¿ªÔÆÌåÓýA long time ago Tektronix made a little board for adjusting scope probes and Seeing frequency response. I think I have a couple here. ¨C Mike ? Mike B. Feher, N4FS 89 Arnold Blvd. Howell NJ 07731 ? -----Original Message----- ? Jinxie, Ultimately, the probe is supposed to measure things. If it measures things well, that¡¯s all. ? Get a square wave going from a generator, put it through a couple of fast inverters in series, surface mount, well decoupled, on a ground plane, and measure with known good probe and the eBay probe. Not at once! Save the response traces in the scope or on a computer to compare them. Be very careful that each probe attachment is reproducible. Ideally you¡¯d solder a ground sleeve to the pcb, then slide the probe tip into it. That way the geometry will be reproducible between probe insertions. ? If their response is very similar at all time scales, from 1s and 2Hz wave down to 100MHz square and 3ns around the edge, then that¡¯s all you care for. Then, to ensure that the low output impedance of the inverter doesn¡¯t obscure things, add a series resistance on the output, say 500R. If both probes do the same thing then too, you don¡¯t even care if the eBay deal is a fake or not. It works just as well. ? Sure, a fake may be less durable, react more to humidity and so on, but it¡¯s really unlikely that it¡¯d have identical response yet different construction. The parasitics determine the probe response too, it¡¯s a part of the design process to make them work for the design and not against it. In practice, you could copy the geometry of the genuine probe perfectly, with all conductive parts idebtical, but unless you used the same compositions of plastic dielectrics as in the original, you¡¯d still end up with a different response - different enough to measure! Unless you got very lucky that is. ? Measuring two probes at once from the same source is tricky. Using fast inverters as a signal source you could fabricate two machine-assembled PCBs, carefully solder the probe sleeve using a fixture so that it¡¯d be within <0.1mm from where it is meant to be on both boards. Then feed the boards from the same square source and measure probes in parallel. The probe tip should be touching a slightly enlarged pad on the output of the final inverter. This will keep the geometry much tighter than playing with BNC tees and such. ? Cheers, Kuba ? ? ? |
