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: Ali Chen who wrote (34477)	11/14/1999 6:47:00 PM
From: John Walliker	Respond to of 93625

Ali, Reflections and things... There must be all sorts of data collisions and resulting data distortions that does not improve setup and hold requirements. And the junction between "unloaded" 28-ohm lines and "loaded" 56-ohm portions is still a compromise to match unmatchable impedances, no matter how high some people get excited about "modulated width transmission lines". Was that me you meant? :-) Just for fun I tried investigating the reflection coefficient of a section of 28 ohm transmision line with the 56 ohm matching section, a 2.5pF capacitance to ground to simulate a device input and via and another 56 ohm matching section. I used the Motorola Smith charting package with raw data from the Rambus design guide. I found that the signal reflected from each chip and its associated matching network was about 1% of the incident signal. No guarantees of extreme accuracy here - this was just for fun. However, it does suggest that reflections are not as big as some people seem to think. Why don't you try examining such things yourself - maybe run some HSPICE simulations to get an even more accurate result. All the raw data you might need is in the design guide. John

To: Ali Chen who wrote (34477)	11/15/1999 7:11:00 AM
From: John Walliker	Read Replies (1) \| Respond to of 93625

Ali, Some more thoughts about reflections. In my earlier posting I described a simulation of the Rambus matching network which indicated that about 1% of the incident signal is reflected at each device. It occurred to me that some people might then assume that if the system were populated with 32 devices, 32% of the incoming signal would be reflected. Of course, it isn't as simple as that because all the reflections arrive with different phases and tend to cancel out. So I redid the simulation with 6 devices in a row - this being the most that the Motorola package can cope with on-screen. At 400 MHz the reflected signal was still around 1% at 800 MHz the reflected signal is about 1.8% at 1200MHz the reflected signal is about 4%. Bear in mind that the signal level at 1200 MHz will be very much lower than that at lower frequencies because of the 1ns rise and fall times imposed by the slew rate limiting of the RSL outputs. I am very pleasantly surprised by just how good the matching networks used by Rambus are. In particular, the way that performance is maintained when multiple devices are cascaded together. This is much better than I had expected. The Rambus channel really does seem to be of very high quality. Try it for yourself. John