Voltage regulators fail high

You can use a SCR crowbar which blows the fuse,or something like this.

This will a least save you on fuses if its just due to wiring errors.

The comparator senses the output voltage and compares it against the reference if it's greater the comparator pulses the base of Q1 which with Q2 makes a latch this switches off the P-FET at the 7805 regulator input shutting it down.

I didnt include any caps other then C1 for some noise imunity, I also didnt calculate resistor values but you should get the idea.

Theres allso dedicated monitoring IC'S.

You may instead want to figure out why they keep failing do you have them properly heatsinked?

Two things I forgot; put a small signal diode at the comparators output cathode towards the transistors just 1n4148 will do . The other thingis place a small capacitor across R6.If you don't have that cap there the transistors will latch as soon as you apply power to the circuit.Start out with 10pf increase as needed.

Usually most devices operating off 5V can take an absolute maximum of

7V so set your protection to activate at around 5.8 or 6V. But check the datasheets on what the devices are that the 7805 is supplying and Check the absolute max voltage they can withstand. Then set your Protection accordingly.

You could also place an led in the transistor circuit so you know that it has been tripped.

First question. What temperature are you running them at ? Do you understand thermal calculations ? What is the max die temp ? Usually it's

125C on 7805s.

Graham

I've abused the 7800 series of regulators. They have thermal protection I know it works from experience;). The only possible explanation for having ten volts at the output is hooking it up backwards, that's about all I can think of.

Yes, you put a fusible link in there with a TVS diode as the cross bar to blow it..

"

or not having the common pin connected.

"

Yep never thought of that.That would do it.

I've accidently left them into a dead short for a few minutes and they survivied even the T0-92 L series. Your eyes start to get buggy after a couple of hours looking at a breadboard;)

Are you using thee proper capacitors across the input and output pins to ground?

--
http://improve-usenet.org/index.html

aioe.org, Goggle Groups, and Web TV users must request to be white
listed, or I will not see your messages.

If you have broadband, your ISP may have a NNTP news server included in
your account: http://www.usenettools.net/ISP.htm


There are two kinds of people on this earth:
The crazy, and the insane.
The first sign of insanity is denying that you\'re crazy.

The SOT-89 version (78L05) looks similar, but has the input and output pins in reverse order to its larger cousins...

"TonyMS"

** A 5.1 volt zener ( 5 watt rating ) across the supply rail will save any parts that could be damaged.

Add a fuse of fusible resistor in series with the supply to the reg IC too.

Anecdote:

Once had to repair a suit case size box of electronics that accompanied an Italian made electronic piano accordion.

It was chock full of 7400 series ICs, all running from the same 5 volt regulator - a TO3 pack type.

The reg IC was not properly fitted to its heatsink nor was there any thermal grease used - so it eventually failed HIGH -at about 11 volts IIRC.

Took nearly all the digital ICs with it - over 120 of them, thankfully they were all in sockets.

When I completed all the repairs, I added an SCR crow bar to the 5 volt rail.

..... Phil

Many thanks for all these helpful replies, which I'll study at greater leisure.

One other suggestion that occured to me last night was simply using a high wattage resistor to drop the 12V down to say 7V so that a failure can't lift so high

Cheers

Tony MS

78xx regulators need a minimum of 2V 'headroom' so they might as well not be there if only fed 7V. If you feed therefore feed them a minimum of 8V, and they fail short, well 8V can smoke the logic nearly as well as 12V.

The one thing that hasn't been mentioned is a reverse protection diode, connected cathode to the 7805 input, anode to the output. If you have anything that dumps current into the 5V rail, e.g. multiple supplies, large reservoir caps downstream of the regulator, or badly designed motor drive circuits, reverse current through the regulator can blow it.

Personally, I'd make up a PSU delivering the voltages I commonly need for breadboarding and stop trying to run regulators on the breadboard. The only exception to that is if I'm deriving a low current clean supply for one chip using a 78L05 or similar.

I've been using the same hacked Sinclair Spectrum +2 PSU as a small bench supply for the last 15 years with no trouble. IIRC it uses a LM723 regulator which let me hack in a 1.5A current limit circuit and I modded it with a pair of 78M12 and 79M12 regulators for the +/- 12V supplies. The other mod it has is a 4 pole toggle switch on the output so the project on the bench is isolated when its off and I can solder stuff without blowing out CMOS inputs. Its not the supply for every job, but it's pretty handy for breadboard work.

If you *must* continue with the main regulator on the breadboard, make up a little module on veroboard with your heatsinked 7805, a fuse or polyfuse on the input, the decoupling caps, a reverse protection diode and a SCR crowbar circuit on a suitable header to plug into your breadboard. Use separate pins for 0V in and out even though its the same rail so you don't loose regulation if the ground lifts and you'll not over voltage any more logic.

--
Ian Malcolm.   London, ENGLAND.  (NEWSGROUP REPLY PREFERRED)
ianm[at]the[dash]malcolms[dot]freeserve[dot]co[dot]uk
[at]=@, [dash]=- & [dot]=. *Warning* HTML & >32K emails --> NUL:

I have heard of people shorting the output to ground and using them as thermostatically controlled heaters - they heat until the thermal protection temperature and then stop heating above that temperature. I think the reliability is probably not good when used like that, because the manufacturer may have set the thermal cut-out at a temperature that will in the long term damage the die, the die attach or the moulding compound, and also the electromigration lifetime due to current density in the metal tracks would be reduced at high temperature.

Still, if you need to perform that heating function without needing high reliability, then I can't think of a cheaper way to do it. It would be nice if someone would make a part that was designed and specified for that application, so that it could be used with more confidence.

Chris

There is one scenario that can kill 78xx regulators: if the output voltage is higher than the input, damage can result. That doesn't SOUND likely, but remember that power-loff events always kill the input voltage, and the output only drains slowly if there's an output filter capacitor.

The recommendation is to add a diode, cathode to (input) and anode to (output) when using large filters on the output.

That is totally insane!

"

Yes, I think this must have been my problem. I had a too large cap on the 5V.

Still doesn't explain why it failed high though

Thanks

Tony MS

"TonyMS"

** Course it does !!

You shorted the input voltage to ground.

BJTs fail short C to E.

Dick wad.

.... Phil

Here are some pictures of the heating plate than one guy mounted to his diesel injection pump, so that vegetable oil would be warm enough to lower the viscosity closer to that of diesel.

I count 10 regulator chips. He mentions that as well as all of the tantalums, he needed an additional large capacitor otherwise he says it oscillates and the tantalums blow up. I suspect they probably would sooner or later anyway.

Chris

And I know from experience that if you abuse the thermal protection, they can explode.

Sounds to me like you're abusing the part through ignorance. How about answering my question too ?

Graham