DIP and test stage

Data belongs in the system, not on the tester

DIP relies on manual insertion, where wrong parts and reversed polarity are largely caught by eye; test data mostly stays on the equipment. Both share the same gap: results are not bound to the SN, so any problem is answered from memory.

Get a solution & quote View all modules

Process nodes

Eight nodes, from insertion to assembly handover

The critical part of this stage is test and repair: parameters must land in the database against the SN, repairs must leave a trail, and disposition must be able to block non-conforming units. The order follows a typical DIP line and can be changed per line in the admin.

  • InterceptionNon-compliant work is stopped — no way around it
  • RecordedAuto-logged, queryable afterwards
  • VerdictRule-based verdicts bound to the SN
  1. 01 Interception

    Insertion

    Assembly drawings at key stations, scan to route; material changes mid-run trigger re-confirmation

  2. 02 Recorded

    Wave soldering

    Process parameters bound to work order and equipment; changes are logged

  3. 03 Recorded

    Touch-up and cleaning

    Routing records and touch-up reasons, used to trace back to upstream issues

  4. 04 Verdict

    ICT

    Each parameter stored, not just PASS/FAIL; failures located to the specific test point

  5. 05 Verdict

    Functional test

    Test items bound to the SN; failure causes classified by test item

  6. 06 Recorded

    Repair

    Defect symptom, cause, replaced parts and technician recorded; repair count accumulated

  7. 07 Recorded

    Burn-in

    In/out times and temperature profile recorded; searchable by SN

  8. 08 Verdict

    Handover to assembly

    Boards with unresolved over-limit repairs or failed retests are locked and cannot enter assembly

Shop-floor issues

Four breakpoints in test data and repair records

The common issue is not the absence of data, but data that is not tied to the SN — so specific questions cannot be answered afterwards.

01

Insertion: wrong parts and reversed polarity stay invisible

On-site

Resistors, capacitors and diodes look alike, and polarity and orientation are judged by eye. After first-article approval, a change of operator, material or batch during the run is not re-checked. Wrong parts often surface only at test or at the customer.

System

Key stations show the assembly drawing and the current part number, and routing is allowed only after scanning. The first article is confirmed by a designated role with a record, and any material change mid-run triggers re-confirmation.

02

Test data: only PASS/FAIL is kept

On-site

ICT, FCT and ATE each keep a local file, and MES records only a result. To check one board's parameters you have to find the file on that machine; if the machine is rebuilt, the file overwritten or the machine moved, the record is gone.

System

Testers write each parameter into MES through an interface, one-to-one with the SN. A board's record shows the measured value, limits and verdict for every item, with no dependence on local files.

03

Repair: nobody records what was fixed or how often

On-site

Defective units are set aside and retested once repaired, then released. Which component was replaced and who did it is usually not recorded. A board repaired three times still moves on, and the yield report shows nothing.

System

Repair entry must be bound to the SN, recording the symptom, cause, replaced part and technician. Reaching the repair threshold locks the unit until engineering releases it.

04

First-pass yield: hidden by retests

On-site

A board fails the first test and passes a retest, and the report counts it as good. First-pass yield looks high, real problems are diluted, and improvement has no direction because nobody knows which item failed.

System

First results and retests are stored separately, and first-pass yield is calculated from first results. Failure rates per test item are counted separately, pointing to the station and batch.

Modules involved

Modules used in this stage

Only the modules this stage requires are listed. Others are added or removed to match the line; master data is shared across systems.

Other process stages

Acceptance

What the system must answer after go-live

Acceptance is not a feature list — it asks whether the data can answer specific questions. If these four cannot be answered, the data is still on the equipment, not in the system.

  1. 01

    Any SN replays its test parameters

    Measured values, limits and verdicts are in the system — no need to open the tester and find a file

  2. 02

    Defects point to the test item

    You see the specific test point, station and batch, not a photo of a defect and a PASS/FAIL

  3. 03

    Repair counts are visible

    How many repairs, what was replaced, who did it — all recorded, and locked once the threshold is passed

  4. 04

    First-pass yield is not hidden by retests

    First results and retest results are counted separately, so improvement has a direction

Next step

See the traceability chain on site

At the model factory, look at the full record of one board from routing through test to repair, and verify how complete the chain really is.

Address
Building 8, Room 303, Yilisheng Industrial Park, 602 Fengtang Avenue, Tangwei Community, Fuhai Street, Bao'an District, Shenzhen

Shenzhen Baichuan Zhike Technology Co., Ltd. · Since 2015 , serving electronics manufacturers across handsets, tablets, automotive electronics, FPC, WiFi modules, energy storage and more.

Used for this conversation only; never shared externally. By submitting you agree to be contacted about this request.

Questions?