Voltage on Frame

[Greg]

When we shipped the robot, there was one small problem remaining that we could not fix. When we checked the frame for voltage there was about 3V going through it. Last year we had a similar problem, but there was 12V. It turned out to be a limit switch with pins touching the frame.

This year we were unable to find the source of this mysterious voltage. I pulled all fuses out. There was still voltage. Unplugged the RC and all limit switches. Still there. I wasted about and hour on this before putting the bot into the crate. I think we'll need to eliminate this problem to pass inspection (the frame is not supposed to be used for power distribution). Any ideas on what to look for when we get to our first regional?

[jacob_dilles]

well if you dont tell them they probably wont check it. but thats bad GP . so to get rid of frame voltage...
*check all MOTORS. all connectios where wires touch motors. unplug the motor and see if it is going thrugh the winding or watever. if it is, thats a BIG problem.
*electronics board. last yeah i build my own "ground stud". all of the electronics were on lexan, but we had a huge frame-grounding problem. it turns out that the screws that held the standofs that held the groundstud were touching a structal member.
*check wires. sometimes wire can get pinched and short with the frame. bad striping jobs, exsessive heat... etc
*sensor circutry. as well as custom circutry. make sure it FAR away from the frame. be libral with your electrical tape use!
*and finaly, the battery. one year our battery holder was made out of aluminum. must i go farther?

[KenWittlief]

There is no place to get 3 volts from the robot RC or power system, unless you have something like the tether cable (RS-232 port) or maybe the link to the radio shorting to the chassis.

chances are you are getting a false reading - ifyou put a digital voltmeter on something that has very high resistance, you can pick up stray static and false readings on the meter. try putting something like a thousand ohm resistor across the voltmeter leads, and Ill bet your "3 volts" dissapears.

If by some chance it remains, then the only place I can think that it could come from is the radio cable from the RC. Does it run through the frame anywhere? is it possible it got pinched or sliced somewhere?

BTW - Im assuming you measured this with the black lead from the voltmeter on the negative (gnd) lead of the battery? another way to check for shorts to the frame is to unplug the batterys (both) and use the ohm meter part of the DVM - see if you have low resistance from the +12 or the gnd posts to the frame anywhere - it should be in the thousands or ten thousands of ohms -if you see <100 ohms to the frame from either one, then you have a shorted cable to the frame somewhere.

[jacob_dilles]

Originally Posted by KenWittlief

There is no place to get 3 volts from the robot RC or power system

or so one would think. resistance can drop voltage. it doesnt have to be a direct short. if a oxidized piece of aluminum 1/64 of a inch wide pierces a 5v wire partialy, it could have 3v on it easly.

just a thought.

[KenWittlief]

Originally Posted by jacob_dilles

or so one would think. resistance can drop voltage. it doesnt have to be a direct short. if a oxidized piece of aluminum 1/64 of a inch wide pierces a 5v wire partialy, it could have 3v on it easly.

just a thought.

only if the short was delivering power (currrent) to some device - but that would require two shorts, one to source the power and one to consume it.

[jacob_dilles]

power --> short --> multimeter --> ground

Originally Posted by KenWittlief

only if the short was delivering power (currrent) to some device - but that would require two shorts, one to source the power and one to consume it.

[KenWittlief]

the impedance of a Digital Voltmeter (DVM) is over a million ohms - so the 'short' would have to have millions of ohms of resistance - that is a possiblility -if someone spilled something on the breaker or ground, and it was salty or only slightly condutive, then you would see voltage on the frame with a DVM - but if the resistance is that high its not a problem

you can put your fingers on the probes of a DVM and it will register several thousand ohms - much less than the input resistance of the meter iteself when its on the volts scale.

[Yan Wang]

Each year, we build most of the robot on the frame before taking stuff off to have the frame welded (it's initially bolted together). We didn't take off electronics last year and the electromagnetic field caused by the welding completely fried all the electronics... that little messup cost us about $1000. Moral of story - large EM fields are bad around the electronics.

[Tom Bottiglieri]

Just use a wood base.... duhHH!!!

[Solace]

................. or you could just build your robot out of plywood, the way all the real teams do..............

on a side note, does anyone remember the HDPE at the top of the ramp last year creating some serious static charge on the robots? has anyone had any problems like that this year? just wondering.

[jacob_dilles]

i could only see it being a problem if you made your wheels out of something like wool, and had a wire running from them into your control board...

[Solace]

never underestimate the creativity of tired people...

[Mike Betts]

Originally Posted by Greg

When we shipped the robot, there was one small problem remaining that we could not fix. When we checked the frame for voltage there was about 3V going through it. Last year we had a similar problem, but there was 12V. It turned out to be a limit switch with pins touching the frame.

This year we were unable to find the source of this mysterious voltage. I pulled all fuses out. There was still voltage. Unplugged the RC and all limit switches. Still there. I wasted about and hour on this before putting the bot into the crate. I think we'll need to eliminate this problem to pass inspection (the frame is not supposed to be used for power distribution). Any ideas on what to look for when we get to our first regional?

Greg,

If your frame is indeed floating (as it should be) the voltage is UNDEFINED and not "zero" as it has no reference. It could just as easily be +100V or - 500V... Is is just not defined... In general, it will usually come to rest between 0V and 12V due to leakage resistance of the wires.

Take Ken's first suggestion to heart and use a 100K resistor to ground and measure. Then 100K to +12 and measure. If you get 0V in the first case and 12V in the second case, you have no problem.

We had an inspector at NE about 4 years ago who knew his stuff. He asked the students to make that exact measurement (frame to ground) and then asked them to explain where the voltage came from. It got them to thinking about Ohm's Law. I later used an extra 100K pot to prove to the students that there was no short.

[Mike Betts]

In an eariler post, I had said:

If your frame is indeed floating (as it should be) the voltage is UNDEFINED and not "zero" as it has no reference. It could just as easily be +100V or - 500V... Is is just not defined... In general, it will usually come to rest between 0V and 12V due to leakage resistance of the wires.

Take Ken's first suggestion to heart and use a 100K resistor to ground and measure. Then 100K to +12 and measure. If you get 0V in the first case and 12V in the second case, you have no problem.

We had an inspector at NE about 4 years ago who knew his stuff. He asked the students to make that exact measurement (frame to ground) and then asked them to explain where the voltage came from. It got them to thinking about Ohm's Law. I later used an extra 100K pot to prove to the students that there was no short.

A former alumnus from that year's team responded via private message:

...You explained it to me after the inspector drilled me for an answer... but all i can remember is induction.... maybe you can shed some light on this issue for them (and me!)...

To which I replied:

Let me pose this in the form of a question. What is the voltage potential between a fragment of Eta Carinae and Haley's comet?

Don't know the answer? Neither do I. If the systems are truly isolated, the answer is undefined. In the real world, nothing is undefined but the coupling is so weak that the voltage could be trillions of volts. The probability that the voltage is zero is, for all purposes, zero.

In our robot, the DC "coupling" due to the leakage resistance of the system is quite weak (on the order of 10s to 100s of megohms so I would expect the DC voltage to be between 0V and 12V. Note also that the act of measuring with a voltmeter which is not ideal (less than infinite ohms) will change what you read (Heisenberg is alive and well).

Induction would likely have a more measurable effect on the AC coupling. If you measure AC volts, I can almost guarantee that the voltage will not be zero either.

Now to the 100K resistor: Its effect is much stronger than the weak coupling of the leakage resistance or of inducted fields. When I connect a 100K between ground and frame, I expect near zero volts unless there is indeed a short in the system. Likewise, a 100K between frame and +12 will raise the frame to a 12V potential unless there is indeed a short in the system.

The 100K resistor “defines” the undefined system and has a impedance low enough such that the non-ideal aspects of the meter’s input impedance is reduced dramatically.

Does this make more sense?

I hope this helps others understand the phenomenon better.

[Andy Brockway]

In 2002 we ran into the same problem at the Championship. We finally traced it to the CIM motor. We can only quess that we had overheated during our Regional and caused a short in the motor somehow. This took four frantic hours of pulling the robot apart, we replaced the motor and the problem was gone!