|Ask the Experts|
January 24, 2020
Can Solder Joint Geometry Change Resistance?
We have a battery powered unit that draws 10-15 micro amps off the battery. When testing using a regulated supply, we see a high occurrence of units that are drawing between 20 - 30 micro amps and cannot explain why.
Could solder joint volume, fillet profiles or flux residues result in solder joints with higher resistance and as a result draw more current?
|Expert Panel Responses|
Flux residues especially those from HASL PCB'scan cause this issue. Wash the bare PCB thoroughly then test the drainage. If you can give more information as to board finish and assembly flux residue may be able to help in the source of the problem.
Technical Sales Manager
BLT Circuit Services Ltd
Typically these types of higher current draws are due to flux or fabrication residues on the capacitors or at the hand solder wire connections, but can be inner layer as well.
These residues are able to cause leakage but not enough moisture to cause electrochemical migration (dendrite shorting) with hard evidence of corrosion on the component areas. Localized ion chromatography analysis typically finds these residues and quantities and then can be optimized.
President/Senior Technical Consultant
More than likely you have an increase in surface leakage.
CEO & Managing Partner
Why would higher resistance - presumably in series with the battery powered unit - result in higher current? Higher resistance results in lower current, given constant (regulated) voltage.
This is "Ohm's Law". Maybe there's a leakage path in parallel with the unit responsible for the other 10-15uA. Without knowing more about the setup it's difficult to assess this.
Design Chain Associates, LLC
From Agilent Technologies focus on premier measurement solutions, we frequently require a stringent PCA cleaning and handling process to ensure our most sensitive electronic measurement circuits perform as designed, especially in cases where mico-amp/volt (or even pico!) make a substantial difference in functional performance.
While I am not a printed-circuit-assembly process expert, I do know from experience, that that ensuring the overall assembly(especially the PCB surface) is very, very free of residues is important in these cases.
If the solder joint is in the thermal path of the device, and the solder joint thickness varies by a significant factor, say 2:1, this could result in a rather large variation in thermal resistance. Under these circumstances, a measurable difference in current draw is possible.
Excessive voids in solder joints, if in the thermal path, can also cause an elevated operating temperature of a semiconductor device, and a higher current draw can result. A simple X-ray of the device could reveal voiding issues.
f the device is not power oriented, other non-solder materials could be used to attach the die, such as conductive epoxy. In this case, the thickness of the conductive epoxy needs to be controlled to maintain a narrow range of thermal resistance. Otherwise, higher currents could result.
If the suspect solder joint is in a wire of single connection, remelting the solder to a sufficient temperature (a small amount of flux could be required to remove oxides) can be used as a way to determine if a "cold" / improperly formed solder joint exists. If the device has leads, remelting with flux can help reflow intermittent joints.
Global Product Manager
I doubt that the solder joint or its shape has anything to do with your leakage currents. I have only found the shape of the solder joint to be significant in certain RF applications and not in DC applications.
What I suspect is happening is you have leakage currents to ground occurring in one or more locations, adding up to the additional current draw. This can occur through flux residues or if you are cleaning, incompletely cleaned flux residues.
If you are using a low solids flux, are you sure the flux is fully reacted by your heat flow? If you are using a water soluble flux, are you fully cleaning it from the assembly and have you verified this with ion chromatography?
Principal Materials and Process Engineer
Determine whether the intermetallic differs between two electrically-differing joints (micro-section analysis).
If cross-sectional views indicates the intermetallic is visually thicker on the higher current specimen, therein may lie the source of increased resistance. If-so, consider using a PCB surface finish with known reduced intermetallic propensity.
Circuit Connect, Inc.
First and foremost, adding a resistance in the circuit should not increase the current. If you model it, you should find that for a simple DC circuit, adding a series resistance will decrease the current.
That said, any resistance from the solder joint will be in the milliohm range. Your circuit's impedance is around 100 k ohms or more, almost eight orders of magnitude higher.
The source of the high current draw is most likely a circuit-related problem, but could be caused by dendritic growth. In this case, however, the current would not be stable over time, and would increase with increasing humidity.
Assuming the circuit (device under test) drawing the current is the same when powering with a Battery vs a Supply, this may simply be an internal resistance or a measurement accuracy issue. The battery's internal resistance (likely higher than your supply) may contribute to a lower current reading. Or, if you are using the supply's current meter when using the power supply vs some other current meter when using the battery, you could easily see a difference due to the different meters' accuracies. And the supply's meter will likely have a different internal resistance than the meter you use when the battery is in circuit affecting the overall resistance in the circuit. The supply itself may also "leak" a few uAmps as well.
Senior Project Engineer
Electronic Controls Design Inc
|Submit A Comment|
Free Newsletter Subscription
Circuitnet is built for professionals who bear the responsibility of looking ahead, imagining the future, and preparing for it.
Insert Your Email Address