Skip to content
S.B. Electronics Printed circuit boards · since 2000, home page

Engineering guide · Assembly & standards

SMT vs THT assembly — choosing the right method

How surface-mount and through-hole joints are made, which parts belong in each, and how a mixed-technology board is built and inspected on one line.
Mixed-technology assembly: an SMT microcontroller and passives beside through-hole headers, connectors and a capacitor
FIG. 01Mixed-technology assembly: an SMT microcontroller and passives beside through-hole headers, connectors and a capacitor
Written by
S.B. Electronics Engineering Team
Published
Updated
Reading time
7 min · 1,538 words
On this page · 10 sections
  1. 01What is the difference between SMT and THT?
  2. 02How does the SMT process work?
  3. 03How is through-hole soldered?
  4. 04When should you use through-hole instead of SMT?
  5. 05When is SMT the better choice?
  6. 06How are mixed-technology boards assembled?
  7. 07Why do moisture-sensitive components matter?
  8. 08How are SMT and THT joints inspected?
  9. 09How do SMT and THT affect assembly cost?
  10. 10What do you need to send for an assembly quote?

Building this board?

Nearly every modern board is surface-mount at heart, yet through-hole survives on most industrial designs — because for connectors, transformers and anything a person plugs, pulls or drops, it is still the right answer. Choosing well means matching each component's job to the process that suits it, and laying out the board so both processes run cleanly on the same assembly.

What is the difference between SMT and THT?

Surface-mount technology (SMT) solders components onto pads on the board surface: solder paste is printed through a stencil, parts are placed by machine and the board is reflowed in an oven. Through-hole technology (THT) inserts component leads through plated holes and solders them on the far side by wave, selective or hand soldering. The joints differ in shape: an SMT joint is a fillet on a flat pad, a THT joint fills a copper barrel that runs through the whole board.

Fig. 02Solder joints · cross-section

SMTFillet on a surface pad

Paste is printed, the chip placed and reflowed; the joint is a fillet up the end of each termination.

THTLead anchored through the barrel

Solder wets the lead and the plated barrel; IPC Class 2 and 3 look for at least 75% vertical fill.

Cross-sections, to scale against a 1.6 mm board: an SMT chip joint on its pad, and a THT lead soldered through a plated barrel with the 75% vertical-fill line.

Table 1 · SMT and THT side by side

FactorSMTTHT
MountingSoldered onto surface padsLeads through plated holes
DensityVery high; both sides usableLow; every hole passes through every layer
ProcessPaste print → place → reflowInsert → wave, selective or hand solder
Smallest parts0201 (imperial) passives; fine-pitch BGA and QFNLead pitch typically 2.54 mm, down to about 1.27 mm
Mechanical strengthAdequate for most partsHigh; the joint anchors through the barrel
Cost at volumeFully automated; lowest cost per jointOften manual insertion; higher cost per joint
InspectionAOI; X-ray for hidden joints under BGA and QFNVisual and AOI; fill visible from both sides
ReworkHot air or a rework station; fine pitch needs skillSimple with an iron and a desoldering tool

How does the SMT process work?

SMT is a four-step line: print solder paste through a stencil, place the components, reflow, inspect. Each step has a measurable window, and a board only works if every step stays inside it.

  1. Paste print. A stainless-steel stencil, laser-cut from your paste layer, deposits paste on every pad. Paste volume is the single biggest driver of joint quality — see SMT stencils for foil thickness and aperture limits.
  2. Placement. Pick-and-place machines read your centroid file — designator, X, Y, rotation and side — and align to the board's fiducials.
  3. Reflow. A typical lead-free profile preheats at 150–180 °C for 60–120 s, holds the joint above the 217 °C liquidus for 45–90 s and peaks at 235–250 °C; the paste datasheet sets the exact profile. Our line: 8-zone lead-free reflow; leaded profiles on request.
  4. Inspection. Automated optical inspection checks presence, polarity, offset and fillets. X-ray checks the joints nobody can see, under BGAs and QFNs.

How is through-hole soldered?

Through-hole parts are inserted — by hand on most mixed boards — and soldered by one of three methods. Wave soldering passes the whole underside over a standing wave of molten solder. Selective soldering uses a small programmable fountain to solder only the THT joints. Hand soldering suits low volumes, heat-sensitive parts and rework.

Table 2 · Through-hole soldering methods

MethodBest forWatch out for
WaveBoards that are mostly THT, at volumeBottom-side SMT must be glued and wave-tolerant; shadowing behind tall parts
SelectiveMixed boards with SMT on both sidesA keep-out around each joint for the nozzle; slower per joint than wave
HandPrototypes, low volume, heat-sensitive partsOperator-dependent; slow at volume

A lead-free wave typically runs at 250–265 °C with 3–6 seconds of contact. Whatever the method, the target is the same: solder wetted to the lead and the barrel, with at least 75% vertical fill for IPC Class 2 and Class 3 work. We run reflow, wave and selective / hand soldering in-house.

When should you use through-hole instead of SMT?

Use THT where a joint must carry mechanical load or high current, or where the part is simply not made in a surface-mount package. A barrel-anchored joint resists pull, torque and vibration far better than a fillet on a pad.

  • Connectors that see insertion and pull forces: terminal blocks, power jacks, pin headers, receptacles with through-hole shell tabs.
  • Heavy parts — transformers, large electrolytics, power inductors, relays — where pad adhesion alone becomes a fatigue risk.
  • High-vibration service: motor drives, vehicles, rail and industrial machinery.
  • High-current paths, where a lead through a plated hole connects to copper on every layer.
  • Boards a field technician will repair with an ordinary soldering iron.

When is SMT the better choice?

Almost everywhere else. SMT parts are smaller, cheaper and available in more variants, and machines place them with repeatable accuracy. Short leads mean low parasitic inductance, which matters for switching regulators and fast signals. Both sides of the board can carry components. And because the process is automated, cost per joint falls sharply with volume. For a new design the useful question is not 'SMT or THT?' but 'which few parts justify a hole?'

How are mixed-technology boards assembled?

Mixed boards are built in a fixed order: SMT first, then through-hole. Reflow temperatures would damage many THT parts, and leads sticking through the board would block the stencil, so the surface-mount work is finished before any lead goes into a hole.

  1. Print, place and reflow the bottom side, if it carries SMT parts.
  2. Print, place and reflow the top side. Heavy bottom-side parts may need adhesive to survive the second reflow upside down.
  3. Inspect both sides by AOI, and X-ray any BGA or QFN joints.
  4. Insert the through-hole parts from the top.
  5. Solder them by wave (bottom-side SMT glued, or shielded by a pallet), by selective soldering or by hand.
  6. Clean if the flux system requires it, inspect the THT joints and run any functional test.

Layout decides how smoothly this runs. Group THT parts in one region and leave clear space around each joint for a selective nozzle. Keep tall parts away from fine-pitch SMT so cameras and rework tools can reach it. If you want a wave process, limit bottom-side SMT to wave-tolerant chip parts placed across the direction of the wave.

Why do moisture-sensitive components matter?

Plastic IC packages absorb moisture from the air. In reflow that moisture turns to steam and can crack the package or delaminate the die — the so-called popcorn failure. J-STD-033 grades parts by moisture sensitivity level (MSL), each with a floor life: the time a part may spend out of its dry pack before it must be baked.

Table 3 · Floor life by moisture sensitivity level (J-STD-033)

MSLFloor life at ≤ 30 °C / 60% RH
1Unlimited
21 year
2a4 weeks
3168 hours
472 hours
548 hours
5a24 hours
6Time on label; bake before use

Parts that exceed their floor life are baked before placement, typically at 125 °C: up to 24 to 48 hours for thick packages and less for thin ones, as set out in the J-STD-033 tables. Unused parts are resealed with desiccant and a humidity indicator card. Through-hole parts soldered by wave or by hand never take the full reflow heat in their bodies, which is one more reason large THT packages remain common.

How are SMT and THT joints inspected?

SMT joints are small and numerous, so machines inspect them. AOI after reflow checks every part for presence, polarity, offset, tombstoning and fillet shape against a programmed reference. X-ray checks hidden joints — BGA balls, QFN thermal pads — for bridges, opens and voids; around 25% void area per ball is a common acceptance limit. THT joints are larger and visible from both sides: an inspector, or AOI on the solder side, checks wetting to the lead and the barrel, vertical fill and the top-side fillet. Both are judged against IPC-A-610 at the class you specify.

How do SMT and THT affect assembly cost?

Cost follows setup and labour. SMT has a fixed setup — stencil, feeders, placement program — and a very low cost per joint afterwards, so it gets cheaper per board as quantity rises. THT is mostly labour, so its cost per board stays nearly flat. On a board with five through-hole connectors and three hundred SMT parts, the THT steps can dominate the assembly cost at volume. Converting a part to SMT, or grouping THT parts so one selective program covers them, often saves more than any component price negotiation.

What do you need to send for an assembly quote?

  • Gerbers and drill files, including the paste layers.
  • A BOM with manufacturer part numbers, designators, quantities and do-not-place flags.
  • Centroid (pick-and-place) data: designator, X, Y, rotation and side.
  • Assembly notes: polarity, special parts, adhesive, conformal coating, test requirements.
  • Your build model — full turnkey (we source the parts), partial turnkey or consigned (you supply them).

We run sMT, through-hole (THT) and mixed technology builds in-house. Inspection: AOI at every stage; X-ray for BGA; functional test on request. Workmanship: IPC-A-610 Class 2 standard; Class 3 on request. Send the data pack through PCB assembly or our contact page; when we buy the parts, component sourcing reviews the BOM for availability and lifecycle first.

Written by

S.B. Electronics Engineering Team

The CAM, production and quality engineers at our factory in the Gandhinagar G.I.D.C. Electronics Estate — the people who review incoming data and build the boards. Published ; last reviewed and updated .

Ask the engineering team a question

Questions

Questions about SMT and THT assembly

Short answers to what engineers ask us most often about this subject.

Is SMT more reliable than through-hole?

Neither is more reliable in general. SMT joints perform very well under thermal cycling and normal handling; THT joints are stronger under mechanical load, pull and vibration. Reliability comes from choosing the right method for each part and building it to the right IPC class, not from one technology.

Can a board have SMT components on both sides?

Yes. The bottom side is reflowed first, then the top. Light parts are held by surface tension during the second reflow; heavy bottom-side parts may need adhesive. Through-hole parts go in after both reflows, usually by selective or hand soldering; an open wave would wash over bottom-side SMT unless it is glued or shielded by a pallet.

What is the smallest component S.B. Electronics can assemble?

Smallest part: 0201 (imperial) passives. Fine pitch: 0.4 mm minimum BGA ball pitch; QFN, QFP and fine-pitch ICs. At these sizes the stencil and the board finish matter as much as placement — ENIG's flat pads print more consistently than HASL below about 0.5 mm pitch.

Should I convert through-hole parts to SMT?

Convert where it costs nothing in function: resistors, capacitors, diodes, ICs and small inductors all exist in SMT packages. Keep THT for connectors under load, heavy magnetics and parts that see vibration. Each part converted removes a manual step, so the saving grows with volume.

What is the difference between turnkey and consigned assembly?

In turnkey assembly we buy every part through authorised channels; in consigned assembly you supply the parts and we assemble them; partial turnkey splits the BOM between us. Consigned kits need overage for machine losses, typically 2–5% on passives and more on the smallest packages.

Start a project

Send the data. We will check it before we build it.

Price one- and two-layer boards online in seconds, or send the full data pack — Gerbers, drill, fab note, BOM and centroid — and an engineer replies with a reviewed quote.