04:17 <azonenberg> ooooh i am liking this

04:19 <azonenberg> lain: so, the version of my probe that i sent to fab, plus a 1pF compensation cap to ground after the five resistors, is flat to +/- 1 dB out to about 4.5 GHz

04:23 <azonenberg> https://www.antikernel.net/temp/newprobe.schematic.png

04:23 <azonenberg> oops

04:23 <azonenberg> https://www.antikernel.net/temp/newprobe-schematic.png

04:23 <azonenberg> https://www.antikernel.net/temp/newprobe-freqresp.png

04:23 <azonenberg> https://www.antikernel.net/temp/newprobe-timedomain.png

04:24 <azonenberg> just the tiniest bit of a ring on the rising edge, and the frequency plot rolls off smoothly out to 10 GHz without any weird spikes

04:24 <azonenberg> If (big if) this model is accurate, this is going to be a *very* nice probe

04:25 <sorear> not sure I've seen that symbol for a transmission line before

04:26 <azonenberg> That's what qucs uses

04:27 <azonenberg> Having the five resistors seems to greatly damp the resonances i was getting with one since the impedance change isn't a sudden step

04:27 <azonenberg> i also attempted to model the impedance of the probe tip and mid-span probing on the DUT

04:28 <azonenberg> So we'll see if reality matches in 2 weeks or so

04:29 <azonenberg> Flat to +/- 0.25 dB from DC to 2 GHz, and +/- 1 dB out to 4.5 GHz would be really nice if it works

07:25 <lain> nice

22:30 <azonenberg_work> https://i.imgur.com/8dUgjOm.png early tests suggest a pair of these should work nicely as a compensation cap rather than using a discrete MLCC

22:31 <azonenberg_work> lain: ^

23:30 <lain> :o

23:44 <azonenberg> welp, it's official https://www.antikernel.net/temp/tip-v04.png the new probe design is using rf black magic :p

23:44 <monochroma> :D

23:45 <azonenberg> Just two 500 fF caps

23:45 <azonenberg> (first time i've used that unit in a schematic lol)

23:50 <lain> :D