CMOS 555 BURNS???

[Kain]

Here is a circuit that I tryed yesterday. I first tryed with NE555 which is the regular 555 timer but then I wanted to have less power consumed so I decided to try with TLC555 which is same timer but CMOS (LINMOS) version from Texas Instruments. While the NE555 seems to work ok, the TLC555 burned out right away

[audioguru]

Hi Kain,
I suspect that the supply voltage is too high because with only a 12V supply, the LED would look dim with only 1mA or less. Did you exceed the absolute max supply voltage of only 15V for National's or 18V for TI's Cmos IC?
Is a relay or something connected to its output?

D1 is used to keep the negative end of the cap from going too far below ground (negative) when the cap is charged and the power is turned off. It might not be good enough since the absolute max voltage is only -0.3V.

Maybe the supply voltage is coming up too quickly causing the IC to behave like an SCR. Most circuits have a supply bypass capacitor which would slow down the supply voltage rise.
Do you dare trying a new IC with a 100uF supply bypass cap?

[Guest Alun]

What are you trying to do?

Because there are probably far easier ways to do it, I think a re-design might be in order here.

I have a feeling the IC is been overvoltaged in some way, perhaps a pin gets taken above or below the supply for a period. I found that CMOS ICs in general seem to be more sensite to abuse in general than their bipolar counterparts. Even though the CMOS 555 is rated to 18V it might blow at 20V but even the TTL 555 is rated to 16V it'll probably take 20V (even more) for short periods. CMOS ICs are especially sensitive to their inputs been taken out side the supply as they contain a parasitic SCR that'll short out all the pins and will melt down if a pin is taken above or below the supply by as little as 0.5V

[Kain]

Alun:  "What are you trying to do?"

I am trying to use 555 as a delay-on timer. In this particular circuit when there is nothing applied to the base of the transistor it shorts out the capacitor which causes pins 2 and 6 to have positive with respect to ground - the output on 3 is low. Once we connect the base of the transistor to ground the capacitor is allowed to charge exponentially, thus eventually we get high on 3, untill we remove the ground from the transistor's base, in which case the transistor quickly discharges the capacitor and the output 3 is forced to low.


Audioguru: "I suspect that the supply voltage is too high because with only a 12V supply, the LED would look dim with only 1mA or less. Did you exceed the absolute max supply voltage of only 15V for National's or 18V for TI's Cmos IC?
Is a relay or something connected to its output?"

The supply voltage was 12V when I tested it and current limited to 15mA, so no - I didn't overvolted the chip. For now it drives an LED which lights up bright at 1mA even so I haven't tryed to source even 10mA , thus I haven't exceed the secifications in this way at least. :D Also if the diode is trying to remove any excess negative voltage from the capacitor I am not sure if it can do lower than about -0.7V (Si) thus if this is enough to burn the chip .....oups. Just a guess though. I haven't tryed to bypass the chip and see if it burns again but I have the feeling that I am about to waste another chip by doing it. The bypassing is normally suggested for the bipolar version of 555 but the manufactors claim it's not necessary for CMOS types. I hope I won't burn another ... ::)

[Guest Alun]

Try this, I haven't tested it, it's just an idea.

The 1nf capacitor charges up when you turn on the power and triggers the timer/ Don't put a capacitor to pin 5 as it probably won't work because the timer won't trigger when the control voltage is 0V, this'll also happen when the power is turned on so it'd ignore the trigger pulse.

I hope this works.

[audioguru]

1) Every single one of the applications circuits in National's datasheet for their LMC555 Cmos IC has a 0.1uF supply bypass capacitor.
2) Cmos ICs are known to "latch-up" across the supply like an SCR under certain conditions.
3) SCRs have a dv/dt limit of the rate of rise of their supply voltage so they don't turn-on when they shouldn't.

All of the above shows that a supply bypass capacitor is needed in Cmos circuits.

[Kain]

Ok now,  I decided to try this once again before I totally change my mind about the circuit design and so far is seems like Aidioguru hit the jackpot  ;D. I tryed the exact same circuit with a bypassing capacitor of 100nF across 1 and 8 as close as I can hook it and another capacitor of 100uF electrolic across the same  pins. I have triggered this thingy about 50 times so far trying to burn it hard but it refuses  ;D. Maybe 20V would do  :P Just kidding but yes - it seems that the bypassing capacitor is needed, indeed. Oh, and about my previous post - I checked again about the bypassing of CMOS versions of 555 and it seems like I misread it sometimes ago - they still need to be bypassed but the value of the bypassing capacitor can be much smaller. I guess it was one of those nights when I was pulling all-nighter and didn't quite know what I was reading when I did  ;D

[audioguru]

Hi Kain,
I'm glad to have helped you fix it. ;D
Without a supply bypass cap, the supply voltage came up so fast that the TLC555 became a turned-on SCR across the supply.

A regular 555 needs a supply bypass cap because its high-current 200mA+ output transistors are both conducting when they switch, causing a 400mA supply current spike.
The Cmos 555 is rated to source 10mA (pullup) and sink 50mA (pull-down) near its max supply voltage and lower output current with lower supply voltages.