Solder Thief Pad Design And Paste Control

A solder thief is a pad placed next to a joint specifically to take solder away from it. It is not connected to the circuit, or it is connected only to a net where the added solder does no harm, and its purpose is to reduce the volume of solder that reaches a joint that would otherwise be overfilled or bridged.

This article explains when a thief pad helps, how it is dimensioned, and the situations in which it does more harm than good.

The Problem It Solves

Excess solder causes two defects: a joint that is too large and masks a pad that should remain exposed, and a bridge between two adjacent joints. The excess can come from the paste deposit, which may be larger than the joint needs because of the stencil design, or from the wave, which delivers more solder than a fine pitch row can hold. A thief pad gives the surplus somewhere to go that is not the neighbouring conductor.

In hand and wave soldering the mechanism is straightforward: the pad wets readily, and molten solder flows towards it because of the surface tension gradient, taking the surplus with it. In a reflow process the pad draws solder from an adjacent deposit that is overfilled, which reduces the volume in the joint itself. In both cases the pad has to be wetted by the same flux and at the same time as the joint, or the solder has no reason to move. The pad geometry that governs this behaviour is described under PCB pad design standards.

Where A Thief Pad Is Used

The classic application is a fine pitch row that has to survive a wave or a selective process with a high solder volume, such as a connector with a row of closely spaced pins. A thief pad at the end of the row, or between the last pin and the board edge, collects the solder that the row cannot hold and prevents the last joints from bridging. Similar pads are used beside the pads of a through hole connector where the hole volume is large.

A second application is a component that must not be flooded, such as a small chip part beside a large pad. A thief pad placed downstream of the flow path takes the excess before it reaches the small part. A third is a test coupon or a mask opening that would otherwise fill with solder and block a probe or a connector. In each case the pad is placed where the solder naturally flows, which means the direction of the wave or the print matters.

Dimensioning The Pad

The size of a thief pad is set by the amount of solder it has to absorb rather than by a standard figure. It has to be large enough to take the surplus without itself becoming overfilled, and small enough that the surface tension does not strip solder from the joint it is meant to protect. A practical approach is to make the pad area comparable with the pad area of the joint, and to adjust from the result: a joint that is still bridged needs a larger pad, and a joint that is starved needs a smaller one or none at all.

The pad also needs its own mask opening, and the opening has to be sized so that the pad wets. A pad that is partly covered by mask has a reduced area and behaves unpredictably, and the mask edge itself can hold solder. Where the pad is close to a neighbouring conductor, the clearance has to be enough that solder flowing onto the thief cannot bridge from it, which would simply move the defect rather than remove it.

Drawbacks And Risks

The most serious risk is starvation. A thief pad that is too effective takes solder from the joint, leaving a fillet that is thin, concave or incomplete, and the result looks like a wetting problem rather than a pad problem. On a fine pitch row with a small paste volume, removing even a fraction of the deposit can leave a joint below the acceptable minimum, which is why thief pads are used more often in wave and selective processes than in reflow.

The second risk is that the pad changes the local thermal behaviour. A large exposed copper pad beside a small joint conducts heat away, which slows the melting of the neighbouring joint and can produce a cold joint on the very component the pad was meant to protect. On a densely populated board this effect is not negligible, and the pad should be kept away from the joints whose thermal behaviour it would disturb.

Design Rules And Documentation

A thief pad is a feature that exists only for the process, and it must be documented as such. The drawing should identify it, because a pad that is not connected to anything looks like an error to a designer reviewing the netlist, and it may be removed in a later revision by someone who does not know why it is there. Naming it and adding a note prevents that.

Where a pad is connected to a net, the connection should be deliberate, and its effect on the net has to be considered. A pad connected to ground adds capacitance and, if it is large, a place for a stray solder ball to lodge. A pad connected to a signal adds a stub. In most cases the pad is left unconnected, which avoids both problems, and its electrical isolation is then checked like any other floating feature. The way process driven features are recorded with the rest of the layout is part of design guidelines for manufacturability.

Testing The Result

A thief pad is evaluated by the defects it was added to prevent. The measurement is a count of bridges and overfilled joints before and after the change, on the same board and the same process, and it should be taken on enough boards to distinguish a real improvement from the normal variation. Where the pad is added for a wave process, the direction of the wave and the pallet arrangement are part of the test, because the pad only works if the solder flows past it.

Where the result is inconclusive, the more reliable change is usually to reduce the solder volume at its source rather than to divert the surplus. A stencil aperture that is oversized, a wave that is too high or a pot temperature that is too low all produce excess solder, and correcting one of those is a better answer than adding a feature that has to be tuned to compensate. A thief pad is a fine adjustment, not a substitute for a controlled process. The placement of the past deposit that determines that volume is described under placement order and pad positioning.

FAQ

Does a thief pad work in reflow? It can, where the joint is overfilled, but its effect is smaller than in a wave process because the solder does not flow as freely. Reducing the paste volume is usually the better remedy.

Should the pad be connected to a net? Usually not. A floating pad avoids adding capacitance or a stub, and it behaves the same way for the solder. Where a connection is made, its effect on the net should be checked deliberately.

How many thief pads does a row need? One is normally enough for a short row, placed where the solder exits, and none is needed where the row is short enough that its own ends absorb the surplus. Adding several pads increases the risk of starving the end joints.