Hi Ken, I remember that thread. I think I even started it!! There is no way that FCC + FCC = FCC but the current rules and regs allow that to happen. its no surprise that Derek is finding PCs that fail. Question to Derek: Do the PCs have the infamous "assembled from FCC compliant components" label??
From: owner-emc-p...@majordomo.ieee.org [mailto:owner-emc-p...@majordomo.ieee.org]On Behalf Of Ken Javor Sent: Tuesday, March 25, 2003 6:19 PM To: lfresea...@aol.com; emc-p...@majordomo.ieee.org Subject: Re: OK, what's going on? All, I would suggest a little literature survey and analysis before approaching the FCC. Answers to the following questions would be helpful to have in hand before approaching authorities: What bands were the outages in? I don't mean a list of frequencies vs. dB above limit, but outage frequencies vs. spectrum usage. If not many people are trying to use the spectrum where the outages occur, then you don't get many complaints. For instance, the broadcast television band extends from 54-87 MHz, and then picks up again from 174-216, and then from 470-890, except maybe the cell bands overlap the last part nowadays. But how many people in fringe TV reception areas in the USA are still trying to receive broadcast, vs. satellite or cable distribution? How many of this subset would operate their PC and also need their TV to receive on a channel that encompasses a problem frequency? The FCC responds to consumer complaints, or the potential for such. If the answers to these questions fall a certain way, it might make more sense to relax the limits... Related to the above: What does 20 dB above the limit mean? Look at the derivation of FCC Class A/B RE limits and it is tied to received signal quality. But how far are you from a transmitter before the broadcast signal is low enough that a 20 dB outage would cause a problem? Finally, I know Derek and am fully confident he is making accurate measurements, and is smart enough to differentiate an ambient from an EUT-sourced signal. But what anyone needs to know, before running to the FCC, is what is the configuration of the PC you tested, relative to the "golden unit" that was qualified to Class B. How did the design change between qualification and mass production? How many vendor changes were made, how many vendors changed parts but didn't change part numbers? I think this is most likely the root of the problem, but you will find the FCC very unwilling to crack down here, because the implication of enforcing this level of configuration control would be to kill the PC peripheral market. There was a thread here some time ago, about FCC + FCC = FCC? It is a very real EMC problem, but the economic forces here are very strong. I think that all these issues need to be addressed before making a case to the FCC. I don't believe the test houses are the weak link. -- Ken Javor EMC Compliance Huntsville, Alabama 256/650-5261 on 3/25/03 3:05 PM, lfresea...@aol.com at lfresea...@aol.com wrote: Hi all, This e-mail has been prompted because of a number of things that have all come together. This may take a little reading, but please stick with it. Last note... this is not intended to pick on any individuals, or organization, but I do want to stir the pot. I operate an engineering lab, helping clients harden their designs to meet EMC requirements. In this particular instance, I was working for a small client, on a card that goes in the PC . In order to test I need a host PC. So, to save money, the card maker supplies 2 clones. Neither of the two PCs passed emissions testing with the card, in fact, above 100 MHz, they fail even the Class A limit: badly! So, before calling my client, I pull his card, the PC is no different, I pull the monitor, then the keyboard, then the mouse... No different. I test just the PC chassis one at a time. On their own, booted and then the peripherals removed. Not even close to passing. Disgruntled, I get my office PC... Fail. I get my kids PC.. over 20 dB over the limit! So, I think so much for clones... I buy 2 Dell ( sorry, no point trying to hide names... ) desktops, both fail, quite badly. However, they have very similar noise profiles... Can 5 PC's all fail? I think my measuring system is set -up wrong. So I verify this. I am within 1 dB of what I expect when I inject a signal from a signal generator and account for antenna factors. Here lies the question: why can I not find a PC that passes? Worse, since they don't pass, who is chasing them down to enforce the requirements? I'm unhappy, because I am taking a clients money to make him meet the requirements, when it seems no one else is. Now, what's making this worse for me, is that I am an EMC Lab assessor. So, I go to labs and make them jump through hoops so that they produce, as consistently as possible, data the characterizes a product. Exercises, like those performed by USCEL, show that labs can have very consistent results. Anyone that stands up and says EMC is not a field where consistency can be achieved, should not be in the compliance business: please close your lab. So if the test are consistent, why the HUGE variations? In the 20+ labs I have assessed, I feel that almost every one had an ethical approach. Ironically, I felt that the bigger companies I visited like HP and Intel were exceptional: both ethically and technically. The rest of the labs were between good to very good. So cheating is unlikely.. I have now spent about 60 man-hours looking for a PC that passes FCC Class B emissions. Something that I should just be able to go to the store and get. As yet, I have no PC. Our field, it appears, is not a level playing field. It appears more like a rugby game in which we have no referee! So why are there no fines being levied? Especially since it seems I can find non-compliant products everywhere! Is the self policing approach out of control? I intend to take this up with the FCC. Is there anyone out there that is supportive of this action ( which means you must be doing things right.. )? Am I wasting my time ( in which case if this is all lip service... why should we even test!!!! )? Or am I missing something ( I listen to 2 by 4's )? Derek Walton Owner of an EMC Lab EMC Lab Assessor NARTE EMC Engineer 30 years of EMC experience