> On attenue pas en multimode / SR : les optiques sont prévues pour. C'est en > monomode que les optiques sont dimensionnées pour des portées bien plus > longue et peuvent être éblouies lorsqu'elles sont face à face avec une simple > jarretière. > > > J'ai justement allumé il y a peu une OM3 et je trouve les débits pas > > terrible, je voudrais m'assurer que le soucis ne vient pas de la. :) > > SFP+ Intel d'un coté et HP de l'autre. > > En 10G c'est plus souvent la mesure qui pose problème. Mais si la distance > est courte (moins de 3m), tu peux utiliser un câble SFP+, ça élimine la > question ;)
Please correct me if I am (and my french translation) understood the question wrong. For SFP+ SR an LR there is almost no issue with attenuation on a direct link (e.g. 2m). A SFP+ SR has a TX power level of max. -1dBm (the receiver overload is also defined to -1dBm for most SFP+) --> overload almost not possible For SFP+ LR there is a small change to overload the receiver because the max TX power level can be up to +1dBm and the receivers overload is +0,5dBm. That means, if you get such a SFP+ LR and you have a perfect LC-connection(no scratches, dust and dirt on the ferule of the plug) than you might run into overload issues of the receiver and this might end up in CRC errors on your interface. But it is very rare that you get such a strong laser because the production range for these 1310nm Lasers is -6dBm to +1dBm. In the 1G world it looks similar. The values for max. TX power level for SX and LX is around -3dBm and the receiver overload is -3dBm For sure these values can differ from manufacturer to manufacturer in a range of +/-2dB. To be on the safe site have a look in the datasheet of your SFP / SFP+ for the Transmitter Output Power/ TX max. level and the Reciever Overload / Average Receiver Power / RX max. level. Thomas CTO Flexoptix --------------------------- Liste de diffusion du FRnOG http://www.frnog.org/