[FTC]: Static Electricity Discharge Video

[PhilBot]

Hi

As a veteran team (2818) I've heard a lot of talk over the years about NXT's locking up due to static electricity.

I've never experienced it, until this weekend.

At the end of a match, as the robot was going up the ramp it just stopped, and it was determined that the NXT had frozen.

At the time I put it down to murphy's law.

However, my opinion changed as I was driving our robot around on our practice field tonight. As it touched the ramp a HUGE (maybe 3/4") spark jumped from the ramp to the robot frame. And proceeded to lock up the NXT.

Holy Cow... it was amazing.

So, wishing that I had a video, I attempted to recapture the event.
It didn't take me any time at all to get the robot to cause sparks.
So, I've attached the video with two events.

http://youtu.be/LanVM9ogLl4

Note: Neither of these times did it reset the NXT... I guess it wasn't big enough.

In both cases I just had the robot do a few loops and then approach the ramp. The sparks actually occurred at the exposed metal portions of the ramp.

This would be hard to avoid.

I've seen ALL the FIRST recommendations re Static Discharge. But like many of my friend teams... I think eliminating the cause would be the preferred method (ie: anti-static mat tiles)

Phil.

[MattRain]

Very interesting! We have been having a lot of problems this year when it comes to static on the ramps, compared to all of our past years as well. (Our team is 2844, and 8640). For the 8640 bot, we have a bar that is only 3/8 inch off the tiles... as a result, we noticed a large static shock coming from it as well. For a fix right now, we have covered the bottom of the robot in electrical tape.

Another thing to keep in mind, is the fact that almost every field element of the past was made of wood, and not pre-manufactured by Andymark. Having elements that are made of mostly metal is not helping the static problem. I would love if FTC started to go away from the smooth foam tiles. It just creates way to much static.

[PhilBot]

Hi.

Yes, we are designing OUT our metal goal guide and replacing it with a plastic one. Anything that keeps conductive elements away from the ramp seems like a good idea.

Didn't think of just taping it

Phil.

[driesman]

Hey Phil - Nice but unfortunate video. FYI, Lydean says the fields have been sprayed down this year. Don't know how well, how often or with what. Qualitatively, I've seen significantly fewer dead bots at our qualifier and the one that I ref'd.

How well are you protecting the input to your Samantha? As much as I dislike black boxes, we've been using the L-Com USB surge protector recommended by FIRST w/o adverse effect. Last year we used chokes on the USB cable and the power to the samantha.

With that said, giving no path for a discharge is a good idea. Let me know how things work out.

-Andy

[maths222]

FTA here, with a few suggestsions.

• Even if you have not tended to have issues in the past, please follow the recommendations in the static document from FIRST.
• If possible, use the L-Com surge protector; if you can't, at least use a ferrite bead.
• This year, FIRST has officially recommended we use anti-static spray at events, although they are not requiring it. (In Illinois, we have been using it whenever possible, and it seems to be effective.)

(Also, my guess is that the weather is drier than it has been recently if you start seeing a spike in lockups and sparks-humidity greatly affects that.)

[PhilBot]

Originally Posted by driesman

How well are you protecting the input to your Samantha? As much as I dislike black boxes, we've been using the L-Com USB surge protector recommended by FIRST w/o adverse effect. Last year we used chokes on the USB cable and the power to the samantha.
-Andy

Without knowing it, we have been implementing most of the recommend precautions all along. We always have the NXT and Samantha mounted on a plastic carrier with a short interconnecting cable. That may have been why we have been immune for so long. Strapping the Samantha (and therefore cable) to a metal C-Channel always seemed like a bad idea to me just based on radio reception.

Based on the finding of that FIRST report I've ordered a surge protector. Although it does go against my long term belief that connectors are the key problem in mobile systems most of the time. Adding the Surge protector adds two more connectors. I'd have gone with the version with a pigtail, except they are all so long Do'h.

I've also started spraying our practice field with Staticide. Mmmm smells nice

Phil.

[westom]

Originally Posted by PhilBot

In both cases I just had the robot do a few loops and then approach the ramp. The sparks actually occurred at the exposed metal portions of the ramp.

This would be hard to avoid.

This failure is routinely avoided. But comprehension of some basic concepts such as grounds and impedance are required. If you have only one ground, then this and other events cause failures.

Parts such as interface semiconductors and switches are rated at 15,000 volts and 20,000 volts. But that number only exists when part of a 'system'. For example, a switch connected to 5 volt electronics may be rated for 20,000 amps. But one switch contact would connect to the digital ground or Vcc. Whereas that switch body must connect to a chassis ground - a completely different ground. Another ground might be analog ground. Of course power supplies can have a floating ground. Other grounds may also be necessary. If any ground are electrically connects to another, it must only be at a single point. Only then are each ground electrically different so that transients (such as static electric discharges) cause no problems.

We would do this with computers. Place that computer on a glass table (because other materials such as wood are electrical conductors). Then create static electric charges in a body. Discharge that charge into the computer chassis and keyboard at various locations. If a motherboard digital ground is electrically different from its chassis ground, then a computer does not crash. In fact, where the discharge exists relative to where that current exits a computer to connect to feet (shoes) can help determine where a ground defect exists. Follow the current.

A PCB board typically has a solid copper ground plane. However voltages vary significantly across that copper plane. If a static discharge current enters the chassis, goes up a metal standoff, crosses the PCB board via that copper plane, then returns to the chassis via another standoff; then a crash happens. Current passing across the copper ground plane (a digital ground) causes major voltage differeneces across that copper ground plane. All semiconductors reference ground. Therefore need a constant ground voltage.

These concept are often foreign most techs or electricians. Because it involves concepts only taught to engineers. One key parameter is impedance. Impedance in that sollid copper ground plane is why electronics boards have capacitors all over it. Because normal mode voltages vary significantly across that copper sheet. For same reasons why you are suffering crashed from something so trivial and static electric discharges (a longitudinal current).

Your system apparently has its digital or analog ground connected in multiple locations to the chassis ground. Then specification numbers for interface semiconductors (ie 15,000 volts) are compromised by bad grounding.

[westom]

Originally Posted by PhilBot

That had got to be one of the most condescending replies I've ever received on Chief Delphi.

Nothing was or is intended to be condescending.

Everything in a system is somehow grounded. Apparently your system is using a floating ground. Fine. Others who post have no idea what that is. If that makes you angry, well, I am sorry. But ABS does not eliminate a bad grounding problem and resulting static discharge currents.

Failures created by static electric discharges are not eliminated by isolation. For example, in their early days, Apple computers would crash because static electric discharges would flow through their plastic keyboard. Those failures were eliminated by painting keyboard interally with an electrically conductive paint. And connecting that paint to the chassis ground.

Early IBM PCs also had this mistake. They also assumed plastic would isolate electronics. A fundamental difference between PC and AT style keyboards - a chassis ground wire was added so that non-conductive plastic did not conduct static currents into digital circuits.

Again, nothing was condescending other than your nasty reply. We often found the most advanced designs compromised by overlooking these concepts. In one case, they used a new floor wax that created static electricity. Then operators touching a plastic switch crashed the computer controller. We discovered a switch, normally chassis grounded, was electrically disconnected by an expoxy paint. Painters in later versions had painted the inside of that switch panel. Neither techs nor electricians understood why that interior surface must not be painted. So static discharges were connecting to the controller via thick ABS plastic on the mushroom switch.

Isolation alone does not avert intermittent crashes created by static discharges. That says nothing about anyone's emotions. The mistake is often overlooked even by geniuses. ABS isolation alone does not avert the resulting currents.

Any tech painting a surface that was not previously painted or forgetting to add a critically important tab or wire can create problematic static electric discharges through ABS isolation. Please only read what was posted rather than insert your emotions. You should have known that isolation alone does not avert static electric discharge currents into electronic controllers.