Manga forseningar i embedded-hardwareprojekt borjar inte i firmware. De borjar nar teamet forsoker pressa in for manga grannsnitt, for hog densitet och for manga mekaniska begransningar i en konventionell stackup som redan ligger nara sin grans.
I industriella gateway-enheter, styrmoduler och kompakt kommunikationsutrustning kommer brytpunkten ofta nar 0.5 mm BGA, DDR, radio, shielding och teta connector dyker upp samtidigt. Da ar HDI inte langre lyx, utan ett praktiskt satt att undvika ytterligare en layout-omgang och extra EVT-forsening.
Why HDI PCB Matters
HDI ar motiverat nar elektrisk densitet, mekaniskt format och tillforlitlighetsmal kolliderar samtidigt. Om ett standardkort bara fungerar genom langre routning, for manga layer-byten eller framtvingad flytt av connector, bor HDI utvarderas pa riktigt.
| Product type | Typical HDI trigger | Common stackup starting point | Main sourcing risk |
|---|---|---|---|
| Embedded SOM carrier board | 0.5 mm BGA, DDR routing, limited outline | 6L or 8L with 1-N-1 microvia | Escapes work in prototype but yield drops in volume |
| Industrial gateway | Ethernet, CAN, RS-485, wireless module, isolated power | 6L with selective microvia | EMI and creepage constraints compete for space |
| Compact HMI controller | Display connector density, processor + PMIC crowding | 6L HDI | Assembly warpage and rework difficulty |
| Radio or telecom module | Controlled impedance, shielding, dense RF + digital coexistence | 6L or 8L HDI | Impedance drift and stackup inconsistency |
| Edge AI or vision board | LPDDR, CSI/DSI, multiple regulators, thermal crowding | 8L HDI | Prototype passes, mass production gets copper balance issues |
| Rugged embedded I/O module | Small form factor plus harsh-environment test margins | 4L or 6L with microvia | Buyer under-specifies test plan and documentation |
"The expensive mistake is not choosing HDI too early. The expensive mistake is staying with a conventional stackup one revision too long, then paying for a rushed redesign after the enclosure, cable set, and firmware architecture are already frozen."
— Hommer Zhao, Engineering Director at FlexiPCB
Embedded Systems vs Communication Equipment
For embedded-kort ligger smartan ofta i integrationen. For kommunikationskort ligger den oftare i marginalen: impedance, return path, shielding, loss och upprepbarhet mellan loter. Samma microvia loser alltsa olika risker beroende pa produkt.
See our HDI flex PCB service page, impedance control guide, and flex PCB prototype guide for supporting detail.
Stackup, Cost, and Lead Time
Det racker inte att bara be om “ett HDI-kort”. Det viktiga ar att valja ratt HDI-niva. Ett 6L eller 8L 1-N-1 racker for manga verkliga konstruktioner. Ett 2-N-2 eller filled via-in-pad bor bara anvandas nar routingbehovet verkligen visar att det kravs.
| HDI build option | Typical use case | Relative fabrication cost | Relative lead time | Procurement comment |
|---|---|---|---|---|
| 4L with selective microvia | Compact industrial controller | 1.2x-1.5x | +2-4 days | Good first HDI step when density is moderate |
| 6L 1-N-1 HDI | Embedded compute, gateway, HMI | 1.5x-2.2x | +4-7 days | Most common balance of density and manufacturability |
| 8L 1-N-1 HDI | Dense processor plus memory plus comms | 2.0x-3.0x | +5-10 days | Strong option when routing density is real, not speculative |
| 8L 2-N-2 HDI | Telecom, RF-digital mixed boards, high escape demand | 2.8x-4.0x | +8-14 days | Only justify when layout proof shows 1-N-1 is insufficient |
| Via-in-pad + filled microvia | Ultra-dense BGA, shortest path, thermal pad escape | 3.0x-4.5x | +8-14 days | Excellent technically, expensive if overused |
"A buyer can save 20% on bare board price and still lose the program if the chosen stackup adds one more prototype loop, two more weeks of validation, and a redesign of the shielding or connector geometry."
— Hommer Zhao, Engineering Director at FlexiPCB
RFQ Checklist
En anvandbar offert kommer inte av att bara skicka Gerber. Den kommer nar engineering-intentionen skickas med: outline, kritiska package, stackup-mal, volymer, impedance-krav och faktisk driftmiljo.
- board outline and mechanical drawing
- Gerber or ODB++ data plus drill files
- BOM or at minimum the key fine-pitch packages, connectors, and RF parts
- quantity split: prototype quantity, pilot run, and annual demand
- operating environment, service life, and target lead time
- compliance target such as RoHS, UL, or customer specification
Prototype vs Production Risk
Det forsta HDI prototype visar bara att kortet kan byggas en gang. Det visar inte att flatness, via filling, impedance och monteringsprestanda forblir stabila i volymproduktion.
"If you want prototype results to predict mass production, the fabricator must know your intended production volume, test level, and qualification target at the quotation stage. Otherwise the prototype is optimized for speed, while production is optimized for repeatability, and the two do not match."
— Hommer Zhao, Engineering Director at FlexiPCB
Review assembly impact together with your flex assembly strategy and detailed routing constraints such as those in our component placement guide.
Qualification and Testing
Definiera redan i RFQ vilka bevis som kravs: impedance coupon, microsection, plating quality, traceability, bekraftelse av surface finish och vid behov environmental testing. For tuff industriell miljo maste detta skrivas in fran start.
Use IPC, embedded systems, and telecommunications equipment references as part of the supplier review discussion.
FAQ
Nar bor ett embedded-kort ga fran vanlig PCB till HDI?
Nar BGA escape, DDR fan-out, teta connector eller enclosure-granser tvingar fram kompromisser i signal, EMC eller manufacturability. Om ett 6-layer-kort bara fungerar med for manga omvagar ar det dags att titta pa 1-N-1.
Racker 1-N-1 for de flesta kommunikationsprodukter?
For manga gateway-enheter, controllers och kompakta communication modules, ja. Ett 6L eller 8L 1-N-1 ger ofta bast balans mellan densitet, kostnad och lead time. Tyngre RF-designer kraver mer validering.
Vad ska en inkopare inkludera i en RFQ for HDI PCB?
Drawing, Gerber eller ODB++, BOM eller lista over kritiska package, volymer, target lead time, environment, impedance target och compliance target. Utan detta kan leverantoren ge ett pris, men inte en stark rekommendation.
Varfor kan ett HDI prototype fungera medan produktionen far problem?
Darfor att prototype ofta optimeras for hastighet, medan produktion kraver material control, registration, copper balance, via filling och assembly flatness. Om produktionsavsikten inte sants tidigt skiljer sig resultaten.
Vad bor en leverantor skicka tillbaka efter review av ett HDI-projekt?
Minst stackup recommendation, DFM comments, lead-time options, tooling assumptions, test suggestions och de punkter som kan paverka yield i volymproduktion.
Next Step
Skicka drawing eller Gerber, BOM eller lista over nyckelkomponenter, prototype- och production-volym, operating environment, target lead time och compliance target. Vi skickar tillbaka DFM review, stackup recommendation, prototype-vs-production-risker och offert med lead-time-alternativ. Borja via quote eller contact.
