This is a caution to my fellow critics of the RoHS directive not to overstate the consequences of complying with it. While we agree that it does make things worse, not better, we need to be careful not to make dire predictions that are not based on solid evidence and may not materialize. Doing so only makes the critic appear biased and undermines his credibility. 

Of all the consequences of RoHS compliance I believe that the biggest is turning out to be the continuing cost of compliance verification. However, I'm ignoring simple economic consequences because RoHS proponents have made it abundantly clear that they don't care about added cost. (That's because they see the cost as being borne - involuntarily - by others, not themselves.) I offer here comments on inadequately substantiated claims of consequences that I've seen from RoHS critics. 

Environmental consequences of disposal. The environmental consequences of unregulated disposal of electronic products (apart from unregulated recycling) have always been zero, and the consequences of disposing of RoHS-compliant products will also be zero. No one is going to be poisoned either way, so to say that RoHS actually increases environmental damage is, I believe, disingenuous. As for unregulated recycling, which cannot be prevented, the consequences will be equally bad either way. 

Environmental consequences of increased mining. There will be more mining to produce the extra tin and silver and copper, but will the increase be enough for anyone to notice? After all, we're the ones who keep declaring that electronic products don't use very much metal compared to all the other uses. 

Increased energy use in manufacturing. Anyone who makes this claim needs to show that it is a noticeable increase. 

Reduced solder connection reliability. Before RoHS, the few field failures of solder connections turned out to have been due to inadequate stress relief in the design, or to noncompliance with soldering standards, not because of inherent limits in tin-lead solder. While I'm no expert on solder connection reliability, from all I've read with few exceptions for those assemblies required to comply that will continue to be true with SAC solder.

I have seen the results of lots of studies, including the JG-PP study, and I just haven't seen a technology precipice. There are much bigger differences in the number of thermal cycles to failure among different kinds of surface-mount components (all of them, except large leadless chip carriers, deemed acceptable) than between tin-lead and SAC, regardless of the thermal cycling conditions.

Reduced reliability because of higher reflow temperature. Certainly raising the reflow temperature prevents the continued use of previously adequate materials. Here is a case where the technology improvements that the RoHS proponents blithely assumed would occur have actually occurred. Although the process window is certainly narrower, there are now improved materials, for which we are indebted to the polymer chemists, for plastic encapsulation and board fabrication that allow the higher reflow temperatures. 

There is one concern that I regard as unresolved: whether every SAC BGA can be replaced on a RoHS-compliant assembly without damaging it or the assembly. I suspect that there are some BGAs for which the answer remains "no," but I also expect to see a drop in such cases.

Reduced reliability due to tin whiskers. While the increased use of tin plating as a termination finish on most components does surely increase the whisker risk, it must be acknowledged that all attempts to quantify that risk have been stymied. That is partly because it is so difficult to prove that a failure (permanent or intermittent) has actually been caused by a whisker. (It is also because it is so hard to get reliable causes of any kind for field failures.)

It may be that a large fraction of the "fault not found" or "retest OK" results are due to tin whiskers, but we really don't know. So even if the increased use of tin plating were to, say, triple the failure rate due to this cause, who knows whether there would be a noticeable increase in the number of failures, hard or intermittent, due to all causes?

Besides, the risks can be mitigated. Component manufacturers claim that as a result of the recent unprecedented attention that has been given to tin plating processes, they can now provide tin plating with an acceptably low risk of growing whiskers for several years. It appears that the sales-weighted majority of their customers believe them. 

Or maybe many of those customers just don't care. If all your competitors buy  the same components you do, and if few customers will ever be able to prove that a failure was due to a whisker, then all of you will experience the same whisker risk and hence there won't be a market discriminator. Besides, a higher failure rate will result in increased sales.

Admittedly, even if the tin plating process is in control at most manufacturers most of the time, one must assume that the process on some batches of components from some component manufacturers has not been controlled as it should have been. But it will take many years before anyone will be able to assess the overall whisker risk as a function of time since plating. 

Let's not overlook the fact that the RoHS directive does not, after all, demand the use of tin plating. I'm amazed that fine-pitch components have now been given permission to add lead to the tin since there are other possible compliant finishes. Some have already been accepted by the marketplace. So if tin whiskers should be found to cause an intolerable (to customers) number of failures, that fact alone could not be used to seek further legislative or regulatory relief, since presumably the marketplace would be able to drive a switch to a non-whiskering compliant finish. 

Also, a tin whisker causes electrical malfunction only if it contacts a metal at another electrical potential, and then only if the available current is either too small to melt and vaporize it or so large that a self-sustaining arc results. 

In fact, where the electrical potential difference is small whisker contact cannot be assumed to result in a short circuit. Bob Hilty of Tyco Electronics has shown by some ingenious experiments (unpublished) that a tin whisker can actually touch the surface of a metal (including gold, which does not itself form an oxide layer) at a different electrical potential without a measurable amount of current flowing - at least for the time that he was willing to wait. This is due to the contact resistance of the tin oxide that is always present on tin. It may take a potential difference of as much as ten volts to break down the contact resistance. Many electronic products have voltage differences far below this.

Further, conformal coating, widely used anyway to increase system reliability, can prevent a whisker from reaching another metal surface, even if it does manage to penetrate the coating from below. Two whiskers that had penetrated a coating from separate metal surfaces could meet, but the risk of a permanent short circuit in this way has got to be very low. A manufacturer's decision on whether to reduce whisker risk by paying extra to apply coating on a product that up till now has not had it is simply a business decision. (It is of course lamentable that the decision needs to be made.) 

The biggest consequence of component manufacturers' widespread adoption of tin plating as a termination finish is to electronic systems not even covered by the RoHS directive, namely space hardware and missiles. Where a needed component is available with no other finish, manufacturers of those systems are having to replace the tin, and to accept the costs and risks involved in doing that. I'm sure we can count this as an unintended consequence of the directive. 

The "logistics nightmare." Perhaps there are a few cases for which recordkeeping is needed so solder connection repairs can be made using the original kind of solder, but it certainly is not needed for most. Recognize first that the RoHS directive does not apply to the solder used to make repairs. So there is no reason not to use tin-lead solder where that makes sense. Recognize also that the RoHS directive does not require a manufacturer to maintain records of what kind of solder was used where. The cost for doing that is self-imposed.

If you count yourself as a believer in the need for keeping detailed records, I'll put the question to you: beyond the risks inherent in doing field repairs of this nature, how many cases are you personally aware of where it has been shown that using tin-lead solder to replace a connection originally made with SAC has significantly reduced the reliability?  

I know of only two that are even possible. One is where bismuth is present, which it can be because it is added in small amounts to some tin platings. Given the availability of enough bismuth there are two low-melting ternary alloys that can form. However, the amount of bismuth remaining where a SAC solder connection has been removed is too small to have a discernible effect on the resulting tin-lead solder connection - composition or properties. 

The other case is using tin-lead solder paste in replacing a SAC BGA. If the reflow temperature is below about 225°C, the lead does not get uniformly distributed throughout each solder ball, and the connections will not survive as many thermal cycles. For original tin-lead reflow soldering of an assembly there are good reasons for not wanting to allow the peak temperature to rise too far. But for replacing a SAC BGA there is no need to so limit the temperature. Also, it may not be necessary for a BGA replacement to use any solder paste. In that case, concern for the effect of lead does not arise.

So I openly invite anyone to present data supporting the need for all the logistics that it appears many are already involved in. I also urge that the next time you hear someone make the "logistics nightmare" claim, challenge him. You can be polite, of course. Just say that you've been looking for corroborating evidence. Don't be surprised if he cannot answer you. 

The assertion that this recordkeeping is necessary has quickly attained the status of an urban legend. I call this "boogeyman thinking," and bluntly speaking, it is unprofessional. There's too much money involved for any engineer just to recommend "playing it safe," even if he sees it as reducing his personal risk. Companies pay their experts to give the right answer, not just the safe one. 

Cost, benefits, and risk. In considering risk associated with various options, one must remember that each option entails some risk. The issue is not whether the risk is nonzero but whether it is tolerable. That is, in considering options, one must compare risk, benefit, and cost. 

That bit of simple common sense has, unfortunately, been ignored by the environmental activists responsible for RoHS, presumably because acknowledging it would pose a risk to their organizations' cash flow. But everyone at times disregards it, for the same reason - fear or greed. Who wouldn't be tempted to choose or recommend the option entailing the least personal risk or the greatest personal benefit if the associated cost would be borne by someone else?

In conclusion, environmental activists and politicians fully expect howls of outrage from those affected by their schemes. I believe that the most compelling argument against RoHS is not all the difficulties that it has occasioned but is, and always has been, that it doesn't solve a problem. It entails cost without a benefit. To raise any of the claims that I have just discussed and not to use the clincher is to weaken your position. 

As RoHS critics, let's be honest and careful not to open ourselves to the charge we have leveled against its proponents, namely of irresponsibly having asserted claims without adequate supporting evidence.

 

Gordon Davy 

 


-------------------------------------------------------------------------------Leadfee Mail List provided as a service by IPC using LISTSERV 1.8d
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Leadfree
To temporarily stop/(start) delivery of Leadree for vacation breaks send: SET Leadfree NOMAIL/(MAIL)
Search previous postings at: http://listserv.ipc.org/archives
Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815
-------------------------------------------------------------------------------