hi, i was replacing a DRAM chip (i dont remember what the package name) type HM514170AJ8. i heard from someone that its very sensitive to static charge, so you better dont touch it by hand. this morning, i've got the new one and trying to replace it. well, maybe i am careless or somewhat, but after i test the board, its showing message error U14, which is mean that the DRAM chip was error. now, i am afraid that it is broken already, and have to buy one again >:(

anyone could tell me how to do it safely?

thanks
Posted on 2003-08-06 00:51:11 by dion
very sensitive to static charge

Hi!
I remember, long time ago, i SPECIALLY tryed to destroy DRAM chip with static electrisity
(just antic trick with my cat :grin: ), but without any result.
Nowadays, i think, internal protection of chips is more efficient

May be problems are in quality of ironing?
Posted on 2003-08-06 02:24:03 by S.T.A.S.
ESD??? What ESD ;)

As long as you don't discharge your self, the iron, etc., THROUGH the chip, you shouldn't have any problems.

Now you could be toasting the chip by over heating it. Or there is something else in the circuit that is frying the chip.
Posted on 2003-08-06 02:59:09 by eet_1024
ESD = Electrostatic Sensitive Device

You need to work at a grounded workstation using precautions for ESDs. Like using a ESD wrist strap.
You can find this equipment in Electronic catologs.

Basically, you need to drop the Electrostatic potential in your body ( body movement generates this all the time ) to zero or ground reference. The worst thing you could do is to work where is carpet.

Heat magnifies the ESD problem, making it easier to kill the IC.

Regards, P1 :cool:
Posted on 2003-08-06 08:53:44 by Pone
ESD = Electro-Static Discharge

I agree that a grounded workstation is ideal but, in dion's case, not necessary. The idea is to be at the same potential as the device, not necessarily ground. Neutral is the more correct term. If your body is charged to 120v, you can safely touch the hot side of your house wiring without fear because the potential "difference" between the two would be zero (don't try this at home).

ICs are a lot tougher than most people think. I can only recall one time ever zapping a chip and that was when I touched a board handed to me by someone else and felt the zap. At one of my jobs we had carpet (very little shag, easy to clean, flat stuff) all over the "lab" and in 6 years I never killed any. Both cmos and ttl.

The main thing is to occasionally touch a common object, preferably a grounded object, before you handle anything.

After all that, I'm not saying you just don't have a bad chip. Of course, you could have a bad board.
Posted on 2003-08-06 10:22:50 by drhowarddrfine
Thanks Howard for the clarification.

The TLAs in here are as thick as nats.

Regards, P1 :cool:
Posted on 2003-08-06 12:11:43 by Pone
Although the ICs may be tough, I've had a coworker zap and kill discrete FETs used in some power switching circuits.

If you're overly worried about static, use a grounded (to earth!) working surface and a wrist strap with a wire connected to the working surface. I've been told that wetting the skin area that touches the metal pad in the wrist strap will ensure a conductive contact.
Posted on 2003-08-06 16:50:54 by tenkey
THis is why i leave my CPU cord plugged in while working on my CPU.

THe CPU cord has a solid GROUND pin, which directly connects to the chassy of the CPU box. Its far more likely i will nudge the frame before my fat fingers get close enough to do damage on a chip etc.

As well i make a point of keeping one hand on the frame when pulling cables etc.

You might want to think about something similar. You dont need to get all crazy and buy expensive stuff, just get GROUNDED. (then you can be sure its your soldering job ;) )

:NaN:
Posted on 2003-08-06 17:34:05 by NaN
Ditto to NaN's reply.

Just touch the earthed chassis before touching anything else and you will discharge yourself.
For home projects, that works very well.
Most likely the soldering iron itself is earthed so you don't need to worry about that.
Posted on 2003-08-06 19:47:42 by VVV
Not most likely. One way to tell is if your soldering iron is a 3-wire or two wire. Of course, the 3-wire would have the tipped grounded while the 2-wire one will not.
Posted on 2003-08-06 20:44:00 by drhowarddrfine
wow... sorry i am late in reply. thanks for all the lessons guys :))

abot grounding problem. yes, i have the tools, wrist strap. but, i am confused where i am gonna strap it to. after discuss with my boss, i know that earth from AC cord is disconnected. now, i am confused again after reading those posts. earth?? which earth? can you guys measure it? with tester i mean? or should i strap it to other? do i have to make a connection to a real "ground"?

the DRAM chip is new FYI. i had 2/3 again in stock. its ...uhm... maybe about $10 to 50 each. uh... very expensive you know...

my solder dont have earthed wire. so, what should i do?

the package is shrinked type, dual in line, 40 pin with J type pins. any tips on soldering this kind of package??

uhm... yes, might be the board itself have problem. the first time i tested the board after soldering the chip, err msg U14 was showed up. after i trace it with scope, now the msg is scrambled... obfuscated... :(

thanks
Posted on 2003-08-07 06:34:07 by dion
When you work on multi-leaded SMDs, you need a SMD reflow/rework machine for best results. Check around on the internet for vendors.

Problem is that, if you don't do alot of that kind of work, it's cheaper to buy another PCB assembly.

Regards, P1 :cool:
Posted on 2003-08-07 12:44:38 by Pone
Ground depends on your Hydro system. I dont know all the world's standards, but common sence should *hopefully* persist.

The AC ground should be the third pin on the chord. In the receptical in your wall, it should eventually be wired back to a ground bar or something, which in turn runs directely into the ground (or under ground), exposed.

Around here, in residence anyways, they often bring the common ground point to the COLD WATER pipes (copper pipe). This is an excellent ground since the pipe runs under ground.

So if I statically discharge myself onto a CPU chasis, the current runs throught the chasis, to the chord's third middle pin, into the receptical, down ground wires to the common ground point in the basement, down a fat copper braided wire that is bolted to the cold water piping, and outside to the earth. From here charges are dispersed "through out the planet" ;)

This is kinda what you need. Some way of grounding your wrist to some electrical ground scheme that will carry charges freely out to the earth ground.

90% of this problem should be taken care of my your local hydro athority. However, the 10% is figuring how to connect yourself to the ground provided in the receptical ;)

Oh, ya, PS: Dont trust Power supply grounds for lab equipment. They are often an ISOLATED GROUND. This means they dont ground furhter that the equipment itself (which in terms of mass, is considerably smallar in volume than the plant Earth for sinking charges ;) )

:NaN:
Posted on 2003-08-07 13:44:46 by NaN
I was being sarcastic with my "ESD?" The dangers of ESD stem from when someone does something stupid like grab one lead of the device, walk on carpet, and touch ground with another lead; thereby forming a circuit THROUGH the device.

Discharging into a chip shouldn't damage it. The idea is not to pass a bunch of current through it.

Nan:
Most newer PC's keep power to the MOBO, specificly the PCI bus; this allows such things as WAN. I first discovered this when I plug a LAN card into my Celeron 433 and the LED's blinked. Now I disconnect power. As long as you touch the case, or just rest your hand or arm on it while installing or removing components, you won't have any problems.
Posted on 2003-08-09 05:17:34 by eet_1024
The dangers of ESD stem from when someone does something stupid like grab one lead of the device, walk on carpet, and touch ground with another lead; thereby forming a circuit THROUGH the device.


That is not the case.
When you discharge yourself to earth a HIGH CURRENT FLOWS between you and the earth.
If you touch a pin of the device instead of the ground, then the current DOES FLOW THROUGH the device into the ground and finally to earth.


Discharging into a chip shouldn't damage it. The idea is not to pass a bunch of current through it.


ESD can damage the chip. Therefore chips contain protection diodes (zeners) on their inputs. They are supposed to conduct the current to ground in case there is an electrostatic discharge. So the current flows THROUGH the protection diodes to ground.

However, these diodes cannot take any amount of current (read energy). Therefore the ESD protection on a device (or device pins, in case only some are protected) is usually specified using the so-called "Human body model", which models the human body with a 100pF capacitor and a 1.5 kohm resistor connected in series. This provides for consistent test results, since known parameters are used.

The capacitor is charged to the test voltage, one of its pins is connected to ground and the other end of the series combination is connected suddenly to the pin of the device. The capacitor thus discharges through the resistor and through the device to the ground (this simulates a worst case scenario).
The ratings in the datasheet for ESD protection actually specify the maximum voltage that the capacitor can be charged to prior to the discharge, for which the device is still protected. This is a measure of how good the protection is.
So, if the data sheet reads 2kV ESD protection, that device has less protection than another which is rated at 15kV.

If you do the math for these two values (2kV and 15kV) you will see that peak current is actually "a bunch". However, the time constant is so short (150ns), that there isn't much ENERGY delivered to the device (protection circuitry) during the 3 time constants it takes the capacitor to fully discharge itself.

Back to the real world, if you are charged with enough charge you CAN damage even an ESD protected device, if its rating is exceeded.
The fact that a device survives a discharge only means that the protection worked and the static charge did not produce enough energy.


P.S.
This is intended as information/ clarification, not criticism.
Posted on 2003-08-11 11:50:14 by VVV
I was refering to a chip that isn't installed (ie just sitting on a desk).

If it's on a board touch the board first (a conductive part of the board)

If the board is on something conductive, touch that first.
Posted on 2003-08-12 03:05:24 by eet_1024
wierd, i googling from days long ago and found no page describing that this chip is ESD chip :(

oh, is there someone know which DIMM/SIMM/whatever memory slot that have those chips in it? i mean memory card for your desktop computer?

thanks
Posted on 2003-08-12 05:51:17 by dion
I don't believe there ever was a time when MOSFETs were felt to be truly static resistant. Nowadays, all of the devices have protective diodes built into the packages, but the protection has limits, as VVV pointed out. It's easiest to assume all MOSFET devices (including pMOS, nMOS, and CMOS) are static sensitive, and to be extra careful when handling them.
Posted on 2003-08-13 20:07:29 by tenkey
thanks for the tips tenkey :grin:
Posted on 2003-08-14 06:45:53 by dion