Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Rambus (RMBS) - Eagle or Penguin -- Ignore unavailable to you. Want to Upgrade?

To: Jdaasoc who wrote (31396)	10/2/1999 1:26:00 AM
From: Tony Viola	Read Replies (1) \| Respond to of 93625

Jdaasoc, How does FCC testing differ from CE testing. VC820 already has CE Declaration of Conformity. The CC820 board does not. The plots, or graphs of allowable decibel/microvolts vs. frequency are different for CE testing vs. FCC testing (the higher the decibel/microvolts, the worse in this case because higher means more energy to interfere with television and other communication media). The allowable levels don't differ by a whole lot, and the difference is that the Europeans (CE) have tighter db limits, for example, over the frequencies where their UHF and VHF television stations transmit. Don't want computers or anything else causing interference. At frequencies where no TV, or police radio, ship to shore radio, air traffic control, or other communications deemed critical is taking place, it gets looser. Same with the U.S., i.e., we are tighter on new electronics units under test over our critical communications frequencies. Again, the U.S. and Europe don't use the same frequency bands for TV and other communications industries that are protected by FCC and CE testing. Hence the differences in the db vs. frequency graphs. Usually, a good design that passes EC will also pass FCC, and vice versa, unless you happen to be unlucky enough to be extra noisy at the critical frequency bands of one or the other. I hope everyone was designing with FCC and CE in mind all this while. I know that Intel has been careful about designing to meet EMC regulations for a long time. But with the screwups, and all the companies involved in this whole new memory technology, who knows? Tony

To: Jdaasoc who wrote (31396)	10/2/1999 1:32:00 AM
From: Bilow	Read Replies (1) \| Respond to of 93625

Hi Jdassoc; It has been some years since I had to work on radiation testing, (so these comments may be in error or out of date,) but I recall that FCC and CE differ in the specific radiation limits (which are defined in frequency bands). I seem to recall that the European standard was more difficult at the high end, but that there were frequency bands where FCC clearance was tougher. (The FCC used to have two classifications that impinged on me. The tighter one, (FCC-B?) is the one that would apply to a mother board.) Possible reasons VC820 (Rambus mother board) has CE, but CC820 (SDRAM mother board) does not: (1) SDRAM inherently has higher radiated noise. (2) Rambus technology is inherently difficult to alter once it is built. For this reason, an engineer would make sure that he met the most difficult radiation specs as early as possible in the program. Not so easy to correct later. (3) CC820 board isn't as far along as the VC820 board, it being started later. (I doubt this.) (4) CC820 not intended to be sold in Europe. (No way.) Clearly it is possible to get the CC820 board to CE compliance. If it weren't, there would be a whole lot of computer companies in deep doo-doo. My guess is solution number 2, radiation is a big deal with Rambus' long lines and high frequencies, so they proved it can pass early on. I've sometimes felt that the CE requirements are primarily there in order to cause trouble for US companies, and that the Euros don't enforce them so strictly on their own companies. My suspicion is that a good bit of the advancement in meeting radiation emission requirements is in the computer case, and in better filtering on the signals that exit the mother board. (I.e. the power cord, mouse cable, monitor cable, etc.) I've always had a certain amount of admiration for the way that Apple builds its equipment, but it isn't that big of a deal, everybody eventually passes FCC and or CE. Some notes for those not familiar with radiation (i.e. "static") testing. This is basically done so that the government can be sure that your equipment doesn't prevent your neighbor from getting Gomer Pyle, USMC on his rabbit ears. (And also to help ensure that your equipment doesn't interfere with other equipment's operation.) The early HP programmable calculators were famous for putting out radiation, people would put them next to a radio so they could hear it. (Big thrill!) There are also restrictions on how much cr@p you put back down the power cable, and the US and European standards on these differed also. It is traditional in the industry to send out prototype systems without having all the clearances in hand. I think this is more or less legal provided they don't actually sell them. They probably are supposed to send some sort of warning with them, but everybody has gotten so used to government mandated mumbo jumbo that I doubt that anyone actually reads it. There are some beautiful free-range testing facilities out west, but back east they have to do it in enclosed buildings with radio absorbtion material on the walls. I used to enjoy driving out to the middle of some nowhere, so far from civilization that they couldn't get TV or radio, to do radiation testing. But it was never my specialty, I can only remember going on about four trips my entire career. The sites are staffed with engineers who specialize in reducing emissions, (generally by reducing your signal levels, which frequently cuts into your signal integrity. This is why Rambus no doubt looked into these issues very early.) Incidentally, I've always felt that spread spectrum (where you deliberately jitter the system clock) techniques for "meeting" radiation compliance were a bad idea. It seems like I'm the only guy with this opinion, cause all my EE buddies tell me I'm wrong. Spread spectrum would be a lot easier with SDRAM, as it has a bit more margin to jitter around. It would surprise my buddies, but not me, if the FCC did something so as to prevent this particular technique. (Basically it broadens out the frequency peaks, so you stay under the limits, but it doesn't reduce the total amount of radiated energy.) -- Carl

To: Jdaasoc who wrote (31396)	10/2/1999 5:06:00 AM
From: John Walliker	Respond to of 93625

Jdaasoc, How does FCC testing differ from CE testing. VC820 already has CE Declaration of Conformity. The CC820 board does not. For radiated and conducted (transmission into the air and along the power cable respectively) emissions the tests are the same. However, CE marking requires some additional tests for immunity to external disturbances. These include swept frequency exposure to a field strength of at least 3V/m 80% modulated with a 1kHz sine wave from 30 to 500 MHz, simulated electrostatic discharge to all external surfaces and connectors and mains supply burst transient testing. These immunity tests must be passed without affecting the operation of the system. The usual process is to take a sample of the equipment to a certified test house and spend about two days there with them. They do the emissions tests first in case the immunity, or most likely the static discharge tests, destroy the equipment. It is possible to make modifications as you go along, but there is very little time to do this. It can cost around GBP 1000 per day. I have been through this process several times (successfully) on behalf of clients for whom I have designed systems. A simple check is to sweep the antenna of a GSM cellular phone slowly over the circuit while it is transmitting at full power. If everything works properly, then the system will almost certainly have immunity of at least 30V/m - ten times the normal limit. Medical equipment can be required to withstand higher levels than information technology equipment. John