Solving static electricity problems by grounding the frame

[Stuart]

no you cant ground your frame to your battery. there is a rule against this, and rightfully so. Remember ground is only ground when it really is ground (as in earth . . mostly) Ive personally measured a 3 volt potential between 2 "grounds" on the old 2006 batteries. If 2 robots were to come in to contact with each other even as little as 3 volts between their frames could be very bad.

I really hate to be a Debbie Downer on this subject.

[AdamHeard]

Originally Posted by Stuart

no you cant ground your frame to your battery. there is a rule against this, and rightfully so. Remember ground is only ground when it really is ground (as in earth . . mostly) Ive personally measured a 3 volt potential between 2 "grounds" on the old 2006 batteries. If 2 robots were to come in to contact with each other even as little as 3 volts between their frames could be very bad.

I really hate to be a Debbie Downer on this subject.

and don't forget that a grounded frame allows a single wire failure to short the battery (a single positively charged wire coming loose).

[ErichKeane]

The battery short issue is easily solved by using fuses and circuit breakers. For a competition like this, I've been very surprised that there isn't a rule requiring inline fuses on the positive source lines of all the batteries. Otherwise, the fuses will prevent a short from causing any issue.

As for the difference in potential between two robots, it would cause no problem whatsoever, since only grounds would touch, so there wouldn't be any power loop.

Think about this example: you have a car that needs a jump. What is the first step? You connect both grounds on the batteries to eachother. In a jump situation, the difference in potential is going to be much greater than 3 VDC, and yet there is never a problem.

I will connect negative terminals of batteries all day, there is no danger.

By disallowing the grounding of the frame, FIRST is causing an even more dangerous problem, the situation where a single positive wire contact would could give the entire frame a POSTIVE charge, which, through a human resistor, could cause extreme physical damage.

There is a very good reason all cars have their bodies/frames grounded, it is for the obvious safety benefits.

[RyanCahoon]

Originally Posted by ErichKeane

The battery short issue is easily solved by using fuses and circuit breakers. For a competition like this, I've been very surprised that there isn't a rule requiring inline fuses on the positive source lines of all the batteries. Otherwise, the fuses will prevent a short from causing any issue.

There is, that's what the breakers in the power distribution block are for, as well as the 120A main breaker. There'll always be some wire that has to connect directly to the battery, which gives you a risk of an unprotected short.

Originally Posted by ErichKeane

By disallowing the grounding of the frame, FIRST is causing an even more dangerous problem, the situation where a single positive wire contact would could give the entire frame a POSTIVE charge, which, through a human resistor, could cause extreme physical damage.

While I don't think there's any debate that a shorted chassis is bad, as you have pointed out yourself, in order for current to flow through a person and cause damage, they would have to be holding on to a negative connection while they were touching the chassis.

Conversely, with a grounded chassis, if a positive lead touches the frame, you've created a short to negative, meaning a massive current flow, causing the battery to heat due to internal resistance. Hot material expands, and the last thing first needs is a leaking lead-acid battery, much less the fire hazard.

Originally Posted by ErichKeane

There is a very good reason all cars have their bodies/frames grounded, it is for the obvious safety benefits.

Not really, it's more a convenience incentive, that way only one power wire (positive) has to be run to devices, leading to less cable runs.

--Ryan

[Al Skierkiewicz]

OK, time for a break here...
The static buildup most of you are experiencing is the plastic wheels moving inside a metal housing and the wheels slipping on the plastic floor. Plastic being an operative word here as it is non conducting. Therefore any of the normal methods one uses to dissipate charge have no effect since there is no path for the static to drain off other than through the humidity in the air.
The robot frame is insulated from the control and electrical system for one very important reason, we want you to be able to play. If you were allowed to add fuses or any other kind of protection device, when they tripped your robot would be dead in the water. The simple act of isolating the electrical system from the frame of the robot prevents abnormal robot to robot contact from taking you out of the match. A side benefit is that the control system remains unharmed and the chance of fire is reduced. Remember that the battery is capable of over 600 amps when fuly charged for about 7000 potential watts.
In addition to all of this protection, there is a 3 amp breaker built into the PD in the negative lead of the Crio power to protect the Crio from any stray shorts that may develop. (Thanks Russ and Erik) Please remember that Inspectors will be checking for electrical shorts to robot frame. Most often these come from poorly insulated sensors but this year will also come from the Crio and camera. Both of these devices must be insulated from the robot frame per R41...


<R41> All wiring and electrical devices, including all control system components, shall be electrically isolated from the ROBOT frame. The ROBOT frame must not be used to carry electrical current (e.g. this is necessary due to polarity reversals that occur under certain operating conditions such as during motor direction reversals).
The chassis for the cRIO Mobile Device Controller and the supplied camera have grounded enclosures. Under this rule (and for their protection), it is REQUIRED that they be electrically isolated from the ROBOT frame when installed on the ROBOT.

You are probably asking yourself 'has he really seen anything like this before?' and the answer would be YES. Smoke, flame, and robots out of commission. I know that inspectors may give you pause but our job is to insure that you play as long as you want or can. We want to see you play and will do everything in our power to insure that.

[EricVanWyk]

A drag chain is a good idea.

ErichKeane -

As several others have pointed out, connecting the battery's return to the chassis is a bad thing that is illegal for very good reasons. This has examined and agreed on by many EE and safety experts, including our friends at UL.

I've gathered a few quips from this thread:

AdamHeard - and don't forget that a grounded frame allows a single wire failure to short the battery (a single positively charged wire coming loose).

RyanCahoon - Not really, it's more a convenience incentive, that way only one power wire (positive) has to be run to devices, leading to less cable runs.

Al Skierkiewicz - The robot frame is insulated from the control and electrical system for one very important reason, we want you to be able to play.

And I'll add a slightly less elegant one:

The battery short issue is easily solved by using fuses and circuit breakers.

This is not true. In order to do this, all of the smaller paths need positive AND negative breakers. If a 5A return is sourced by a 40A source, the 5A return will fail. The converse is also true.

"One Breaker per Path" is only viable if your topology is a star. If the ground is already shorted, you are one mistake away from breaking that assumption.

[ErichKeane]

A fuse in the shorted circuit would prevent any discharge from causing significant damage if the frame was grounded.

Vehicles are grounded because it is safe. The simplicity of the electronics system is a byproduct of running the whole body as a ground, but in the end, it is still safe enough that every car made in the last half century does it!

Why would you need a ground-breaker? I've never seen a system where this is even done before. Again, in your car, each circuit has a fuse, but the master fusible link prevents the positive line from getting grounded and causing damage.

When it comes down to it, a grounded frame would be safer. I'd rather have a short cause a fuse to break, than have the chance at harming a person. A grounded frame makes a short immediately obvious! Otherwise, it causes massive electriction risk to anyone around the battery! Touching a ground wire on the bot is fairly easy, all it would take is someone installing/removing the battery while in contact with the frame.

I also still don't see how two colliding bots with grounded frames cause any issues. Attaching the ground wire from two batteries causes absolutely no problem.

[Al Skierkiewicz]

Eric,
Remember that in a car the frame is not "grounded" it is the common connection for the power supply. It was and is used to make manufacturing cheaper by eliminating the need to run two wires eveywhere in the vehicle. However, things go awry when body parts have high resistance, loose hardware and defective welds. And in a vehicle collision, electrical faults usually serve to stop the vehicle operation, a result we don't want to happen.
The battery is 12 volts and does not produce an electrocution hazard to First participants.
Electrical contact of auto body parts does produce electrical problems in modern day auto electronics. Although "jumping" a dead battery is normally though of as OK, two autos entwined in a collision can cause serious currents to flow in the bodies of the vehicles. If a car battery is capable of 5500 cold cranking amps and that passes through your car's computer instead of directly to the battery as in jumping, your are looking at replacing a very expensive piece of hardware. It depends on the path the current takes.

[EricVanWyk]

Originally Posted by ErichKeane

I also still don't see how two colliding bots with grounded frames cause any issues. Attaching the ground wire from two batteries causes absolutely no problem.

I don't either, I think this reason is simply wrong.

The other reasons I still agree with.

In any case, grounding the chassis is illegal in FRC.

[ErichKeane]

Originally Posted by EricVanWyk

I don't either, I think this reason is simply wrong.

The other reasons I still agree with.

In any case, grounding the chassis is illegal in FRC.

I realize it is illegal, I was going to ask for a rules clarification in the Q&A, but in reading through here, I'm convinced that is what the rule means.

I just still haven't heard a single reasonable reason why grounding the frame is a bad idea. If car bodies are grounded because it is cheaper, why don't higher end cars electronically isolate the bodies if it is at all safer? The reason is that grounding the body has some huge advantages, 1st- It prevents static (which is especially unsafe in fuel-based vehicles), and 2nd- it isolates, identifies and removes any shorts immediately through smart fusing.

Another household device which has the body grounded is your computer case, for the same reasons. Most metal cases are grounded (as are many laptop cases, if not, at least the trays) for these reasons, and the computer manufacturers don't face huge risks from awry wiring! Again, it is to protect the user.

The ONLY somewhat reasonable reason i've seen so far is the delicacy of the cRio. In a grounded system, it is possible that a positive short to a grounded frame could cause problems with it. However, that begs the question: why isn't the cRio designed with removable diodes to prevent this problem?

We've already seen that the cRio is susceptible to ESD (a WHOLE 'nother rant!), which should be avoidable if ITS frame was grounded.

As for two vehicles exchanging electricity in a collision: the only way this happens is if the positive cables touch. Vehicles touching grounded bumpers happen all the time, and I've never seen, nor heard of a single one causing computer/fuse issues. Again, the only situation this causes a problem, is if one has a positive short that touches another ground.


So I guess this discussion has dropped into an academic one (since FIRST's rule, is FIRST's rule). Does anyone have a plausible true reason why this would cause an issue?

[Russ Beavis]

The cRIO isn't ESD sensitive - the Driver Station is. Have you ever caused the cRIO to "fail" with an ESD event?

The auto industry is ALWAYS doing safety vs. cost analyses. There are plenty of single-point failures that can result in dangerous situations but they've done everything "reasonable" to ensure that such events are unlikely. Imagine, for example, that your steering rack snaps or your brake pedal becomes disconnected. Automobiles are NOT fail-safe under every circumstance.

I don't agree with the rationale provided in Section 8 regarding the need for chassis isolation. Collisions would need to have multi-point/multi-differential voltage contact to get really exciting. For example, birds don't seem to mind sitting on high tension wires. Current needs to have a return path in order to flow.

I do, however, agree with the FIRST isolation rule for this very specific reason - if the chassis is grounded and ANY cable touches it, current will flow.

The chassis current will be limited by the breakers in most such failures but there are arcs that can be struck and maintained at < 40 amps. I've personally experienced this in non-FIRST applications. This is one of the primary reasons why UL often limits low-voltage devices to 8 amps.

The most dangerous instance, by far, is when the battery shakes loose and its positive rail touches the chassis. Now THAT'S exciting - a dead short on a 12V/18AHr lead acid battery!! The 120A main breaker won't help you then.

The second most exciting failure is when the output from the 120A main breaker is shorted to the chassis. That breaker takes a while (relatively speaking) to open since the short circuit current for the battery is "only" a few hundred amps. That's a lot of damage before the breaker opens.

The auto industry and PC industry have obviously weighed the pros and cons. I would NEVER consider shorting my car battery's positive output to the chassis to test for isolation and breaker operation but, if you care to try this on your own car, please post the video on Youtube for everyone to watch.

Russ

[Al Skierkiewicz]

OK,
Let's hit these one at a time.
Computer cases and most home appliances are at ground/earth not power supply common. No current flows in the case unless there has been a catastrophic failure and then the power breaker or fuse trips. This is an attempt to protect the user from death, not prevent failure in the components.
Cars do have static buildup but they are huge and produce a rather nice capacitively coupled connection to earth through that nice plate and conductive additives in the tires. Static sensitive components in your car are designed to operate in this environment through the use of shielding, isolated power supplies and internal protection. Ask any car electrical engineer and they will tell you that the electrical environment on a modern day vehicle is one of the worst environments they have to contend with.
Auto body common power supplies are not ground nor are they earth they are power supply common. No current flows between the car battery and the ground beneath the car. If the addition of one wire would add $.025 cents to the cost of a car, the manufacturer making 1 million cars would be spending an additional $250,000 a year per wire.
Metal bumpers are the exception rather than the norm these days. You will find them on some vehicles, mostly SUV or truck designs. Again you have to visualize the current flow. Where does it go. Does current flow if the positive lead of one battery is connected to the body of another car? No, there is no complete circuit. If the two bodies come together than the path is through the positive lead of the first battery, through the second car body, then through the common body contact and then through the first car body and back to the first battery. Where the current flows is anybody's guess and it is only limited by the resistance of the connections in between. There is a real reason why firefighters cut the battery leads when they first appear on an accident scene. High curent translates to sparks, heat, and explosive batteries.
Grounding the case of the Crio to the robot frame will not prevent ESD. Again the issue is with the discharge path, huge currents develop huge voltage potentials across resistance. Think about this...how much voltage is developed by 1000 amps flowing through a one ohm resistor? Ohm's Law says V=I*R. V=1000*1=1000 volts.
As I have discussed earlier, First wants it's teams to be able to play, as long as they want or can. Do you want to be dead during your finals match on Einstein because of an electrical mishap in a simple robot collision? Do you want your Crio to smoke when you have no money left to buy a new one? Do you want your students to watch in horror as smoke and flame rises from their robot or that of one of their alliance partners or both partners?

[MrForbes]

Originally Posted by ErichKeane

Another household device which has the body grounded is your computer case, for the same reasons.

Household device wiring is quite a different case...one side of AC line power is connected (literally) to the ground. The case of the computer is grounded to keep the exposed metal parts of the computer at ground potential.

Anyways, I don't see any advantage to connecting the wood frame of our robot to the negative battery terminal.

[ErichKeane]

Er, your computer case IS connected to the negative terminal of the DC power supply, through the brass stand-offs of the motherboard. The power supply itself is the only one routed directly to the 3rd pin AC Common.

And yes, grounding your wood frame would have minimal advantages...

[MrForbes]

Originally Posted by ErichKeane

The power supply itself is the only one routed directly to the 3rd pin AC Common.

The 3rd pin on the line cord (which most of us call the ground or earth terminal) connects directly to the metal power supply case, which is connected directly to the computer case by it's mechanical mounting. The negative output wires of the power supply are soldered to the ground path on the power supply circuit board, which is mechanically connected to the power supply case also.