LM7812 problem

[Wobulator]

I've got an LM7812 conected to a 24v battery supply with a load of 63mA max. The device has failed on me twice dispite trying heatsinks (which shouldn't be necessary at such a low load).It's used on a relay circuit board for dc motor starting, so I suspect it might be a surge problem but I can't see any spikes with my osciloscope. Anybody got any bright ideas?

 

[OperaHouse]

Are yoy driving relays with it? Only thing that comes to mind is if the kick back of a coil causes the output pin voltage to go higher than the input pin.

 

[RadarRay]

Its been a long time since I did any powersupply design, so I am doing this with what is left of my memory of it. Did you follow the app guidelines? National Semi used to have very good ones. If I remember there was a small cap on the output that was required. Other than that I agree with Operahouse. Are your coils snubbed?

 

[jbartos]

Suggestion: Visit

http://www.wb4hfn.com/DrakeArticles/Receiver_Pwr.htm

for a little filter on the LM7812 output

 

[Nandos]

Wobulator:

7812 need an output electrolytic capacitor as well that an input filtering capacitor.
The relay kick back, needs to be eliminated, place a diode ( IN4002) from the output to the input.

Anode to output, cathode ( the band ) to the input of the 7812.
The 7812 is sensitive to reverse voltages above 2 volts.

Regards

Nando

 

[BrianR]

The failure is probably due to high voltage transients generated when the solonoid coil is de-energised. The energy stored in the inductor's field is returned very quickly when power is removed and a high voltage reverse pulse is generated, (L di/dt). A pulse several hundred volts amplitude is not uncommon).

A capacitor across the output of the IC will not prevent this and may make it worse as it forms a parallel resonant circuit with the relay inductance causing ringing.

Place a reverse biased diode such as a 1N4004 across the relay coil, (located at the relay coil if this is practical), to absorb the reverse voltage spike, and now a capacitor across the output of the supply (eg 1uF 35V tantalum) would be of use to ensure the LM7812 is stable all the time.

The reason for locating the diode at the coil is one of good EMI/EMC control as absorbs the high voltage transient which also has a high frequency component, before it can be carried and radiated from the connecting wires.

 

[gerryogrady]

I have just constructed a circuit using a 555 timer (PWM)and two power mosfets as a speed control unit for a 24VDC 20amp motor. The 555 timer is being powered by an LM 7812 and it kept shorting input to output at very low current draws < 100 ma. I garnered some advice from several newsgroups and might have the fix ( at least for my situation). A Shottky diode reverse biased across the motor, a 1N 4005 reverse biased across the LM7812 and a 3300 ufd 50v cap from input lead of 7812 to ground. Possibly spikes were getting through to my 7812 and these changes appeared to fix my problems.
Cheers ... Gerry

 

[BrianR]

As gerryogrady highlighted, there is another common failure mode for 3-terminal regulator IS's. When powered from a transformer/rectifier with a large filter capacitor, common mode mains bourne transients capacitively couple through the transformer unattenuated. The capacitor momentarily fails short circuit and dumps it's charge. (sputters)

It self-heals and recovers next half cycle and continues on as normal, trouble is that for 1/2 a cycle it's output was zero while the output of the regulator IC was held high by it's output capacitor and the IC was momentarily reverse biased.

Many regulator IC's fail with only a small reverse bias and so without any apparent cause, it fails. The fix is to install a reverse biased diode from output to input of the IC to drag the output down if the capacitor sputters.