"Jim Thompson" wrote in
message ...
On Tue, 17 May 2011 12:33:41 -0500, John Fields
wrote:

On Tue, 17 May 2011 14:07:28 +0100, "Ian Field"
wrote:


"Kevin McMurtrie" wrote in message
.. .
In article ,
"Ian Field" wrote:

...
It still overheats (a new Sign' NE555 failed) even after replacing the
16V
zener with a 12V part and adding a second 68 Ohm resistor inseries as
the
dropper.

It occurred to me that it didn't like driving the gate capacitance of
an
IRF740, but it still does it with the MOSFET removed.

I've run the point of a scalpel between all the tracks on the
stripboard
to
eliminate any solder bridges I might have missed.

The only thing left I can think of is maybe I overdid the Vcc
decoupling,
the 680u is a very low ESR from a VCR PSU and the 330n might only be
making
matters worse.

Its well known that the 555 draws large current spikes during output
transitions, totally OTT decoupling could be whats cooking it!

The original 555 does run hot. That's why the CMOS version is popular.

No, it doesn't like directly driving big MOSFETs. When the MOSFET gate
is driven low, inductive ringing on the drain appears on the gate via
capacitive coupling. That ringing wears out the 555 chip by pushing
the
output pin below zero volts.

Before I changed the 16V zener to 12V it was getting hot enough to self
destruct, at 12V it still needs a clip on heatsink to keep the temp' in
save
limits.

By yesterday I was starting to suspect the very low ESR 680u + 330n
decoupling was way OTT and incompatible with the 555 drawing large
current
spikes during output transitions - replacing the 680u with a regular
quality
330u and deleting the 330n made it run a little cooler, but it still
needs
the heatsink.

---
The only reason you'd need the electrolytic is if you had really long
leads from the supply, and 100nF right across pins 1 and 8 ought to be
more than enough to take care of the spike.
---

My original idea required using pin 7 way out of published spec, the
datasheet doesn't say specifically whether it applies to pin 7, but most
pins on the 7555 have an inherent substrate SCR structure so there would
be
a risk of latchup and self destruct if I'd attempted what I wanted to do
on
a CMOS part.

---
So refresh my memory; what are you trying to do?

Find his asshole with a mirror, a flashlight and a helper... but Ian
is failing ;-)


At least I'm not a cantankerous old fart.