SIR endpoint readings tend to be higher following IPA cleaning.

This study investigated the hand rework cleaning process using SIR (surface insulation resistance) via the IPC B-24 coupon test. The final IPA clean was compared to "no final IPA clean." The weak organic flux pen is used on all test coupons as described in the Process Description.

Materials. Materials used were:

• SAC Alloy A: flux by volume 3.3%, flux class ROL1.

• SAC Alloy B: flux by volume 3.0%, flux class ROL1.

• SAC Alloy C: flux by volume 2.2%, flux class REL0.

• SnPb Alloy A: flux by volume 2.2%.

• SnPb Alloy B: flux by volume 2.2%.

• IPC-B-24 FR-4 coupon: OSP with no paste applied (used for wire-only portion of testing).

• IPC-B-24 FR-4 coupon: with Indium SMQ 230 solder paste applied and reflowed onto comb pattern.

• IPA (isopropyl alcohol).

• SnPb tinner.

• No-lead tinner.

• Plato RMA flux braid.

• Sponge Q-tip (for cleaning).

• Synthetic bristle brush (for cleaning).

Process description. The following rework process was used to manually rework SMT components.

1. Apply flux pen (water-based flux containing weak organic acids) on leads or end caps and pads.

2. Remove failed part from pads.

3. Remove excess solder from pads by using a solder braid that also contains flux (RMA).

4. Clean pads and surrounding substrate by applying IPA to a sponge Q-tip and wiping area thoroughly.

5. Apply flux pen on pads.

6. Place new part using SnPb or Pb-free alloy solder wire containing RMA type flux core.

7. Perform final clean by applying IPA to a synthetic bristle brush and then wiping around leads of ICs or end caps of components.

Test procedure. Delphi uses a modified IPC-TM-650, method 2.6.3.3 to accommodate different environments. The original IPC method calls for a stringent vapor pressure level. Currently, the lab uses this procedure with two additional modifications. The measurement bias is not reversed as compared to the test bias and, generally, no soldermask is put on the boards. The pass/fail criteria are based on a specification of resistance (in ohms) to be > 5.0 x 108V; the IPC standard specification is > 1.0 x 108V.

Using the IPC-B-24 coupon the steps described above were followed. Materials were applied using the same parameters on a production product. Solder was applied, covering 100% of the surface of the comb pattern. An IPA clean was used with a sponge Q-tip to remove flux residue after solder removal on all substrates. The rework wire alloy was then applied. For each wire alloy a test consisting of a final IPA clean versus no final IPA clean was performed as described below.

• Test 1: No-lead solder wires.

  - Follow Process Description with final IPA clean.

  - Follow Process Description without final IPA clean.

• Test 2: Standard SnPb solder wires.

  - Follow Process Description with final IPA clean.

  - Follow Process Description without final IPA clean.

To establish an SIR baseline, solder wires were applied to a bare IPC-B-24 coupon.

• Test 3: No-lead solder wires and two SnPb solder wires without applying anything to the coupon except for the wire itself.

  - Weigh an exact amount to apply to the IPC-B-24 PCB.

  - Apply wire only (no flux pen, solder braid, tinner, IPA wash).

Results

No-lead SAC alloy SIR tests showed performing a final IPA clean had no significant effect. The data yielded slightly better SIR readings for the cleaned coupons (Figure 1).

SnPb alloy SIR tests showed performing a final IPA clean had no significant effect. The data yielded slightly better SIR readings for the cleaned coupons (Figure 2).

Wire-only SIR test showed cleaned coupons was at least as good as coupons that had only the wire applied (compare Figure 3 to Figures 1 and 2.)

Summary

IPA cleaning of no-clean fluxed assemblies at repair does not lower SIR readings. In fact, this study shows most SIR endpoint readings to be higher in resistance when cleaned with IPA.

 

Curt Alexander is a senior manufacturing development engineer at Delphi Electronics and Safety, a division of Delphi Electronics Corp. (delphi.com); curtis.m.alexander@delphi.com.