555 timer headaches!

[MrAl]

Hello there,

I hooked up a 555 timer in a monostable configuration with a long
time delay. I looked on the data sheet and i found not only the circuit
connections diagram for doing a monostable but also a nice graph
for choosing the approximate R and C values for a long delay.
It turns out that 1M for the resistor and 22uf for the cap would be
just great. So, i figured, ok, hook this up and connect to the
trigger and there you have it...a one shot with a long time delay.
So, i hooked it up and using a pullup resistor on the trigger when
i grounded the trigger (negative going edge to trigger the device)
the LED i had connected to the output (through a resistor) went
out. Some 10 seconds later it came back on...nice! This meant
the cap charged up and triggered the upper threashold and everything
was just fine. But wait..............................
I went to repeat this seemingly simple experiment and guess what?
The next negative going edge (grounded the trigger for a second time)
did absolutely nothing! The LED stayed lit unlike the first time when
it went out for some 10 seconds.
Disconnecting the power, then reconnecting a bit later (after consulting
the data sheet again) i grounded the trigger and it worked ok again.
The LED went out for 10 seconds and then came back on. But....
when i went to trigger it again, same thing happened...the LED stayed
lit again, meaning the device wasnt working the second time around,
only the first!
Connecting a voltmeter to the cap (+) terminal, i decided to track
the cap voltage to see what was going on. After turning the power off
and back on, the cap was low (near zero volts). Triggering the input
with a negative edge the LED went out and the cap voltage started to
rise...up, up, up. This was normal and very good.
After the voltage got up a bit, the LED came back on. Great!
Funny thing is, the cap voltage continued to rise...not good!
According to the data sheet, once the cap gets up to something like
2/3 Vcc the cap is supposed to be discharged again to get ready for
the next possible trigger input. But, this wasnt happening. The cap
continued to charge until it got to some rather high voltage near the
supply voltage Vcc level. This was not the way it was supposed to work.
Playing with the circut a little bit more, i noticed that when the
connection between pin 6 and 7 was broken and then reconnected,
the cap would indeed discharge, but only after doing that, which of
course is unacceptable as the circuit is supposed to do this by itself
and that is part of what makes it work as a mono in the first place.
So i again consulted the data sheet, which assured me that
"no additional protection for the current through the discharge pin
(which happens to be pin 7) is needed". This is the pin that is
'supposed' to discharge the cap.
Ok, so what the heck was going on?
Everything worked just fine except the cap wouldnt discharge.
Being the cautious person that i am with these circuits, i connected
a 270 ohm resistor (happend to have one laying right there) between
pins 6 and 7, meaning now the chip would discharge the cap through
a 270 ohm resistor and through it's internal transistor instead of
just it's internal transistor...this limited the current that could go
through this transistor.
Now repeating the experiment, the LED would go out, then 10 seconds
later come back on, and also THE CAP VOLTAGE DISCHARGED!
Triggering the device a second time, the LED went off and the cap
voltage discharged and everything worked just fine after that.
I could trigger the device time after time and it would always work
as expected.

CONCLUSION
The only conclusion i can draw from this is that the tiny internal transistor
gets overloaded and stops conducting for some reason, so the cap
wont discharge (and the circuit wont work normally) unless there is
a resistor connected to pin 7 (the discharge pin) to limit current to
stop it from latching up. I didnt try different values to see what
would work (like say 100 ohms) though, so it's possible a lower
value would work too.

So, my question is, did anyone else ever hear of this happening with
a 555 timer or could i possibly have a bad part?
The device marking is the NE part number and it's a brand new part.

[Chris Smith]

For ALL of my 555 experiments,..... I start with a bag of fresh 555s because some do fail, then I added in Forrest’s Mimms book on the 555s to the mix, and Always used a trusty POT or two of reputable value to the mix to “Testâ€

[MrAl]

Hi Chris,

Funny thing is, it works perfectly with the 270 ohm resistor in place.
I was wondering if anyone else has ever seen this happen.

Thanks

[pebe]

I think you have a faulty 555. When pin 6 reaches 2/3 of V+, an internal bistable flips over and switches on the discharge transistor whose open collector is connected to pin 7. With pins 6 and 7 joined, this should discharge the cap connected to pin 6.

When pin 2 triggers, the bistable flips back making pin 7 open circuit and allowing the cap on pin 6 to start charging.

Try replacing the 555.

[Chris Smith]

Stressed 555 was my first pick, which is why I use the socket and a bag full of the chips.

Also Mimms picks the way to hook up the 555 diagram that others do not show?

His never has failed, others have.

[MrAl]

Hi again Chris,

Ok, so what does his show for a monostable with very high
capacitance like 22uf?

[Chris Smith]

I don’t have those books handy, but from memory a 350 uf cap and as high as a one meg pot were some of the values I have done.

I have used 1 uf on average [radial] and 250 k pots, but I have ran the whole spectrum from 50 kHz [and higher] down to once every 24 hours.

I had all this on disk, but that disk died and I need to re scan it all back onto a Cd.

[philba]

A 1 Mohm and 22 uF cap should give you a monostable period of 24.2 seconds. Not 10 seconds. t = 1.1*R*C It sounds like you were looking at the RC table for the astable mode.

the national datasheet shows a decent schematic though I would tie reset high (pin 4). Also, couple CV (pin 5) to gnd.

could you post the schematic so we can be sure you wired it correctly?

[MrAl]

Hi there,

The circuit is the monostable circuit on the data sheet, i believe it's
figure 1, "Monostable". The nomograph for R and C is figure 3.
Clearly it shows roughly a 10sec delay for R=1Meg and C=22uf
(or near that anyway).
The data sheet is for LM555.

[Janitor Tzap]

Hi there,

The circuit is the monostable circuit on the data sheet, i believe it's
figure 1, "Monostable". The nomograph for R and C is figure 3.
Clearly it shows roughly a 10sec delay for R=1Meg and C=22uf
(or near that anyway).
The data sheet is for LM555.

Mral,

Not all 555 timers are created equal.

There are several different version of the 555 timer chip out there.
The NE555 version as you have found gets easily over loaded.
{I bet you got this one from Radio Shack.}

In any case, the one for LM555CM is designed for a higher output current.
Which is probably the one that was used in the data sheet.
There is also a CMO version well, but I can't think of what the lettering is.

Well check the data sheets, or goto the NTE site and look it up.


Signed: Janitor Tzap

[philba]

Hi there,

The circuit is the monostable circuit on the data sheet, i believe it's
figure 1, "Monostable". The nomograph for R and C is figure 3.
Clearly it shows roughly a 10sec delay for R=1Meg and C=22uf
(or near that anyway).
The data sheet is for LM555.

Yes, I see that but the formula they give says 24.2S. I'd believe that first.

[philba]

I did a quick simulation. Here's the schematic

here's the waveform


as you can, see, the period is 24.2S. Did you wire it up as my schematic shows?

by the way, what are you doing with the trigger when it's not grounded? floating? not a good idea. try pulling it high with a resistor (10K would be fine).

By the way, the cmos version of the national part is the LMC555.

[ecerfoglio]

I would tie reset high (pin 4). Also, couple CV (pin 5) to gnd. (Philba)

Good advice.

Also add plenty of power suply bypassing near the IC (at least one 100n = .1 micro between pins 1 (gnd) and 8 (+V))

Without the "extra" resistor, when the discharge transistor turns on there is a *large* current transient between pins 1 and 7.

At the same time, pin 3 goes low (another current transient in the output's "totem pole")

If the power supply's voltage goes down during those transients, or if pin 4 is left floating, it *may* cause the internal flip flop to get into some weird status (like the same level at both Q and notQ outputs at the same time)

-------------------------------------------------------------------------------------

Even if we did't see it in any book, if you have a large timing capacitor in a 555 monostable, and if you can allow a small delay (say 10 to 100 ms) between timing cycles, adding a small resistor between pin 7 and the junction of pin 6 and the cap sounds like a good idea.

You may calculate this resistor using the formula:

5 * R * C = five time constants = allowed delay (10 to 100 ms)

with C = 22 microfarad it gives

R = 10 ms / (5 * 22 microfarad) = 90 ohms

to

R = 100 ms / (5 * 22 microfarad) = 900 ohms

[MrAl]

Janitor Tzap:
Yes it was purchased from Radio Shack, you think that has
something to do with this?

Philba:
Well, i am seeing roughly a 10 second delay but i didnt check
it that closely i guess. Also, it's funny that the time this
cap and R would take to charge to 63.2% of Vcc is 13.9 seconds,
which is less time then it takes to get to 66.6%, where it is
said to get to in order to flip the flip flop.
Your circuit looks good...only difference with mine is these:
1. On the output i have LED in series with 1k resistor
2. On the trigger i have a 120k resistor pullup, and short it
to ground to trigger the device.
I could check the value of the cap i guess, or perhaps time
the timing cycle with a stop watch (or both)...i'll do that
next.

Ecerfoglio:
Yes good idea, but i dont see why the power supply would get
a negative spike on it due to current flow from pin 7 to pin 1.
Maybe a positive spike lifting the voltage even higher?
I like this line of thinking however so i'll try bypassing
the chip and see what happens.

ALL:
I'll get back here with some results hopefully in a few minutes.

A LITTLE LATER:
Ok, i added a bypass cap and timed the cap and it timed out at 23
seconds, not 10 seconds.
Secondly, i decreased the trigger pullup to 1k, but that didnt change
anything.
So far, the only thing that helps is to add a small resistor (270)
in series with pin 7 instead of shorting pin 7 to pin 6.

[rshayes]

The output stage on an NE555 is an oversized version of a TTL buffer. This uses active pull up and pull down devices. If one of these turns off slowly, both devices are on simultaneously and a high, short spike of current flows through the output stage from the positive supply to ground. This current spike only lasts for a few nanoseconds, but it can also be in the 100 milliamp to 1 amp range.

These devices should have good bypasses on the supply pins to prevent interreaction with other circuits through the power supply leads.

The CMOS versions will not have this problem to the same extent, since the output stage on these devices is usually much weaker.