, and poof... magic smoke!

I tested the transistors and diode, which seem to be OK

I did blow one resistor and I replaced that, but the controller still was, a no go.

Not sure how to test the LM741, I'm also thinking it could be the 1K POT.

Are there any gurus out there that would know which pins I could test out.

My eletronics expertise is well below average. It's  getting bad, gotto start hiding the sales slips from the wife, just too much magic smoke.

Should the pic not show up it may be viewed in my file in "user info"

Help

Picture in bmp LINK
John Cox

Ottawa, ON

[dinges]

Magic smoke, you say? can't you put it back in?

Saw the schematic, and there's not much that can be wrong with it; you say both transistors are ok?

My bet is the 741; don't know its maximum rating, but I'm guessing at 18V (can't be bothered to look it up in the datasheet; I'm pretty certain it's toast).

Your resistors shouldn't be defect, they can stand the overvoltage.

A quick test to see whether the 741 is defect: turn your potentiometer from fully left to fully right; while turning it, the output of the 741 should switch from hi to lo (or low to high, depending on direction). If it doesn't, your 741 is in silicon heaven (no, not California).

There's a very small chance that the zener diode is also dead; you can easily test it with a meter, there should be 6V (I think it was a 6V zener) across it. Don't think there's a problem with it, but better make sure.

As far as hiding sales slips: 741s are pretty cheap (have been around since the early '70s, BTW! They were much more expensive back then :-)), not more than 1$, I think. Hope you've used IC-sockets; if not, easiest solution would be to cut off, with a little side-cutter, all the 8 individual pins; removing the black body; and then desoldering the 8 pins individually. That way you run the least risk of damaging your printed circuit. And note where the 'notch' is on the 741, so you can make sure you mount the new one correctly; otherwise, that one too will let its smoke out.

Good luck,

Peter,

The Netherlands.

[dinges]

Noticed just now you did blow a resistor?! Which one? If R3, I'd replace the diode too, no matter what the meter says. But I can't imagine any of your most left three resistors blew?! Strange.

Peter,

The Netherlands.

[dinges]

And I did have a quick look at the datasheet; the 741 should be able to handle 36V supply voltage...

Another check: turn the potentiometer from left to right, while measuring the voltage on the wiper; it should vary from 4.2 to 5.8V; if it doesn't, this means that either R1, R2 or R4 (the potmeter) are gone.

So, check the voltage over the zener: it should be 6V; check the varying voltage on the wiper; it should vary from 4.2 to 5.8V; finally, check the output from the 741; it should go from hi to lo (or reverse) when turning the potentiometer; there should be a very marked 'switchpoint'; turn a little to the left and it goes fully +12; turn it fully right and it goes 0V (or reverse). Check this out and when you know more, let us know.

Peter,

The Netherlands.

[Opera House]

You're a wizard who ever you are.  I can't see how you can burn up any resistor in that circuit.  There should be a resistor to limit current from the op amp to the transistor.  Is a 2N3055 really used as a driver to the other 3055?   The op amp is used as a comparator.  If pin 3 voltage (as referenced to the - supply) is higher than pin 2, the output will go high and turn on the transistor.  While on the surface this circuit basically works, it is a poor design.  The op amp will never go fully to the positive rail and the darlington transistor will cost another couple of volts.  That means a lot of heat.  I would (at a minimum)put the transistors below the motor.  The circuit might have died because of spikes from the motor.  Better put a diode across it.  Simple test:  drive transistor base (op amp output disconnected) from the + rail with a 220 ohm resistor and see if the fan comes on.

[maker of toys]

the darlington pair is because the 741 can't supply the drive current to turn a 2n3055 all the way on all by itself.  I'd have used something more like a 2n2219 for the driver or maybe used FETs, but, hey- maybe the designer got a deal on the '55s?

to test a 741:  short pins 6 and 2 together (tiny blobs of solder holding a peice of resistor lead will do) and disconnect the transistor bases, then turn on the power.  measure the voltage on pins 6 and 3.  if the readings at pins 3 and 6 (relative to the circuit common (-12 V here)) are not VERY VERY close (+/-0.002v) then the 741 is suspect.

at $0.35 for a single 741 (mail order) you'd be time and agony ahead to just replace it. might as well get a dozen and put in a socket while you're at it;  the 741 is tough, but that one is going to have a hard life.  (they are very static and EMI sensitive compared to everything else in that circuit.)  A LM 311 comparitor is (I think- ) a pin compatible replacement for the 741 in this application, and is likely to be a bit hardier. (if the 311 isn't, there IS a suitable comparitor out there, I'm just not remembering it at the moment.)

also, as someone already said, if R3 is gone, it's a good bet that the 741 and the zener are toast too.

and as another someone said, those transistors will be happier with the motor on the collector instead of the emitter (where it is now.)

A big 'free-wheeling' diode (a 1n4004 or whatever radioshack has in a  1 or 3 amp, 400V rectifier will do) isn't a bad idea either;  put it across the motor leads by preference.  (you do not want the diode to conduct in normal service. . . .) a 0.1uF ceramic cap across the motor  (with another between pins 4 and 7 of the 741) will keep wierdness from infecting the 741 so often, too.

[Opera House]

Another potential problem is that this circuit is prone to oscilation.  The moment that the fan turns on, the supply will drop a couple millivolts and then the fan will turn off.  This isn't as big of a problem if the load is just a resistor. The motor being inductive will give an amount of kick back.   Even as large a resistor as 1M between op amp pin 3 and 6 will provide some positive feedback.  Think it should be closer to 300K.

[JohnC]

Thanks guys for your expertise.  The culprit was the 741.

I don't know if I want to continue with this circuit after everyone's feedback.

I just was trying to turn on a fan at my battery box when the voltage reaches somewhere in the vicinity of 14 volts to vent a battery box.

I would like to keep it simple.

Any suggestions would be welcome.

Cheers

JohnC

[Dzisko]

See V-drived_fan.gif in my pictures. I love simple good circuits too.

R1 can be 100-500 Ohm, R2 - 5-20 kOhm, D1...Dn on your choise to reach requered level of voltage when fan begins to run, Dpr - any middle-power contemporain diode, Q1 - almost any middle-power, and Q2 - power npn BJT.

[JohnC]

Thx Dzisko

I appreciate the tips. Seems like a simple circuit I may try.

Could I use my 3055's for Q1 and Q2.

Also you suggest "D1...Dn on your choice to reach required level of voltage when fan begins to run, Dpr - any middle-power contemporain diode", could you elaborate a bit better for me what I should use, just in the learning curve on this stuff...but having fun.

Thx guys for your input!!

[Dzisko]

Of course, 3055 will work well. If I am not mistaking, it allows to 90 V and to 10 A.

All depends of the your fan power. If its power is 10 W, them there will flow current about 1A and if Ku of the transistor 50, it will need 20 mA through its base - if you have Zener diode powerful (allowing such currents) enough, you may leave just Q2, and if your fan take less than 10 W, then, of course, it is not necessary build the Darlington circuit (plug all your diodes and R2 directly to base of Q2).

How to chose number of diodes - take R2 big enough (20-50 kOhm), and then R1 also big enouht - 1...2 kOhms (to not burn diodes and bases of transistors, but if you need the sharpnest, then reduce R1 (R2 does not play a big role - things were changed since times of Germanium transistors :-), and then replace diode by diode, keeping in mind that they have to be in sum about 12 V (1.4 V will fall on bases (if you use 2 transistors asa a Darlington circuit)).

Voltage dropping on the Zener diode is typed on it as a rule, and voltage droppind on diodes is about 0.7V on each (of course, diodes plugged in direct, and Zener diodes in reverse connection).

Interesting property of this circuit is that diodes have negative thermoresistance - the more temperature, the less direct voltage - so you may obtain thermo-drived fan.

I think, that is much more useful thing, than V-drived. The more diodes in serial - the more sensivity, so there are lots of toys to play with :-)