Panache
International Hazard
Posts: 1290
Registered: 18-10-2007
Member Is Offline
Mood: Instead of being my deliverance, she had a resemblance to a Kat named Frankenstein
|
|
c mos handling precautions
if a piece of kit has a warning on it to use cmos handling precautiuons when dismantling i assume this is to preserve the component not because of
some inherent danger to the dismantler because of toxic compounds or stored current?
|
|
not_important
International Hazard
Posts: 3873
Registered: 21-7-2006
Member Is Offline
Mood: No Mood
|
|
Yup - CMOS = HiZ easily damaged by static electricity zaps. When the first CMOS input op amp ICs came out, the kill rate to them from hobbyist
projects was quite high; all the tube users wanted that hi impedance input but weren't used to the sensitivity.
|
|
mr.crow
National Hazard
Posts: 884
Registered: 9-9-2009
Location: Canada
Member Is Offline
Mood: 0xFF
|
|
I subconsciously touch something metal before I have to pick up a circuit board or a chip. I have never fried anything.
Double, double toil and trouble; Fire burn, and caldron bubble
|
|
12AX7
Post Harlot
Posts: 4803
Registered: 8-3-2005
Location: oscillating
Member Is Offline
Mood: informative
|
|
Specifically, always first touch the common terminal (ground, negative power pin, etc.) of whatever circuit or component you're handling. For a
circuit, you probably want to touch screws, or whatever traces run around the perimeter, or a battery contact, etc. Avoid internal traces. Chips in
circuit are somewhat less sensitive than naked pins, but it depends.
When picking up a chip, ideally you should touch the antistatic foam first, then pull it out. If it's not in foam, pick it up by the corners, where
the power supply pins [usually] are.
Another hazard is taking inputs outside of their rated voltage range. This can cause latchup in powered circuits, basically shorting the power supply
through the affected chip. Obviously, said chip will get hot rather quickly! This is a good reason to avoid carelessly poking away inside a powered
circuit. There is no danger of latchup inside an unpowered circuit (after all, the cure for latchup is to turn off power and restart the circuit --
assuming the affected chip didn't nuke itself in the process, in which case it must be replaced).
Series to look out for include CD4xxx and 74Cxx (but not 74xx or 74LSxx), which are the old CMOS lines. 74HCxx is the newer CMOS line, a bit more
robust, but still use caution around it.
These days, ESD isn't such a big deal, because manufacturing processes are more robust than they used to be. Keep in mind that practically everything
digital is now CMOS, and most of it runs at vanishingly small voltages (~1.2V for CPU cores!). Transistors are smaller than ever, but only on the
inside; the transistors at the pins have to be big enough to drive those slow, beefy PCB traces, often with a lot of current, so they'll be rated for
ESD according to the datasheet.
The only thing you really have to remember is charge equalization. Ground your body to the chip, then pick it up; ground body+chip to the circuit,
then plug in the chip. You'll never have a problem. You don't need grounding clips or anything, just equalize it really. (This is more specific
than "touch something metal", since an obviously example would be touching a metallic Van de Graff generator!)
Tim
|
|
peach
Bon Vivant
Posts: 1428
Registered: 14-11-2008
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by mr.crow | I subconsciously touch something metal before I have to pick up a circuit board or a chip. I have never fried anything. |
I tried suggesting that on an electric DIY forum, with an additional background of my vast experience of doing it.
You wouldn't believe the amount of shit I got back about it from people who don't even know CMOS is or how silicon works.
I'd say take your shoes off and stay sat in one place while you do it, but given the electron microscope thing I don't know if taking your shoes off
in this case would be a good idea.
The people I've seen working on computers in clubs and at stores other than PC world just touch the frame whilst it's plugged in to discharge
themselves. You need to do that each time you spend a significant amount of time not picking out components.
But yeah, if you get into the subconscious habit of doing it, it takes zero time. Your hands will be right beside the case anyway.
|
|
peach
Bon Vivant
Posts: 1428
Registered: 14-11-2008
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by 12AX7 |
These days, ESD isn't such a big deal, because manufacturing processes are more robust than they used to be. [snip] transistors are smaller than
ever, but only on the inside; the transistors at the pins have to be big enough to drive those slow, beefy PCB traces, often with a lot of current, so
they'll be rated for ESD according to the datasheet. |
I tried making that point as well, but I think it was lost on 'em.
[Edited on 31-10-2010 by peach]
|
|
Twospoons
International Hazard
Posts: 1324
Registered: 26-7-2004
Location: Middle Earth
Member Is Offline
Mood: A trace of hope...
|
|
Quote: Originally posted by 12AX7 |
These days, ESD isn't such a big deal, because manufacturing processes are more robust than they used to be. [snip] transistors are smaller than
ever, but only on the inside; the transistors at the pins have to be big enough to drive those slow, beefy PCB traces, often with a lot of current, so
they'll be rated for ESD according to the datasheet. |
Doesn't stop me shouting at people walking around the office with naked circuit boards. We have anti-static bags, so there's no excuse.
Worst thing about static damage is that things can get 'slightly' broken, and not fail properly until out in the field. We've still had issues in our
factory from time to time - we have cmos cameras attached to glass, with a cable trailing off just waiting to get zapped.
Helicopter: "helico" -> spiral, "pter" -> with wings
|
|
Contrabasso
Hazard to Others
Posts: 277
Registered: 2-4-2008
Member Is Offline
Mood: No Mood
|
|
CMOS has incredibly small junctions inside the chip which helps with low power usage, and speed. BUT these same junctions die with only a tiny charge
applied. Best to use a wrist strap and grounding wire for you and the board so that you are both at the same potential and nothing can therefore flow
to kill junctions.
The grounding wire should have a resistance of 100K to 1M ohm and the wrist strap should be conductive on the inside but insulated on the outside.
This is to prevent currents flowing of life threatening scale if you touch something live.
Sometimes you ground to the mains ground sometimes to just the same ground (bench top) that the board is sitting on and grounded to.
|
|
mr.crow
National Hazard
Posts: 884
Registered: 9-9-2009
Location: Canada
Member Is Offline
Mood: 0xFF
|
|
I see pictures on the Internet where people put their brand new video card on the carpet. ARGH! I always pick up PCBs by the edges so I don't touch
the circuits.
My soldering iron has a ground terminal on the back. That plugs into a wire that wraps around my finger.
So basically touch something metal and don't shuffle your feet or pet a cat and everything will be ok. Don't Panic
Double, double toil and trouble; Fire burn, and caldron bubble
|
|
Panache
International Hazard
Posts: 1290
Registered: 18-10-2007
Member Is Offline
Mood: Instead of being my deliverance, she had a resemblance to a Kat named Frankenstein
|
|
thank-you ALL for the response, this electron microscope is doing my head in and i'm fairly sure i'm going give up on it and just use it as an
ornament, albeit a 800kg one, one of the drawers has a nice dessicator built into it.
Cutting the loom was a big mistake but after two days trying to run them apart logically i realised i would be there forever. There were so many
wires in the loom i got serious cramp in my wrist from the sidecutters.
At 4am on the last evening i had to get the thing out i almost died when a one inch thick 2 foot in diameter stainless plate came loose ( i had undone
the fastening screw lying on my back underneath it thinking it came out from above the desk, but no the rotating plate was in two pieces) and i was
under it under the desk, i have no idea how i got out of the way, it makes me think perhaps i'm a ninja.
The cmos elements are around 200mm in diameter and around 5mm thick, does this sound right, there are like 8 of them.
|
|
Texium
|
Thread Moved 19-11-2023 at 16:16 |