Bearbotics Engineering Analytics Report (BEAR):

Improving the Efficiency of Power

Circuits in FRC Robots

(v1.0; 12/12/2022)

Collaborators:
David Nguyen (2023 CEO/Captain; 2022 Electrical Team Lead)
Keira Griffin (2023 Electrical Team Lead; 2022 Electrical Team Member)
Mike Hinkle (Head Coach, Electrical Mentor)
Abstract
This is the report-out of a thorough study of high-power circuits and components in FRC robots
undertaken by Bearbotics (FRC team #4068), with specific recommendations. Much of
community knowledge and practice regarding FRC electrical systems are best practices based
on a black-box approach to system design using what is best described as recipes. While there
are many benefits to this state of practice, this inquiry sought to determine if electrical system
efficiency (i.e., the fraction of battery power that is actually delivered to the motors) could be
further increased in practical ways that had minimal negative side-effects. The answer is yes.
The findings and recommendations herein are likely of interest to all FRC teams, but especially
to those seeking to maximize the performance of their robots.

Disclaimers
● We have used most of these recommendations on our own FRC robots and can confirm
they are not detrimental to robot performance.
● Nothing in this document should be construed as a review of a specific product or a
recommendation for or against its value or usefulness. We recognize that vendors have
specific objectives for features and capabilities for their products, and that those may
have some (unintended) impacts on what we’re measuring in this study. Teams should
use the information and recommendations presented here within the context of their
robot plans, goals and practices to make their own selections.
● We did share an earlier draft of this report with some (but not all) of the cited vendors.
Again, our general attitude toward the FRC vendors is respect and appreciation for all
they do to help us build better robots!
● We have no commercial or sponsorship relationship with any of the vendors that supply
parts investigated in this study.

Methods
Our methods focused on identifying and measuring excess intrinsic resistance in the
components of high-power electrical circuits on FRC robots. This resistance is measured in
milliohms (i.e., thousandths of an ohm), and in some cases fractions of milliohms. These are
extremely small values that are beyond the capabilities of conventional tools such as the digital
multimeters we all have.

Our first approach was to measure voltage drops at full current loads; for example, to take
voltage measurements on electrical components while handling current flows as high as 100A.
We found it was very difficult to regulate current flows at this scale with accuracy and stability

Bearbotics (FRC 4068): Improving the Efficiency of Power Circuits in FRC Robots (v1.0)
without investing in very-costly test equipment. And we were concerned about the safety
considerations for students engaged in the testing. Plus, the testing system would not be
portable, so we couldn’t easily extend this type of support to other teams, or publish guidelines
that enabled other teams to do this testing themselves cost-effectively.

Therefore, we decided to measure the intrinsic resistances of the devices under test directly
using a precision milliohm meter. We believe this approach is valid because with one exception
all the devices we investigated are fully passive; i.e., they do not process or modify the electrical
current flowing through them in any significant way, and therefore present a constant resistance
across all reasonable levels of electrical current. (The exception is circuit breakers, which are
highly-linear and are discussed in more detail below.)

To reduce cost and increase learning, we built our milliohm meter using Arduino-class devices
based on our own improvements to a published design. Commercial milliohm meters are
available at prices that may be reasonable to many teams, but we have not tested any of those
devices and cannot make a specific recommendation.

In general, resistance was measured for the full path or circuit through which current would flow
when the device was used in a robot electrical system. For a PDH or PDP device, the measured
resistance is for the full circuit from the main negative (-) input terminal to the corresponding
positive (+) terminal. All values are net, and exclude any connecting wires or connectors, and
inserted devices (like circuit breakers).

Whenever possible, we used 12 AWG stranded (at least 600 strands) wire with silicone
insulation for the test leads. We find this type of wire achieves an excellent electrical connection
with the latching WAGO terminals becoming more common in FRC electrical devices.

Definitions
“Intrinsic resistance” is the internal resistance of an electrical component or device. Using
common current technology, all electrical devices have some amount of resistance. Devices
used in high-power applications (like FRC electrical circuits) may have very low intrinsic
resistance, but it is still an impactful amount that should be considered. Generally, the resistance
of these power devices is measured in milliohms. This is often called “series resistance”.

“Excess intrinsic resistance” is the measured resistance of a device compared to the best case
(lowest) resistance of a comparable device (which is either the same type of device, or a
suitable alternative device). This report recommends the method of finding and minimizing the
excess intrinsic resistance of devices in FRC robot electrical circuits.

milliohm = one-thousandth of an Ohm

Bearbotics (FRC 4068): Improving the Efficiency of Power Circuits in FRC Robots (v1.0)
Summary
For the purpose of this report, we believe there are 3 general characterizations of workable FRC
robot electrical systems:
1) Reliable: can repeatedly and reliably deliver useful power to the motors in the drivetrain
and main mechanisms, despite the rigors of competition matches. This is a highly
desirable robot capability for teams that want to compete well and participate in the
Elimination rounds.
2) Average / Typical: built well using the widely-available guidelines and cookbooks
published to the FIRST community, using purposefully-selected components.
3) Optimized: this is a new level that we introduce here. This level is achieved by
deliberately selecting – and, where necessary, measuring the intrinsic resistance of – all
electrical components to attain an actual build that effectively minimizes the loss of
electrical power between the battery and the primary motors.

Our recommendations address all 3 of these levels to some extent, but are primarily focused on
practical ways for teams to move to the Optimized level.

As we built more complex, capable and powerful robots, we observed that their dependencies
on the electrical system increased significantly. The first need was to improve reliability and
ease of build, but some of the fixes we adopted there (e.g., conversion from Anderson to XT
connectors) had the dual purpose of improving efficiency also. But after we built swerve-drive
robots, we found that the average current draw of our robots throughout the entirety of a match
was approaching 100A. Peak loads could be much higher – we had reasons to believe they
could approach 150A. And indeed, given that a good FRC-spec battery can supply 200A current
for the duration of a match – although we don’t believe our robots achieve that level of demand
(yet) – it raises the important question about whether or not there are impactful but hidden
factors in typical FRC robot electrical systems that impede the transfer of all this power. Put
more simply: do robot electrical systems built per our current standards and best practices
deliver the maximum possible fraction of power from the battery to the motors? We undertook
an investigation that delivered the answer: no. We needed to change some of our selections
and improve several of our practices to minimize the loss of power in our robots, and that’s what
we report on here.

A few simple calculations illustrate the scale of the problem.

● One milliohm (0.001 ohms) seems to be a very-small and perhaps inconsequential
value. However, assuming that the supply battery can output 12.0 volts at a load of 100
amps, adding just 1 milliohm to the circuit between the battery and the output (let’s
assume a motor) reduces the voltage arriving at the motor by 0.1 volts (per Ohm’s law E
= I x R), or 0.8%, which reduces the motor output power (P) by about 1.7% (per Joule’s
law P = E2 / R). (We’ve simplified this calculation by disregarding the motor power
conversion factor, but this is a highly-linear relationship.)

Bearbotics (FRC 4068): Improving the Efficiency of Power Circuits in FRC Robots (v1.0)
 ● Our data shows that excess intrinsic resistances of 5 milliohms could be present in even
well-built robot electrical systems. At a system load of 100A, this would result in a
voltage reduction of 0.5 volts (4.2%) and a power reduction of more than 8%.
● Excess resistance of 10 milliohms is quite possible. At this level and the same system
load (100A), the voltage reduction is 1.0 volts (8.3%) with a power reduction of 16%.
● And, assuming a team’s robot design has pushed everything to the max so that 200A
peak loads are achieved, with just 10 milliohms excess resistance the voltage reduction
is 2.0 volts (17%) and the power reduction is about 30%!

In conclusion, at the electrical current levels commonly experienced in FRC robots, even very
small resistance resistance values can have significant impact on the power delivered to the
motors – and therefore overall robot performance. Therefore, the search for and removal of
excess resistance values on the order of just one milliohm can be worthwhile.

Scope
This report and its recommendations focus on the highest-power circuits in the FRC electrical
system. These are the main power trunk (extending from the battery to the power distribution
panel (PDP or PDH)), and branch circuits carrying currents of 30A or more (typical of circuits
supplying NEO or Falcon motors that drive high mechanical loads, such as chassis drive and lift
motors). These areas are shaded light blue in the simplified FRC electrical system diagram
(below). There’s limited benefit to circuits carrying 20A or less.

(drawing courtesy of stefenacepcion)

Bearbotics (FRC 4068): Improving the Efficiency of Power Circuits in FRC Robots (v1.0)
Key Findings and Recommendations
Much of Optimization can be achieved by better component selection – picking one component
option over another (e.g., selecting Anderson instead of spade connectors). Some is
accomplished by improved use of specific components (e.g., which ports on a PDH to use for
drive motors). But full Optimization also requires measurement of the intrinsic resistance across
multiple components of the same type to determine which provides the greater efficiency (lowest
resistance) so it can be used for the robot build.

There will be decisions for any team to make as they determine how to use the information and
recommendations in this document. Some of this is about the typical trade-offs of cost, time and
quality (where “quality” is determined by low intrinsic resistance in this case). And some of this is
about features vs. power efficiency. The bottom line: in many of these matters there is no right
answer for every team and every robot; so this is not an attempt to set new standards. But we
do seek to help our fellow teams to optimize their electrical system selection, design and build
practices so they can achieve better overall results against their goals. For example, our team’s
own updated selections based on this information actually deliver end-to-end circuit resistance
about 1 - 2 milliohms higher than the theoretical minimum, because we want the features and
functionality that certain parts offer (this results in a functionality boost but a performance
penalty of about 3%).

1) The Rev Robotics PDH power distribution panel

provides much-needed additional power ports that
enable the building of more-complex robots. But,
although they offer the ability to place any value of
circuit breaker in any circuit, the ports do not all offer
the same power efficiency. Ports 0, 1, 18 and 19
should be preferred to supply motor circuits with the
highest expected loads (like chassis drives and lifts).
Ports 5 - 14 should be avoided for the highest-load
motors. (In the PDH, Ports 0 and 19 have at least 1
milliohm less resistance than Ports 9 and 10.) (Similar
guidance applies to the CTRE PDP unit, although the
excess intrinsic resistance is lower in that case.)

2) Circuit breakers presented the widest range of

measured resistance values, while offering little to no
external clues regarding the actual device resistance.
In retrospect, the test values were so wide-ranging
and inconsistent that although we tested about 40
different units, we wish we had tested many more samples. We will continue to focus on
testing and characterization of circuit breakers in the future. Main findings:

Bearbotics (FRC 4068): Improving the Efficiency of Power Circuits in FRC Robots (v1.0)
 a) 120A main breakers ranged from a low of 0.4 milliohms to a high of 2.6. The
resistance reduction opportunity is as much as 2.2 milliohms. Manually cycling
the breakers did change the measured values, sometimes by as much as 0.5
milliohms, but the effects were not predictable. We recommend teams select a
unit with intrinsic resistance of 1.0 milliohms or less.
b) 40A CTRE PDP breakers averaged 5.0 milliohms, with a high of 31.1! Using our
Optimize (measure and select) method, we had to discard ~70% of our samples
but got average unit resistance down to 2.24 milliohms.
c) 40A Rev Robotics PDH (ATO) breakers were an interesting and changing story.
Our 2021 product set averaged 1.55 milliohms with a low of 1.25 and a high of
2.0; they all passed the Optimize review. But in the set we purchased in
November 2022, only half performed as well as the 2021 models – and the other
half was much worse. Applying the Optimize method, we rejected the
underperforming units (~35%) and achieved an average of 1.41 milliohms for this
multi-year mix of units.
d) Unexpectedly, across all the PDP and PDH breaker types we tested, the 30A
PDH (ATO) breakers had the most consistency and lowest average resistance of
1.21 milliohms. We had to reject less than 10% of the units to achieve this.

(Note: in all cases except for the 120A units, the breakers were new and unused, so
mechanical or electrical wear were not factors in these measurements. In all cases
units from multiple production years were tested, including the PDH breakers.)

Recommendation: there is no easy answer in this case. The only way to ensure
optimal (lowest) device resistance for circuit breakers is to measure a large set of
devices and select the best for use in the robot. We recommend that units showing
more than 2.0 milliohms resistance be rejected (that’s the standard we used in the
Optimize step for this report), and units with less than 1.5 milliohms be preferred
(with PDH breakers there’s a good chance you can find enough of these).

Speculation: since we apply current limiting on all our swerve-drive motors, we think
we can keep them below the trip thresholds of 30A PDH (ATO) plug-in breakers.
Therefore, we will experiment with using 30A breakers for our chassis drive motors
as we do robot testing for the 2023 FRC season. Clearly there are tradeoffs.

3) Connectors are what inspired the original version of this investigation. There are many
styles and degrees of convenience to choose from, and so many are used in each robot
build – so the total impact on robot build time and performance is
significant. Here are our main findings:
a) Spade connectors are not the best option. They have the
highest resistance of all the quick connector solutions; even
using excellent technique, their best resistance is 0.8 to 0.9
milliohms per connection.

Bearbotics (FRC 4068): Improving the Efficiency of Power Circuits in FRC Robots (v1.0)
 b) Butt splices can be very good. However, this depends fully
on the quality of the wire insertion and crimp. Done well, a
good old-fashioned butt splice delivers resistances as low as
0.05 milliohms!
c) WAGO lever connectors are surprisingly good. Across many
samples and styles using 12 AWG stranded wire, measured
resistance averaged as low as 0.1 milliohms per
connection! Given the high-quality connection without the
need for soldering, plus the flexibility to easily remake
connections, this is an excellent choice. But this applies
only to authentic WAGO connections (more about that in
(e), below ). And be careful about the excess wire length
required for the side-by-side style of WAGO lever nuts;
it’s very easy to add an extra 2-5” of extra wire to make
one of these connections along an otherwise-straight wire path; and with 12
AWG wire each extra inch adds 0.11 milliohms of resistance (0.8 for 10 AWG
wire). We therefore prefer the WAGO inline connectors.
d) XT connectors are no better than soldered Andersons. This was the specific
inquiry that started this whole investigation, and was quite a surprise to us. But
this applies only to Andersons with crimped-and-soldered connections; when
crimped only, Andersons do not perform as well
as XT90 connectors. Additional insights:
i) The above comments apply only to XT90
connectors. For XT60 and XT30
connectors, Andersons perform better
whether crimped or soldered.
ii) And WAGO lever connectors perform
significantly better than Anderson and XT
connectors in all cases. The lowest per-connection resistance we
measured for either 45A Andersons or XT90 connectors was 0.4
milliohms, compared to the 0.1 milliohms achieved by WAGOs.
e) Be cautious about inexpensive lever connectors. When
we say WAGO, we mean authentic WAGO parts only.
The widely-available cheaper knockoff parts have lower
costs but much higher internal resistance, averaging
0.7 milliohms per connection (to be clear, that’s for
each side of these 2-sided devices). Notes:
i) Despite their claim of supporting 12 AWG wire,
we were not able to insert high-strand silicone
12 AWG wire into these components; 14 AWG
did fit.
ii) And, to be fair, these devices perform better than Spade connectors and
are easier to make..

Bearbotics (FRC 4068): Improving the Efficiency of Power Circuits in FRC Robots (v1.0)
 4) From a power-efficiency viewpoint, which power distribution panel is better: CTRE PDP,
or Rev Robotics PDH? We waited to publish this report until we could test the 2022
version of the PDH (we received ours in November 2022), and the answer is: if you
measure and select all the related components, you may achieve a very slight edge with
the PDH. Details:
a) This consideration must include the performance of the circuit breakers too,
because they are a requirement for the final system in the robot.
b) The CTRE PDP actually has the lowest intrinsic resistance. But the Rev Robotics
circuit breakers have significantly lower best-case and average resistance,
swinging the two systems back to equal, or perhaps a slight advantage to the
Rev Robotics system. (For full disclosure, if we could use a full set of the
very-best PDP breakers that we could find, the PDP system would have a slight
advantage; but we couldn’t find enough of those top-performing breakers, even
among all the spares we had in stock. The newer PDH breakers have a
significant advantage in average and mean resistance, which gives the
advantage to the PDH system.)
c) Overall we purchased 3 of the PDH units for use and testing: 2 in 2021, and 1 in
2022. Two of those units tested well, but one of them (from the 2021 batch) had
remarkably higher intrinsic resistance (averaging 6.4 milliohms per port, vs. 3.5
milliohms for the other 2 PDP units). We still don’t have an explanation for the
poor performance of that one unit from 2021, but were very relieved to see the
2022 unit was solid – so we believe that high-resistance 2021 unit is the
exception.
d) The bottom line: there’s simply no way to optimize this highly-important
component without test and measurement. Our measured difference between
average and optimized for this one electrical system component alone is 6
milliohms, so we recommend that any team seeking the highest robot
performance invest in measurement.
e) Why do PDP and PDH units have such high intrinsic resistance (the 2-3 milliohm
resistances are much higher than just a straight piece of #12 wire would have)?
In addition to the WAGO wiring terminals and circuit-breaker mounts, there are
voltage-dropping resistors in these units to support the measurement of electrical
throughput current. The efficiency cost of this feature is estimated to be about 1
milliohm of resistance.

5) With battery leads, we investigated the relative advantage of 4 AWG wires (with SB120
connectors) vs. 6 AWG (with SB50 connectors); in both cases, we used new units fully
manufactured by Andy Mark. Using our methods and measurements, we found no
difference in intrinsic resistance. (Full disclosure: the 6 AWG leads actually measured
slightly better, but the difference was within the margin of error.)

6) Should passive electrical components be re-used? It wasn’t that long ago that re-use of
circuit breakers and battery leads was a common strategy for us to reduce build costs.
But our testing shows teams should be cautious about re-using some electrical

Bearbotics (FRC 4068): Improving the Efficiency of Power Circuits in FRC Robots (v1.0)
 components. Our general recommendation is to test components before re-using them if
the goal is high performance. Details:
a) COTS battery leads have improved a lot since 2019. We tested some new 2021
models against used leads received in FRC kickoff kits in 2018 and earlier, and
found excess resistance as high as 8.3 milliohms in the older units. We have
more confidence in re-use of battery leads produced in 2020 or later, but will
always measure them just to be sure.
b) With 120A circuit breakers (which we also have re-used a lot), we found excess
resistance as high as 2.2 milliohms in used units. Given the very-high electrical
currents this device will experience, we recommend testing and selection to
identify the best-performing unit for the robot. We recommend that the 120A
breaker on a high-performance robot have an intrinsic resistance of 1 milliohm or
less.

7) The wires between a NEO motor and the SPARK MAX controller should be kept as short
as possible. This recommendation is speculative, because we were not able to verify it
by testing; however, we are highly confident in it. These wires (between a brushless
motor and the connected brushless motor controller) are the only place in the robot
electrical system where an alternating (reversing) current occurs – and would therefore
be subject to the skin effect which increases the effective resistance of the wires. If you
need to locate a high-power motor a significant distance from the power distribution
panel, extend the wires on the DC (input) side of the controller only – not the wires
between the controller and the motor. (Of course this is not an issue for Falcon motors,
in which the wire distance between the motor and controller is extremely short and
fixed.)

Bearbotics (FRC 4068): Improving the Efficiency of Power Circuits in FRC Robots (v1.0)
Theory
Overall, this is a story of very-high electrical currents – frequently exceeding 100 amps – and
the resulting impact of very-small (but higher than necessary, or excess) electrical resistance.
The laws (Ohm’s and Joule’s) and calculations are simple, and are basic to electrical
engineering, But the impacts are over-sized and are most likely to affect teams seeking to
maximize robot performance. Therefore, we recommend a rigorous but achievable process of
identifying (through planning and measurement) and removal of excess intrinsic resistance in
FRC robot electrical systems. This process is performed on a component-by-component basis.

The overall theory is illustrated in this conceptual model:

Intrinsic resistance is the circuit component this report focuses on. All electrical devices using
current technology have some amount of intrinsic resistance; the approach we recommend is to
identify and eliminate or reduce excess intrinsic resistance. Even though shown above in one
convenient location RINT, the intrinsic resistance is actually distributed in multiple discrete
elements throughout the entire robot electrical system – including both the positive and negative
sides of each circuit. For the purposes of calculation though, lumping all the intrinsic resistance
into one entity (as shown in the diagram) simplifies the approach.

Relevant calculations to demonstrate the scale of the opportunity are shown in the side text
(and in the “Summary” section). Our analysis shows that simply going with a typical “average”
approach to building a robot electrical system can result in excess intrinsic resistance of 5 - 10

Bearbotics (FRC 4068): Improving the Efficiency of Power Circuits in FRC Robots (v1.0)
milliohms – and perhaps even more. This amount of resistance will reduce the electrical power
available to the motors by more than 15%, which is a significant impact on robots designed for
high performance in the first place.

The motors and associated controllers are outside the scope of this analysis, since they have
already been thoroughly discussed in other documents and threads. The purpose of the
recommendations in this report is to maximize the power delivered to the motor systems.

The Role of Batteries

Using the CTRE Battery Beak, the lowest intrinsic resistance we’ve ever measured for a battery
+ cable connection is 14 milliohms. We will use batteries with measurements up to 19 milliohms
for competition matches. This resistance is important because it’s the largest single chunk of
distributed intrinsic resistance in the robot electrical system, so it has real impact. Reducing the
battery intrinsic resistance by 5 milliohms will benefit overall robot performance as much as
eliminating 5 milliohms of excess intrinsic resistance across several other electrical components.
(Insights into how to reduce battery intrinsic resistance are published elsewhere.) For our
analytical calculations, we assume a somewhat-ideal value of 15 milliohms for the resistance of
the battery plus the directly-connected leads.

The Role of Wire

Wire has intrinsic resistance too. Overall resistance of wire part of wiring can be reduced by:
● Using the largest (i.e., lowest AWG number) practical wire for the circuit. But due to size,
weight and cost considerations, we never use wire larger than #10 AWG wire for
high-power motor circuits and rarely use larger than #12 for this.
● Keep wiring lengths as short as possible. One advantage of WAGO connectors is they
allow easy quick replacement of wire segments if any were built too short.
● The overall reduction opportunity here is up to 5 milliohms in each motor circuit.

Effect of Supply Voltage on Motor Output Power

An important question is: does a
decrease in the power supply voltage
to the motor controller actually result in
a decrease in motor output power? Or
is there voltage compensation
performed by the motor controller?
Complex high-power testing could give
us the answer – and fortunately that’s
already been done by VEX and
published in its DC Motor Testing data.
The adjacent chart shows that data for
the NEO plus SPARK MAX combo,
re-sliced to portray the relationship

Bearbotics (FRC 4068): Improving the Efficiency of Power Circuits in FRC Robots (v1.0)
between supply voltage and motor output power for a constant motor speed (about 1000 RPM –
which is a good speed for drive motors if the robot is trying to do some heavy work pushing or
pulling on the field). The data clearly shows that a reduction in supply voltage does result in a
significant reduction in motor output power (which would result in reduced robot chassis speed,
for example). There’s no internal compensation in the controller to overcome that. The robot
control program could impose some compensating controls to attempt to maintain mechanical
power and therefore speed, but those have limited range because they would demand more
electrical current (power) to each motor, which would further reduce the supply voltage and the
potential motor output power.

Does this really make a difference?

The big gap in this analysis is comparative performance data. We have not had the time and
resources to do a complete A/B test, comparing the performance of an average robot build to an
optimized one in which all other aspects are identical. So we continue to rely primarily on the
analyses and calculations presented above to demonstrate the opportunity.

In theoretical comparison, we believe that an optimized robot build will have these advantages
compared to an average one (as defined in the “Summary” section):
- Greater acceleration
- Great pushing power
- Ability to support more motors and mechanisms
- Ability to operate mechanisms and manipulators at higher speeds and power levels while
the robot is moving on the field
- Less inclination to experience electrical brown-outs
- Less inclination to experience movement hesitation and other anomalies

Bearbotics (FRC 4068): Improving the Efficiency of Power Circuits in FRC Robots (v1.0)
Estimate of Voltage and Power Impacts
This table summarizes our estimates of the potential positive impacts of our recommendations.
We’ve used the findings and guidance provided in other sections of this document to compile
three potential scenarios for FRC robot electrical system outcomes, including the recommended
Optimized one. All resistance values are in milliohms.

Intrinsic Resistance of Electrical System Scenarios

Electrical System
Component Worst Average Optimized
Measurement and
Scenario Poor choices and/or bad Typical results. Good
selection of all
Description luck practices.
components
Battery + leads 22 18 15
120A CB 2.58 1.38 0.34
CB-PDH Jumper 1 0.7 0.5
PDH 7.84 3.45 2.62
PDH CB 5.03 2.25 1.12
Branch wiring 8.8 5.28 3.08
Connectors (Branch) 1.89 0.81 0.53
Connectors (Motor) 1.89 0.81 0.53

Total Resistance 51.0 32.7 23.7

Excess Resistance 27.3 9.0 0

Power loss @ 50A Load Compared to Optimized

Voltage Drop 1.37 0.45
Power loss 20.5% 7.0%

Power loss @ 100A Load Compared to Optimized

Voltage Drop 2.73 0.90
Power loss 38.6% 13.7%

Notes:
- Branch wiring assumes use of #12 AWG wiring throughout. The differences are due to
longer wire lengths from poor component locations and less attention to wire routing.
- In some areas we’ve been kind to the “Average” scenario. For example, if we had not
measured components for our 2022 season robot build, the PDH values could have
easily been in the “Worst” category, even with all our other good practices.

Bearbotics (FRC 4068): Improving the Efficiency of Power Circuits in FRC Robots (v1.0)
Next Steps
This will be a continuing story. We plan to test and measure important new devices as they
become available. We will also look for existing alternatives to some devices and determine if
they offer advantages regarding power efficiency.

We will improve our understanding and characterization of circuit breakers. These are unique
devices, with very small contact surface areas and in many cases very low force maintaining
those contacts. We want to better understand the effects of aging, and mechanical and electrical
wear and tear on the intrinsic resistance of these devices.

Bearbotics (FRC 4068): Improving the Efficiency of Power Circuits in FRC Robots (v1.0)
Details of Findings / Data Table

Component: Best Potential

(Unit Under Test) (Lowest) Average Worst Resistance Comments and
(note: multiple tests were Resistance Resistance Resistance Reduction Resistance Reduction
performed on each unit) (milliOhms) (milliOhms) (milliOhms) (milliOhms) Method
Main Circuit Breakers
120 Amp Circuit Breaker, 0.34 0.43 0.60 0 This is the standard of
Unused best.Note: we only found
and tested 1 in this
condition.
120 Amp Circuit Breaker, 0.85 1.36 0.51 Test units for resistance
Some Use before use
120 Amp Circuit Breaker, 0.65 1.38 2.58 2.24 Test units for resistance
Moderate Use before use
120 Amp Circuit Breaker, 1.98 2.12 2.28 1.94 Test units for resistance
Extensive Use before use
PDP Circuit Breakers
30 Amp Breaker (V1 / for 2.74 17.04 75.2 14.30 Test units for resistance
PDP), Typical before use
30 Amp Breaker (V1 / for 2.74 2.77 2.81 14.26 Rejected low-performing
PDP), Optimized units (~70%)
40 Amp Breaker (V1 / for 1.66 5.02 31.15 3.36 Test units for resistance
PDP), Typical before use
40 Amp Breaker (V1 / for 1.71 2.24 2.98 2.77 Rejected low-performing
PDP), Optimized units (33%)
PDH (ATO) Circuit Breakers
30 Amp Breaker (V2 / for 1.12 1.30 2.23 0.93 Test units for resistance
PDH), Typical before use
30 Amp Breaker (V2 / for 1.12 1.21 1.35 0.15 Rejected low-performing
PDH), Optimized units (~10%)
40 Amp Breaker (V2 / for 1.19 2.25 5.03 2.78 Test units for resistance
PDH), Typical before use
40 Amp Breaker (V2 / for 1.19 1.41 1.96 0.55 Rejected low-performing
PDH), Optimized units (~35%)
Battery Leads
Battery Leads (6 AWG SB 2.18 2.21 This was best or equal to
50A), New, Unused best in our tests
Battery Leads (4 AWG SB 2.22 2.23 No better than 6 AWG

Bearbotics (FRC 4068): Improving the Efficiency of Power Circuits in FRC Robots (v1.0)
 120A), New, Unused SB50 units
Battery Leads (6 AWG SB 3.79 6.61 10.48 8.3 Test before any re-use
50A), Used
Power Connectors
Butt Splice (AWG 10-12) 0.05 Very good, but not
changeable
Spade Connector (AWG 0.86 0.94 Use a different connector
10-12) type
30 Amp Anderson (Crimped 0.59 0.61 0.40 Crimp AND solder wire
Only) connection
45 Amp Anderson (Crimped 0.46 0.46 0.07 Crimp AND solder wire
Only) connection
30 Amp Anderson (Crimped 0.21 0.21 Good
and Soldered)
45 Amp Anderson (Crimped 0.40 0.40 Good
and Soldered)
XT30 (18 AWG Wire) 0.99 1.02 Limit to 20A max load
XT60 0.79 0.80 0.40 Use XT90 or 45A
Anderson instead
XT90 0.40 0.40 Good
WAGO 221 Inline Splicing 0.18 0.27 0.30 Very good
Connector
WAGO 221-412 Lever Nut 0.01 0.04 0.08 This is the standard of
best
Power Distribution Panels (full circuit resistance,
per port)
PDP V2 (CTRE) 2.05 2.35 2.48
PDH V3 (Rev Robotics), 5.46 6.41 7.84 3.77 Test PDH units and use
Unit #1, 2021 release the best one
PDH V3 (Rev Robotics), 2.96 3.45 4.12 1.16 Test PDH units and use
Unit #2, 2021 release the best one
PDH V3 (Rev Robotics), 2.62 3.33 4.07 1.45 Test PDH units and use
Unit #3, 2022 release the best one

Worse than expected

Better than expected, or
the best

Bearbotics (FRC 4068): Improving the Efficiency of Power Circuits in FRC Robots (v1.0)