EMC Control Plan
We have written many articles on EMC design and troubleshooting. This time, we are going to shift gears and do one on managing EMC during project development.
The first problem we often encounter is the failure to give EMC its fair share of attention. With tight scheduling, the tendency is to concentrate on the most immediate problems, pushing EMC into the background. Often, EMC testing is the last technological effort before production. Unfortunately, it won’t go away, and the longer you wait to address it, the more painful the solution. Problems that should have been avoided up front end up as costly fixes and schedule delays when uncovered at the tail end of the project.
The second problem is failure to make good EMC design decisions or, worse, failure to make any EMC design decisions. This is an engineering problem, but it is up to the program manager or project engineer to ensure the necessary skills are on hand to make sound design decisions.
EMC is far from an exact science, so you need to plan for contingencies. You need to figure on three test cycles. The sooner you get some test data, the more options you have for remedies. If you have good in-house test capability, your early testing will be fairly accurate, so the final test will likely go pretty smoothly. If you have little in-house test capability, your risks at final test are higher. If you wait until the product is designed and built, your risk of failure is very high, and you can expect significant cost and schedule impact.
Achieving successful EMC control is fairly straightforward if you design conservatively, but if you start making compromises in the interest of manufacturing cost control, you can expect to spend lots more time and money on development and test. This leads to the situation that mild commercial EMC requirements take more development and testing than vigorous military designs.
Military contracts often require preparation and execution of a formal EMC control plan. You may not want or need such a formal approach, but your EMC program will go much smoother if you do an informal plan.
The following paragraphs briefly describe such a plan, defining the steps that need to be taken over the duration of the project. Your EMC expert will have a major input, but your entire design and management team will have some involvement.
1. Identify and Evaluate EMI Requirements
Your EMC requirements may be well defined, or you may have to establish some requirements yourself. The easiest requirements to establish are those of the European Union, which are very well defined. But even there, you may still have to decide in what equipment category your product lies if there is no appropriate product family, you will need to pick the closest fit. In any event, you will need to decide which class your product lies in, notably in defining failure during ESD testing. If you expect to operate in a harsh environment, you may want to adopt more stringent requirements to keep your customer happy.
If you are selling only in the USA, you have more flexibility the FCC exempts a lot of equipment from emission testing and doesn’t quantify immunity requirements at all. In this case, we advise following the IEC requirements (generally identical to the EU requirements) if you don’t meet these requirements, you will most likely encounter problems in the field, leaving yourself with unhappy customers, as a minimum.
At the other extreme would be a military application developed under contract. In most cases the requirements will be pretty well defined, but occasionally we encounter a simple statement, like “Meet MIL-STD-461F.” In such a case, you will need to select the applicable paragraphs and design accordingly. Here, it is really important to do a formal EMI Control Plan, even if not contractually required, and get your customer’s approval early in the project. In many cases a preliminary plan is required in the proposal.
Whichever the case, you will need to establish your EMC design goals early in the design cycle, as these drive your design decisions.
2. Establish Product Goals
There are a number of factors to consider, and management will need to be intimately involved. The first consideration is the trade-off between production costs and nonrecurring engineering costs, and schedule. Generally, the conservative approach will result in lower nonrecurring costs and shorter development cycles, but with a higher production cost. If you expect small volumes, per unit costs will be secondary to engineering costs. Accordingly, you will be looking at a low-risk approach. If you are figuring on producing millions of units, your production costs will be of major import, with engineering development and test costs being secondary. But when you cut manufacturing costs, you start taking risks, so you need to budget much more time for redesign and retest.
3. Identify Key Design Issues
Here, you will need to identify the key EMC design issues. All EMI requirements need to be addressed, but some will be more difficult, or painful to handle. Once you have established the basic design concept, you will need to assess the associated risks. For example, if you have decided you are going for minimum repro costs by eliminating shielding, you are taking a major risk you can expect major emission and RF immunity problems, so be prepared to do lots of board engineering and, even then, you may not make it. If you have a good metallic box, EMC effects will be much less of a problem, leaving primarily the transient effects in power and data cables.
Once you have identified the problem areas, you need to decide how to handle the risk you need to have a fall- back position what if your approach fails? Slipping schedule six months is generally not acceptable you need to be ready with Plan B if your preferred approach fails.
4. Get Preliminary Test Data
You need to get test data in the design cycle as soon as you can. You don’t need to have everything working before getting some data most emissions and immunity problems will show up as soon as you power up for initial checkout. As soon as you find a soft spot, try fixes on the spot the intent is to identify a clear-cut fix before redesign.
Preliminary testing need not be exhaustive a quick look at the key areas will usually be enough to steer you in the right direction. Even if you don’t have full test capability, anything you can find early is money in the bank.
Rework and retest as needed. This can be done while the prototyping proceeds. The goal is to have a good idea of what your final test results will be before the final test.
5. Prepare Test Plan
The test plan is usually prepared by the test lab people. They know what equipment they have, and how to set it up. But you need to provide inputs, including pass/fail definition and support hardware and software.
If support equipment is needed, you will have to make sure it doesn’t skew the test data. You may need supplemental text fixtures, such as breakout boxes for cable injection confer with the test lab on this. You will need to identify what software is required to support the test. You will also need to determine how you are going to tell when a failure is observed perhaps as simple as an alarm light, or it may be erroneous data readout.
6. Test and Redesign as Needed
Here, you are looking at the formal test. If you have done your job with pretesting, this should be a simple process. Unfortunately, even with the best intentions, the preliminary test isn’t always right, depending on how well equipped your preliminary test facilities are, so you can expect to see some problems pop up again here. Now it’s a matter of test and rework until you solve the problems.
7. Test Report
This would normally be done by the test lab.
EMC control won’t happen if you don’t plan for it. Even if you don’t have need for a formal program plan, you still need an informal plan, and follow through with it. Those who don’t, end up paying in the end.
As Murphy says, “There is never time to do it right, but there is always time to do it over.”
DI-EMCS081528. Electromagnetic Compatibility Program Procedures.