4. ■ The product development process
■ The importance of planning
■ Ask the right questions
■ Build the right prototype
■ Schedule
■ Test and validation
■ When to ask an expert
■ Case studies
■ Mechanical and electrical prototyping details and resources
WE’LL COVER
4 | S Y N A P S E C ON F I N D E N T I AL
6. THE IMPORTANCE OF GOOD PLANNING
6 | S Y N A P S E C ON F I N D E N T I AL
7. What are you trying to learn?
■ User experience and product definition
• “Will this design fit 95% of my target population”?
• “Is it intuitive for users to walk up to my product
and wave to wake it up?”
■ Technical and engineering challenges
• “How do I make a bluetooth antenna work underwater?”
• “Will my design generate adequate pull force”?
■ Product architecture tradeoffs
• “How do users respond to the product size that results
from the minimum viable battery life?”
Who are you trying to impress?
■ Yourself, a partner, an investor, a CM, etc?
ASK THE RIGHT QUESTIONS
7 | S Y N A P S E C ON F I N D E N T I AL
8. Now that you have your questions, WHAT TYPE OF PROTOTYPES SHOULD YOU MAKE?
Looks like: Production intent size and appearance, non-functional
■ Great for user interaction questions, and impressing key stakeholders
■ Lower cost and lead time than functional prototypes
Works like: Explores some functional aspect of the design
■ Great for for addressing technical risk regarding functionality or performance
■ Saves resources by not worrying about size, appearance
Looks like/Works like: Captures both function and appearance of the intended
design. May be created with a different method and scale from final design.
■ Closer to a real product - answers form, interaction, and function questions
■ Generally higher cost, longer lead time than looks like or works like
MAKE THE RIGHT PROTOTYPES
8 | S Y N A P S E C ON F I N D E N T I AL
9. ■ Don’t forget about testing! Plan to spend a substantial
amount of your prototyping time and budget on testing
■ Testing types
• Reliability - drop testing, water ingress, thermal,
chemical compatibility, etc.
• User interaction
• Regulatory
• Performance variation over multiple units
■ Think through what can be tested on each prototype
• Some types of testing, particularly reliability, require representative materials and construction
• Can you perform multiple tests on the same unit if you go from least to most destructive to
keep build quantities and costs low?
■ Build your schedule so that test results from one prototype inform the next
generation prototype
• Don’t ignore ambiguous or negative test results and charge ahead!
TEST & VALIDATION
9 | S Y N A P S E C ON F I N D E N T I AL
12. Ask for help at key decision points
■ Once you have a prototyping plan, have a plan review
• Did you choose an effective, efficient approach based on the challenges you anticipate
and who you’re trying to impress?
• Are there major design or technology risks that your prototype doesn’t mitigate?
• Have you chosen appropriate materials, processes, and components?
■ Before releasing custom PCBs or tooling, have a detailed technical review
• Custom parts are expensive and have long lead times
• It’s easy to make silly mistakes
■ Part on PCB has wrong package size or pinout
■ Incorrect connector orientation
■ Overconstrained component interfaces
■ Once you’ve processed the test data, have a results review
• If your results are ambiguous, or if there are multiple paths forward, this is a good time to
consult an expert on go-forward strategy
EXPERT REVIEW & FEEDBACK LOOP
1 2 | S Y N A P S E C ON F I N D E N T I AL
13. 1 3 | S Y N A P S E C ON F I N D E N T I AL
CASE STUDY
49ERS STADIUM ENTRY
CODE NAME | KEZAR
14. 49ERS STADIUM ENTRY (KEZAR)
P R O O F O F C O N C EP T P R O T O T Y P E
What we were trying to learn
■ How do users interact with a stadium entry
system?
Who we were trying to impress
■ Key stakeholders in 49ers org
1 4 | S Y N A P S E C ON F I N D E N T I AL
CATEGORY Looks like
TEST &
VALIDATION
User interaction testing
SCHEDULE
5 weeks design & build,
4 weeks test
15. 49ERS STADIUM ENTRY (KEZAR)
1 5 | S Y N A P S E C ON F I N D E N T I AL
P R O O F O F C O N C EP T P R O T O T Y P E
16. 49ERS STADIUM ENTRY (KEZAR)
A L P H A P R O T O T Y P E
What we were trying to learn
■ Is this a solid product design that can pass
environmental testing?
Who we were trying to impress
■ Ourselves, the CM
1 6 | S Y N A P S E C ON F I N D E N T I AL
CATEGORY Looks like / works like
TEST &
VALIDATION
Environmental, user
interaction
SCHEDULE
12 weeks design & build,
8 weeks test
17. 49ERS STADIUM ENTRY (KEZAR)
1 7 | S Y N A P S E C ON F I N D E N T I AL
P O C V S . A L P H A B O A R D S
18. 1 8 | S Y N A P S E C ON F I N D E N T I AL
CASE STUDY
SONICARE AIRFLOSS
CODE NAME | CHINOOK
19. SONICARE AIRFLOSS (CHINOOK)
P R O O F O F C O N C EP T P R O T O T Y P E ( B R E A D B O A R D )
What we were trying to learn
■ Does our pump system design work?
Who we were trying to impress
■ Philips technical team
1 9 | S Y N A P S E C ON F I N D E N T I AL
CATEGORY Works like
TEST &
VALIDATION
Functional test
SCHEDULE
8 weeks for design, 1
week machining, 1 week
assembly, 1 week of test
20. SONICARE AIRFLOSS (CHINOOK)
2 N D P R O O F O F C O N C EP T P R O T O T Y P E ( B R E A D B O A R D 2 )
What we were trying to learn
■ Does our overall system work?
(Gear train, plunger, mixing chamber, pump, seals)
■ Is our volumetric-similar design viable?
Who we were trying to impress
■ Philips technical team
■ Ourselves. Does this system deliver the energy needed?
2 0 | S Y N A P S E C ON F I N D E N T I AL
CATEGORY Works like, size similar
TEST &
VALIDATION
Functional test
SCHEDULE
16 weeks for design, 3 weeks machining,
1 week assembly, 2 weeks of test
21. SONICARE AIRFLOSS (CHINOOK)
A L P H A P R O T O T Y P E
What we were trying to learn
■ Does our housing seal design work?
■ Do our miniaturization details work?
■ What happens in drop testing?
Who we were trying to impress
■ The client (Philips), both technicaland executive
2 1 | S Y N A P S E C ON F I N D E N T I AL
CATEGORY Works like, looks like
TEST &
VALIDATION
Extensive, special emphasis on drop
SCHEDULE
12 weeks for design, 10 weeks for tools,
2 weeks for build, 3 weeks for test
23. 2 3 | S Y N A P S E C ON F I N D E N T I AL
CASE STUDY
SPORTWATCH GPS
CODE NAME | CONSTANTIN
24. NIKE+ SPORTWATCH GPS
P R O O F O F C O N C EP T P R O T O T Y P E S
What we were trying to learn
■ GPS performance, kickstart FW development
Who we were trying to impress
■ Technical stakeholders
2 4 | S Y N A P S E C ON F I N D E N T I AL
CATEGORY Works like
TEST &
VALIDATION
GPS Performance,
display & graphics
SCHEDULE
6 weeks design, 2
weeks fab, used by
FW team for 12
months
25. NIKE+ SPORTWATCH GPS
P R O O F O F C O N C EP T P R O T O T Y P E S
What we were trying to learn
■ How does the proposed form
factor fit our target user group?
Who we were trying to impress
■ ID and design stakeholders
2 5 | S Y N A P S E C ON F I N D E N T I AL
CATEGORY Looks like
TEST &
VALIDATION
User fit study
SCHEDULE
2 weeks design, 2 weeks
fab
26. NIKE+ SPORTWATCH GPS
P R O O F O F C O N C EP T P R O T O T Y P E S
What we were trying to learn
■ Can an FPC survive in a watch band up to our pull
force spec?
Who we were trying to impress
■ Ourselves
2 6 | S Y N A P S E C ON F I N D E N T I AL
CATEGORY Works like
TEST &
VALIDATION
Bend & pull testing
SCHEDULE 2 days build & test
27. NIKE+ SPORTWATCH GPS
A L P H A P R O T O T Y P E S
What we were trying to learn
■ Is this product ready to build at a CM?
■ Does the GPS antenna work in form factor?
Who we were trying to impress
■ Nike (a design driven company)
■ TomTom (a GPS performance driven company)
2 7 | S Y N A P S E C ON F I N D E N T I AL
CATEGORY Looks like/works like
TEST &
VALIDATION
GPS performance,
environmental
SCHEDULE
12 weeks design, 3
weeks fab, 6 weeks test
29. MECHANICAL PROTOTYPING | POSSIBILITIES & PURPOSES
2 9 | S Y N A P S E C ON F I N D E N T I AL
TIME COST QUANTITY HOW REPRESENTATIVE? COMMON PURPOSES
RAPID
PROTOTYPING
(Additive
manufacturing)
Overnight-3 days Cheap
$100s
<10 Not very. Critical to consider the
purpose of the proto, because there
are significant tradeoffs between
processes
- Fit and interface checks
- Form studies
- User experience testing
- Appearance models (if finished)
- Functional testing
MACHINING
1-2 weeks Highly variable <~30 Quite. Available in a range of
materials (metals, plastics, foams)
- Tight tolerance requirements
- Fit and interface checks
- Functional testing
- Performance testing
- Reliability testing
URETHANE
CASTING
1-2 weeks Fairly cheap
$500 - $3k
<50 Very. Great for middle quantities,
with quick timelines. High quality
finish is possible
- Form studies
- Appearance models (if finished)
- Functional testing
- Performance testing
INJECTION
MOLDING
4-15 weeks Expensive
>=$5k
>100 1:1. Can use mass production
materials, so appropriate for
mechanical testing and cosmetic
evaluation
- Fit and interface checks
- User experience testing
- Cosmetic models
- Functional testing
- Performance testing
- Reliability testing
30. ■ Resources for Rapid Prototyping:
• Proto Labs: “Prototyping Process, Choosing the best process for your project”
• Quickparts: “Process Comparison Chart”
■ Fabrication Partners:
MECHANICAL PROTOTYPING | RESOURCES & PARTNERS
3 0 | S Y N A P S E C ON F I N D E N T I AL
• Additive Manufacturing and Cast Urethanes:
■ Fathom (Seattle, WA & Oakland, CA)
■ Fictiv (SF, CA)
■ ProtoCafe (Redwood City, CA)
■ 3DS Quickparts
• CNC Machining:
■ Concept Reality (Vancouver, WA)
■ Made3D (Kirkland, WA)
■ Livewire Prototyping (Vancouver, WA)
• Prototype Injection Molding:
■ Model Solution (S. Korea)
■ AIMMCO (Woodland, WA)
■ Fathom (Seattle, WA & Oakland, CA)
• Sheet Metal:
■ Form Factor Design (Boise, ID)
■ Peridot Corp (Pleasanton, CA)
31. ■ Hobbyist boards from Sparkfun, Adafruit
• Great for quick proof of concept investigation
■ Evaluation boards from semiconductor manufacturers
• Great for getting your firmware team up and running with production intent parts while
you work on custom PCB design
■ Modules
• Gumstix, Variscite, Phytec for embedded computing platforms
• Silicon Labs, for WiFi and Bluetooth modules
■ Radio modules often come pre-certified!
• Can be used for prototyping or in low to medium volume production
■ Semi-customized boards
• Gumstix Gepetto
■ Batteries
• Powerstream for lithium polymer
ELECTRONICS PROTOTYPING | OFF THE SHELF COMPONENTS
3 1 | S Y N A P S E C ON F I N D E N T I AL
32. ■ Custom Rigid PCBs through Batch Service
• Assemble multiple PCBs onto a single panel
• Fixed quantity of prototypes (e.g. 3), limited to 2- and 4-layer designs
• Longer turn times (2-3 weeks)
• Vendors: OSH Park
■ Custom Rigid PCBs on Dedicated Run
• Panels fabricated with multiple instances of your design
• Arbitrary number of layers, quantities, higher precision, higher cost
• 2-5 day turn time
• Examples: Prototron, APCT
■ Custom Flexible PCBs on Dedicated Run
• Longer turn times (1-2 weeks) and higher cost (~2x) than rigid PCBs
• Examples: Streamline, Cordova
ELECTRONICS PROTOTYPING | CUSTOM PCB FABRICATION
3 2 | S Y N A P S E C ON F I N D E N T I AL
33. ■ Do-It-Yourself
• Design in components with suitable packages (e.g. >0402, no BGAs)
• Can reflow and rework using hot air gun for fine pitch
■ Kitted Order with Assembly House
• Parts may be hand placed for low quantities, automated placement for higher volumes
• You supply PCBs and components
• Finished boards are inspected for quality
• Examples: PCA, Out of the Box Manufacturing, Screaming Circuits
■ Turnkey Order with Assembly House
• Assembly house will order components and PCBs and deliver finished PCBAs
• Some fabrication vendors will provide an all-in-one service
• Examples: Out of the Box Manufacturing, Schippers and Crew, Screaming Circuits
ELECTRONICS PROTOTYPING | CUSTOM PCB ASSEMBLY
3 3 | S Y N A P S E C ON F I N D E N T I AL