Mr. Kim,
Thank you very much for what I can trust are the true facts on the Korea EMI Certification. I'm sorry I may have upset you, but the information I had came from several sources and it seemed to be reliable, some of it still is. It seems my main errors were: (a) to assume Korean labs took the same time and cost doing EMI tests as US labs, and that (b) US lab test data could never be accepted for Korean certification. I was also under the impression that the EMI certificate cost was US$712, then US$389 every year after. This must have been the lab testing fees plus the government fees, from what you say. However, you confirmed the most troublesome requirements - yearly retesting and translation of the user manuals. Another label isn't fun either. You see, our products are not like typical ITE that people toss out after a couple years - they continue for years and years and years. One extreme example is our QBus interface card that we still make, it is for a computer (Digital LSI-11) that was first designed in the 1970s! By the end of this year we will have well over 200 products subject to EMI testing for Korea, such as PCI boards, ISA boards, SBus boards, PCMCIA boards, stand alone data extenders and more. A video board is easy to setup and test, set the resolution to maximum and go. Our boards have a variety of I/O methods, analog, digital, timers, and a variety of proprietary and new industrial data busses. We make ISM equipment, mostly PC based, and the majority are Class A. I think the EMI regulation fits the fast paced ITE industry well, but has serious drawbacks for ISM products like ours. We've already done CE for all of our products, we spent over US$1,000,000 on this while also building our own new EMC lab. We don't release new products without CE testing. Some products are also subject to FCC as well. We're also doing the Australia C-Tick. Now Korea has EMI requirements, next Taiwan has EMI requirements, the CE Mark will add several more EMC tests soon.... So, you might imagine that I (and maybe a few others too?) am frustrated by all the new things we must do. How many certificates, declarations, and labels and marks will be enough? A nearby computer company has a large staff (about 12+?) for EMI testing of maybe 40 products per year, and many obsolete, as you suggest, every year. Here, it is me and my technician for about 100 new products per year, and sustaining many old products too. From my experience working elsewhere and dealing with other labs, we're doing a darn good job business-wise! However, I must admire your lab's ability to perform all emission tests and write the test report in a single day. This is good example for the rest of us to emulate, where possible. Do you use an open area site or a semi-anechoic chamber? We use an OATS (we lease from another lab) that is 90 minutes drive away, one way. If all goes well we can complete two radiated tests in one shift. We have a method that cuts our report writing time down to almost as good as yours for a full CE test; we'd still like to make it shorter. Yes, I am aware that the FCC does not have NVLAP setup for Korea. I'm afraid I don't have much influence on such matters, and I agree it is not a fair situation. I plan to submit for NVLAP soon, but I don't enjoy the idea. Let me restate your key points to be sure if I understand the situation now and correct my errors for everybody's benefit in the internet: (1) US labs can apply for laboratory certification to run the tests themselves. (Good! What is the cost to apply?) (2) Retesting is required yearly if the product is still on the market. (This is what I was saying.) (3) Block diagrams are OK. (Good! Though this conflicts with another generally reputable source.) (4) Korean labs charge about US$600 for an entire test and report, usually done in one day. (Amazing!) (5) Two US labs are approved for the EMI certification test at this time. (Not so good, but better than I believed.) (6) The Korean government EMI certificate costs US$70 and is ready in three days after submission. (Good!) (7) User manuals must be translated into Korean prior to certification. (Costly to us.) (8) A label with government division number, emission class, fiscal year, and lab ID is required. Comments are welcome. Best Regards, Eric Lifsey Compliance Engineer National Instruments USA [email protected] on 08/01/97 01:45:42 AM To: Eric Lifsey/AUS/NIC, [email protected] cc: Subject: Re: Korea in a Nut Shell Haitong EMC Inc. Tel : 82-339-376-4117 Fax : 82-339-376-4118 Email : [email protected] Ryan Kim / President of Haitong EMC Inc. ---------- > ?????????????????? > ???????????????????? > ???: Korea in a Nut Shell > ???: 1997???????????????? > > What a surprise! Where did you get those wrong information on Korea EMI regulation. I wish you have confidence before post your opinion. > > A past posting requested info about Korea EMC Requirements. Here's what > I've learned, condensed for your reading pleasure, and interpretted for > it's deeper meaning. > > A foreign company must hire a Korean EMC lab to represent them for > government certification, and pay the lab accordingly for the emission > tests, report creation, and submission of the whole thing to the Korean > authorities - along with the required fee. User manuals must be translated > into Hangul/Korean. You might also have to provide full schematic diagrams > with the test report. > ** What did you mean hire Korean EMC lab. For the Korea EMI, only Korea government authorized lab can issue the report to apply for the final certi from the government. Mostly in Korea (31 labs) and 4 in United States (IBM). Why don't you apply for the Korea EMI lab accredition instead of hiring Korea lab! There is no law that other contries lab can not apply for the accredition. 4 IBM labs approve the fact. Normal charge for the test and report is U$600 per model which is almost one third of America lab charges. Submission fee to the government is U$70 which is more than 1/10 of FCC application fee. Korea EMI lab issue the report and get test and report fee as well as application from the applicant and submmit test report to the government with submission fee without extra charge. What is wrong with that? Also, EMI lab get the certi and delivery to the applicant. Certi is issued within 3 days after the submission which is really short period of time compared to the other countries. User's manual must be translated in to Korean because Korea government wants to let Korean people have the Korean written manual for the purchasing imported devices. If you do not translate your manual and supply with English manual, how Korean people read your manual and understand how operate? We, Korean people have learn English to use your products? This is the reason why Korean government want to see Korean written manual and let foreign manufacturers supply their system with Korean written manual. I don't see any problem with that. If Korea manufacturer supply its system with Korean written user's manual, how American customer will act? Imagine! Full circuit diagram must be provided? Where did you get this wrong information? Only one page block diagran is needed. If you even do not want to provide one page block diagram, its O.K. Korea EMI lab draw the block diagram for you. > This must be repeated every year, and for each product. Every year, every > product (plus fees!). > ** NO! If your system is no more manufactured after one year, just forget it. If your system is keeping imported after the one year, you are subjected to have test report to check the your system's EMI condition to compare to the origition condition when you get the EMI certi. Thus, if you want have the report, apply to the Korea EMI lab and get the test report. We charge U$600 for the test and report which is same as first test and report. If we do not have this kind of survellience system, how Korea government control the inspection. Just get the Certi with good system and sell bad system to the public without and limit? And except test and report fee, no other charge and submission fee is needed. If your system is so good and very popular to the public in Korea, and keep selling for more than year with one model, get just one more report after one year. I bet your system would not be popular and would be closed within one year, then just forget about retest. > Now stop a minute and consider the impact to an imaginary company that > markets a relatively common but noticable number of products. Let's toss > up the number 60 for products, and assume a work week in Korea is five days > long, and that we're using a typical lab with a typical open area test > site... ** What company have the number 60 for products? Let me know. Each model, let say one company have 5 computers with different model name and spec. If this company export 3 models to the Korea, only 3 computers are subject to get EMI certi. Now, let's assume that you use my open site. > Each emission test takes two days to run, plus time to write and assemble > the test report, maybe three days total. (Labs tend to be notoriously slow > at doing reports, though.) ** Each test take approximately 4 hours and issuing the report another 2 hours. Which lab is so notoriously slow at doing report. America lab spend at least one shift (8 hours) for the test. Normally 2-3 shifts for the one model. And charge U$1500 (for example) for the shift. So total test fee would be U$4,500 and spend 3 days. And issuing the report the next day. So, approximately 4 days (if lucky) would be needed to get report with so much money. Korea labs! one day job for the test and report with U$600. Test procedure is same and testing equipment is same too compared to America lab. Now, compare Korea lab to America lab. Which lab is so fast with less money. One big Amreica company visited my lab 10 days ago with 2 computer systems. I finished test within 1 shift and next day I issued report. They were so glad for the finishing so quickly with less money. They expected more than 4 days staying in Korea because they have many such time consuming experience in Amreica. If you want to see the proof, just contact me. I will show you. > > Three times 60 means that 180 EMI lab working days are required to > completely run through all the products our imaginary company makes. That > amounts to 36 weeks or 9 months of almost continuous testing at one lab. ** I even do not want mention about this. Because it is just your imagine and mathemetical multiflying. Last months, I finished 120 models test and report and of course all of them got the certi. We work 24hours a day if we have many require without getting extra pay. That means we can finish 6 models a day with one test site. Lab working days are the term of America's not Korea's. I really don't understand why you put the meaning of working days. Your people prefer to have not Korean. We charge U$600 for the test and report per model not by testing hours or shift. So, I even do not know the meaning of "EMI Lab working days" What does 36 weeks, 9 months mean? Do you think we Koreans are very undeveloped country people? > > Imagine if this company markets 120 products. Now it needs 18 months of > testing. Oh darn, they don't make years that long, do they? It's now time > for the EMC lab to work two shifts, or hire a 2nd lab! > ** Oh God, getting deeper. Send 120 products to me and I can finish within one month for the certi. My question. What company markets 120 products within one month? Let me know. I want to contatc them to make money within one month. And we don't wait until 120 systems get test report. We go for the one test and report get certi as you may, may know. > We have nearly 200 products now and are still growing. (We can't obsolete > things nearly as often as we'd like, people still need QBus cards for their > good old PDP-11s.) Anyway, we'd need 30 months now for a single shift at > the lab, two shifts needs 15 months - whoops, still three months left over! > Gadfry! we'd need three shifts! ** I have one big American company customer. They markets more than 600 products per year to Korea. They don't have any complain and very glad to Korea regulation in terms of time and money. 200 products, peace of cake. Don't worry be happy. Contact me for the Korea EMI. If I spend more than two months to get all of 200 system's EMI certi, I would not charge any cost. Free! Free! > Instead, let's guess that Korea company has 200 products which need marketing to the United States. How long time and how much money that Korea company has to spend!!!! > So we've just determined that, for a company of our size, we would have to > purchase a Korean EMC lab (to assure access and priority, how else could it > be done?) and staff it to run three shifts. It would still take nearly an > entire year to finish the first round of testing. Oh yeah, I was so > worried about just finding TIME to run the tests, I forgot to consider the > COST, or even if there are enough Korean labs to go around. ** Your people work so slowly like that? Then purchase my lab and hire me as your test engineer and pay me U$3,000 per months. Then I will finish 200 systems within 2 months and will relax, sit back for more than 1 year to wait another 200 products. Cost? 200 products X U$600 = U$120,000. You pay me just U$6,000 for two months salery and you can save U$114,000. How much money Korea company has to pay to get 200 products FCC approval? I bet you must underestimate Korea and Koreans!!! > FCC DOESN'T ACCREDITTED ANY KOREA LABS AS NVLAB!!! DID YOU KNOW THIS? > (Ha ha ha ha!) *** No good! What's so funny. > > Good luck to the rest of you that have the TIME. As for me, I'm going to > sit back and watch the fireworks. > ** Good luck to you too. I will sit back and wait until you purchase my lab. Pay me U$ 1 million for my test lab. It is cheap price since my net profit is more than half million dollars per year. Be real! Instead of sitting back, be a business man and try to make money not just complaing. ** Wish to hear from you soon not compain but your opinion with correct information. ** Bye now. I have to go for 20 system test which I got today. Best regards, Ryan Kim
