|Ask the Experts|
January 28, 2020
Concerns Replacing High Speed Components with No-Clean Solder
We have been told that we should not replace High speed/High frequency components rated at 1GB or higher using no-clean solder - that this can cause reliability problems. Should we be concerned?
|Expert Panel Responses|
Yes, you should be concerned, and it is very good of you to ask rather than ignore this warning.
Very-high or ultra-high frequency (VHF or UHF) electronic circuits (generally accepted to be greater than 1 gigahertz for analog signals or 1 gigabit for digital signals) typically operate on the very outside surface of the conductors. This is called the Hall effect, or "skin effect". Maintaining electrical control over the signal at these frequencies is much more difficult from a circuit design standpoint. These high-frequency signals desperately want to travel wherever there is a path of least resistance.
When they find an alternate path, they will pass through that path rather than through the intended circuit. When this happens it is called leakage current or cross talk. At these frequencies, even the smoothness of the copper traces (or lack thereof) can cause performance issues in the circuit. Dust, debris, even single fiber strands embedded in the solder joint can provide an alternate path to the signal and cause performance issues.
No-clean flux residues, while normally considered to bean insulator, can easily provide an alternate path to signals at frequencies near or above 1 gigabit. For circuits operating in this upper frequency range, it is absolutely critical that no alternate path is provided to the signal, or the circuit will not operate as intended. Therefore you cannot leave behind any no-clean flux residues that would act as an alternate circuit path.
So how does one clean a CCA that requires assembly or rework late in the process?
If a no-clean flux is required because the entire CCA cannot be immersed in water due to the fact that it is either conformal coated or because there are now certain components on the CCA that cannot withstand the wash, a no-clean flux can still be used for rework or post-assembly processes BUT the no-clean flux residues must be completely cleaned.For rework of conformal coated CCAs, the target component area can be stripped of conformal coating and then after the rework is completed the area can be cleaned using a localized manual wash method followed by a blow-off using an ionized air blower AND a short-duration bake to remove all moisture.
Allowing the localized wash to simply evaporate will leave behind the flux solids that were suspended in the cleaning solution, whether that is saponified and de-ionized water, IPA, or any other solvent. Do not think that you can simply blow off all of the cleaning solution and skip the 15-20 minute bake, as you will NOT get all of the moisture off prior to conformal coat touch up, and any trace of moisture entrapped under the masking, under the component, inside connector pins, etc., will also provide an alternate path to the high-frequency signals.
Keep in mind that when using a localized wash process, the localized cleaning solution must not be one that will soften or otherwise harm the conformal coating, if the CCA is coated. A localized cleaning process can also utilize an ESD-safe vacuum line to remove the cleaning solution. Repeated flushing and vacuuming will eventually remove all traces of flux and other residues from the area, but should always be followed by a short duration bake to ensure the CCA is clean, dry, and pristine.
I'll assume that by 1GB you meant 1 GHz. In short, yes, there can be problems with some pastes. To make a blanket statement not to use no-cleans is, in my opinion, drastic. The potential for a problem exists because the dielectric properties of the paste residue may cause losses that affect signal integrity. The higher the frequency, the greater the sensitivity to dielectric properties of the materials involved.
Talk with your paste supplier(s), and ask for information specific to the paste(s) you are using or considering. Give them specifics as to the frequencies of interest. They should have test data on their pastes at frequencies at an above 1 GHz, which is really not that extreme these days.
This question comes up from time to time and I shudder at the thought that people would use a flux without evaluating the material for its functionality and its long term survivability on the product. These are all test that are required when building product to the requirements of J-STD-001 as the flux shall be qualified to J-STD-004.
Flux is used for only three reasons in soldering, prepare the materials to be soldered, protect the materials from being oxidized and get out of the way when the solder comes. How can it harm the product? The residues can harm the product and those residues need to be tested to make sure they are benign after being processed or used.
The use of additional flux is a major concern in low-solid content flux applications as this material is not typically heated to the soldering temperatures and therefore not benign and still active, which will impact product performance and reliability.
So check and verify the flux you are using is capable of meeting the demands of your product and working within the environments your products will be subjected to during its operational life.
Vice President, Technical Director
No clean flux residue may create noise in high speed circuit boards as well with very low voltage mV cases as we witnessed. Though we deployed localized cleaning for respective IC susceptible for noise and got this resolved.
If you do not have any specific IC/Area in board can be done this way,suggest to clean full board with solvent recommended by no clean manufacturer and confirm it meeting IPC TM 650 Ion chromatography/SIR test on cleanliness reliability front before you move.
Indeed this is good question and as Mr. Lambert indicates how it can effect product functionality.
But there good case although we witnessed cleaning localized on risk IC have passed the board functionally on a my signal circuit part. This can't be denied with good data on pre-post cleaning and repeatability established and I feel IPC/Jdec need some study on this.
Supplier Quality Leader
Poor surface insulation resistance (SIR) is the greatest contributor to degraded post-rework performance in high frequency circuits. Spot-cleaning of cleanable or no-clean flux is a prime source of poor SIR. It is extremely difficult to remove all active or conductive flux residue ingredients -- especially under low-clearance parts.Julie Silk, Agilent Technologies, USA
We dictate using a limited amount of no clean flux-core wire solder (tested and approved types, of course) and leaving the flux residue on the board. The wire solder ensures the flux residue is de-activated and it encapsulates any remaining conductive ingredients. Trying to clean it just spreads it around. Never never use liquid flux.
|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