Search



The ESD (Electro-Static Discharge) Guide for the Hobbyist

ESD_Susceptible.pngI will attempt to always begin my guides with a bit of background information to let you know why I'm qualified to speak on the topic.  As of the present time while this guide is being written I am a Master's student in electrical engineering, showing I have the knowledge, or at least research skills, to talk about an electrical topic.  I have also held a job for about a year at a military contractor where ESD training and compliance inspections were required.  On a somewhat related note, I am a certified IPC Specialist (sounds fancier than it is), which basically means I had professional training in soldering, including surface mount work.  Finally, I'm a college student with an insatiable curiosity and desire to build things, but without the extra income to spend on non-essential items.

Why should I care about ESD?

I've seen the above argument from hobbyists more than I care to mention.  The usual justification is that components are cheap, ESD problems are rare, etc.  Well, some of this is true sometimes, sometimes not.  Nowadays msot components are designed to take a good bit of ESD abuse.  While it's rarely enough protection that they will survive if you go out of your way to try and damage them, it's sufficient for casual handling.  If your idea of electronics is an 8/16-bit micro, a 555 timer, and a 7805 for power, you're fine.  Go away.  Leave.  This guide is of no use to you.  When you're ready to come back later, remember where this guide is.

What components are not ESD-tolerant then?  The most commonly-damaged components I've personally seen are higher-end digital components.  If you're working with a 32-bit microcontroller such as an ARM or a TMS, ESD can be a very serious problem.  Peripheral support circuitry for these is also sensitive: RAM, fast A/D's, and network communication to name a few.  Also along the same line of thought FPGAs sometimes have very poor protection.  In the analog realm, mosfets are notoriously bad.  They've come a long way over the years, but they still need a bit of consideration when handling.  This is especially true if you're working with high-power (ironic, isn't it?) or high-speed devices.  All of these components are neither cheap enough (for me anyway) nor rugged enough to forego ESD protection at all.

Oh no!  OK, what should I do?

Let me start off by saying what you should not do.  I found a link to a hobbyist ESD guide, and some of the comments scared the hell out of me.  In particular this one:

"Back in the old days if ever we were unsure the easiest way to test it was deliberately charge youself up as much as possible, grab the metal part of a screwdriver (tightly I might add or you feel the shock) then touch whatever we were using as the earth with the tip of the screwdriver. If you saw a small flash and heard the crack then it was safe to use." - James Turner

(see link)

I'd like to repeat it again.  Don't do that!  It's about the quickest way to ensure death that I can think of.  If you can see a flash from you to your ESD discharge area I think you're doing something wrong.  If it makes an audible sound, I know you're doing something wrong.

For those of you that don't know, Jedec is a standards organization covering just about everything I can think of.  One of their standards, JESD625-A covers "Requirements for Handling Electrostatic-Discharge-Sensitive Devices".  As far as I know, this is the de-facto standard companies live and die by for ESD protocol.

For safety reasons, Jedec says your ESD equipment should conduct less than 0.5mA from the highest nearby supply voltage to ground.  This is here for your safety.  What does this mean?  Well, here in the US, your highest supply voltage is likely going to be 120V, meaning your grounding path should have at least a 240kΩ of resistance.  It's going to be 450kΩ for you UK guys out there.  Remember, this is a minimum.  Typical ESD protection offers an impedance greater than 1MΩ to ground.  Again, this is for your safety.  Consider for a moment that you followed that guy's advice that I quoted a minute ago.  You have a huge hunk of steel you're leaning on to keep yourself grounding while holding a soldering iron in your right hand, leaning down, carefully tinkering away.  This iron is a good couple of years old, and a frayed wire accidentally exposes the handle to 120V.  Your body is now a good, low-impedance ground path for this electricity to flow 'safely' through you, where it can do no harm to any of your precious components.  On the other hand, 0.5mA is about the boundary where you can even feel a shock.  If you're safely grounded with more than 1MΩ, you should conduct less than 0.25mA, not only surviving the ordeal, but doing so without even feeling it.

1MΩ or more?  But what about my sensitive circuitry?!?

I've seen this comment come up, even from fellow engineers that should know better.  If you want to just trust me, go ahead on to the next section.  Otherwise you can continue reading.

I will once again digress into a standard.  This time it comes from the department of defense in MIL-STD-883.  Quick warning - the linked standard is a little long and dry.  This is the standard where they define the 'human body model' for ESD compliance.  It basically says that the human body's static charge can be modelled by a charged 100pF capacitor in series with a 1.5kΩ resistor.  This model is probably the most used for semiconductor device protection.  You should see it in almost every datasheet.  You should also see a voltage rating these devices are designed to withstand without degradation.  This means the human body model can be charged up to that voltage and discharged through the device without harm.

ESD.pngYou will likely see ratings of at least 1000 volts for hardier parts designed to interface with the outside world; things like your 555 timer or 16-bit microcontroller.  Most likely anything you'll ever see in a packaged part will be rated for 100V or higher.  If you're working directly with an exposed die, ratings can be 20V or lower.  These devices are outside the scope of this guide; if you're working with them you have an expensive manufacturing facility and can afford to hire somebody for consulting.  In my opinion, 200V is about the worst-rated part a hobbyist would expect to encounter.

Now, I'm not sure what kinds of voltage a person can actually be charged up to.  I did find out, however, that 10kV is the level where you can get a very loud, painful, spark.  If you're charged above this point, you probably shouldn't be working around electronics. 

So, to finally answer the question: "1MΩ or more!?!? Are you sure that'll save my precious devices?" I made a pretty graph for  you.  It's the image appearing immediately to the left of or above this paragraph.  The chart plots your voltage versus the required discharge time to safely handle a susceptible part.  The calculations assume the human body model is discharged through a 5MΩ ESD ground (this value is quite typical) and the part you are handling is rated for 200 volts.  Note that the discharge time is measured in milliseconds.  For comparison, a typical blink of the eye takes several hundred milliseconds.  So, if you can hold off on touching your part for 1/100 of an eye blink, a 5MΩ ground is sufficient.  I suppose the particularly impatient among us could settle for a 1 or 2MΩ ground path.

OK. So what do I need to do?

If you were in a perfect world (a high-tech manufacturing firm?), you would wear a conductive ESD smock, walk on a grounded conductive floor, have a conductive ribbon dangling off your shoes so that it never stops making (electrical) contact with the floor, breathe in moist deionized air, sit down in a grounded chair, put on your grounded wriststrap, lean over a grounded workbench, grab your grounded tools, and get to work.  Yes, I'm serious.  I've worked in a place like this and have visited many others like it.  The protocols are really amazing.  You're grounded all the time.  Parts are carried around in ESD bags.  Carts are sprayed down with ESD spray so they're conductive.  An alarm goes off at your workbench if it detects that you've built up even the slightest charge.  It may seem over the top, but it really isn't so bad when you consider a single damaged component can cost you thousands of dollars in parts, labor, and damage to your company's reputation.  So do I do all of this at home? No.  However, there are some easy and cheap steps you can take to control static and prevent the vast majority of ESD accidents.

I would like to give you fair warning that some of the links in this section are affiliate links.  Not all, but some.  You have my word that I didn't randomly pick them, or pick the most expensive items for the job.  Actually quite the opposite.  I spent a good bit of time finding these products.  Also, unless otherwise stated, I own these exact items or something very similar.  Note that the total cost can be under $50.

Here is my 'must have' list:

I'd like to take a minute here to discuss 'proper' grounding for your wrist strap and ESD mat.  DO NOT ever stick a ground wire into the third prong of your outlet.  It doesn't just look ghetto and scary, it is.  The best plan would be grounding to a metal water pipe.  Failing that, you should attach your ground to the screw on one of your wall outlet plates through a 1MΩ resistor.  There should be a single screw (two for a switch) on most outlets to hold the plate on.  This is attached (internally) to ground.  The 1MΩ resistor is there just in case it isn't.  Building wiring is often wrong - sad but true.  The resistor won't protect your circuitry (well, it might), but it will save your life if somebody attached the outlet's ground to hot.  A low-impedance wire connected directly to a mis-wired ground pin wouldn't be so forgiving.  The best way to accomplish this is with a small, circular wire terminal at the screw, a length of wire, then the resistor, then more wiring to your workbench.  Put heatshrink tubing or electrical tape over the resistor.  All of this is just to keep exposed wires and ghetoness to a minimum.

You should now know and understand ESD control a little better.  Hopefully I've inspired you to do things a little better and a little safer.  Please feel free to add your comments below.

 

Original publication date:

2009-10-04

 

Last Updated:

2009-11-01

Comments: