Recently I used 18 gauge wire wagoed to a neo, and apparently I should have used 14 or some other gauge. My question is how do I tell what wire guage to use, and for what components or voltages. Thanks!
For that specific case, read the game manual. R622
7 Likes
Besides the correct answers above, what if you weren’t doing this for FIRST? There are some general rules dictated by the physical properties of the material in the wires. But as you can see by the chart above, the more amperage a wire can carry, the larger diameter it needs (gauges go smaller in number for larger wires).
This is another chart I found as well, but for FIRST follow the specific rules in the game manual! This chart is based around 120 volts I believe, which is why they show a limit of 9 amps for 12awg cable. This is not the same with 12v DC
If you didn’t know how much amperage it used, look on the component you are connecting to. The AWG is sometimes written on the wires insulation, or you can measure the original and match it. Not matching the insulation thickness, but the diameter of the conductive material inside the insulation.
6 Likes
Then we should point out that aluminum wire or copper clad aluminum (CCA) wire has its own current capacity chart since you need larger aluminum wire than copper for the same current. Al and CCA wire are illegal in FRC (for power circuits).
1 Like
I do not like this particular chart as an EE. TLDR on OP’s question of how to choose a wire size at the bottom of this post.
Where or based on what are they coming up with these arbitrary current capacities?
9.3A on 12AWG is for 99.99% of applications overly conservative. The NEC even with its factors of safety allow 20A on 12awg for a considerable distance from the source (housing and commercial building runs).
History/How many “wire ampacity” charts are made unless they otherwise specify:
Regarding why ampacities in many tables and charts you find online are different than what is specified for FIRST is not actually because of the voltage (120v AC vs 12v DC), but rather a combination of the application itself (load current), circuit protection provided, wire temperature, and voltage drop to the load.
Many charts you find are based on the North American, National Electric Code (NEC). This is a set of rules for the trades that simplify the complex problem that is wire selection for a given application. This way a house or commercial building can be built without needing an Electrical Engineers sign off on a bunch of math - as long as it was built to local code (which generally adopts some version of the NEC). So as long as you stay in the NEC guidelines and the manufacturers installation instructions, the building is not at adverse risk of having an electrical fault cause a fire or building damage.
Commonly you will see 14awg for 15A rated circuits below a certain length, 12awg for 20A circuits or 15A circuits farther from the circuit breaker box, etc…
These NEC ratings also factor in a margin of safety. All that put together results in conservative wire selection for most charts.
For automotive, individual appliances, off-road, boating, anything other than construction you will find the same load (# of amps), using smaller gauge wire than the NEC recommends. These applications are using other charts and rules regarding voltage drop to the load (1% or 10%), doing the math on the heat generated in the wire and individual wire temperature, etc…to properly size the selected wire for a given load (going into detail and doing the math for each circuit). One of the outputs of this is the table that guides FRC wire selection every year and why it remains mostly unchanged every year, because wire selection is actually driven by your load, and more importantly the circuit protection provided. In the end it’s all a thermal problem, make sure your wire is protected by a fuseing element or circuit breaker that will trip before the wire is damaged is a commonly followed boundary condition.
I have utilized this chart in the past. I have not checked every value - but a spot check of the values I was interested in against outputs from other calculations matched within reason so I have trusted it thusfar.
If your application is like most of the FRC robots I have seen with wires not so well cable managed and dangling in “free air” I would consider the value for “chassis wiring” to be closer to reason. For applications where the wires are tightly bundled in a cable conduit, raceway or other sheathing you should de-rate from this value significantly.
TLDR how to select a wire size:
For FIRST: simply abide by the guidelines and table in the manual and there wont be any hazards with the wire or circuit protection selection. Teams may have other motives like maximizing delivered power to the load and choose to size up wire (10AWG or 8AWG for motors, 4AWG or 2AWG for battery leads and the main power run), which minimized voltage drop and boost power (P=V*I or P=V^2/R or P=I^2R), with the downside of it costing them more and they need to come up with solutions for terminating the wire that other teams dont need to think about using the minimum’s from the table.
Overall General application - When selecting a wire gauge for a project that does not need to abide by the NEC or other local code, I would
- First consider the protection device and size it appropriately (Littlefuse and Eaton (Cooper Bussman) have a few whitepapers on the topic of properly sizing a circuit protection device like a fuse or circuit breaker)
- Select an initial wire size based on the chart previously listed that exceeds the rating of the circuit protection device
- Calculate the voltage drop to my load and determine if that is acceptable for my application
- Re-size as necessary for #3
- If this is a volume project and costs are more important you need to go farther and calculate the wire temperature under load and during overload, and consider downsizing the wire to a value that can withstand the temperatures during overload before the circuit protection device trips, and cool down before another fault needs to be survived.
8 Likes
I agree with you and applaud your analysis as worthy of a real electrical engineer.
Adding a little based on my own EE experience*:
Applications like FRC involve relative short duration transient loading of conductors, resulting in temperatures that swing over a range of several tens of Celsius degrees during a typical match, due to intermittent operation of devices (mainly motors) that draw significant current from the battery.
Neglecting heat transport (a reasonable, conservative assumption) the rate at which copper conductor temperature rises is proportional to the square of the density of current (Amperes per cross-section area). The equations are summarized below.
Note that when heat transport (cooling and heatsinking) is considered, temperature rise is more accurately described using the usual first order solution based on a thermal time constant derived from heat capacity and wall heat transfer coefficient at thermal interfaces; here I will neglect that effect, focusing instead on the initial rate of rise. This approach is useful for sizing electric power conversion devices based on their peakiest intermittent load conditions.
Current density in copper conductors of typical electrical energy conversion devices ranges from very low values in continuous operation (e.g., refrigerator compressors) to very high values in highly intermittent operation (e.g., motors used to start internal combustion engines in cars and trucks). SI units for current density are Amperes per square millimeter; many years ago I adopted a convention that is popular in the UK and European countries, calling this are unit a “quadratic” millimeter and abbreviating it “qmm”. This is convenient because it doesn’t require formatting exponents in fractions: units for current density (J) are thus A/qmm. When the load is AC, the current is assumed to a an effective (root-mean-squared) quantity.
So, current density in copper coils of energy conversions devices can be anywhere from about 4 A/qmm when the load is steady and little or no forced cooling is available, to more than 90 A/qmm in an automotive starter motor. It is rare to find applications so extreme, though; most of the ones I have worked on have had peak current density less that 35 A/qmm.
For that range, the rate of temperature rise (Kelvins per second, starting from 20C) can be plotted vs. current density:
A potentially helpful illustration can be provided by looking at temperature rise during fifteen seconds (or less) while a copper conductor of a given wire size (20 AWG to 6 AWG) carries 60 Amperes:
From the graph above, we can clearly see a reason for requiring 12 AWG as the minimum wire size for carrying motor current in FRC robots. 10 AWG is even better, and many teams use it.
As Big Al and many others have noted previously on CD, there’s another good reason to use 10 AWG for motor wiring, especially long runs: voltage drops in the wiring limits power available to the motor, so fatter wire can give a performance advantage, as well as keeping things running cooler.
*I have been a practicing electrical engineer for more than forty years.
9 Likes
Another thing I like to remind new folks to this space is that the current limit / breaker is there to protect the insulation, not the actual copper bits. The failure mode for exceeding the current limit over a sufficient period of time is usually the insulation melting, which can lead to additional shorting pretty quickly and a cascading failure, or a “walking wounded” weak place in the insulation that can fail later even at lower current levels.
Things like the number of wires in the bundle, bundle spacing, whether it’s in a surrounding medium (eg conduit), etc, all have an effect on how quickly generated resistive energy can be dissipated away and thus the temperature the insulation will see. In addition to the wire itself, common sources of heat are wire interconnections, due to both wire-contact resistance and contact-contact resistance.
7 Likes