03:48 <azonenberg> Ooook now i am even more confused

03:48 <azonenberg> photos coming in a few

03:48 <monochroma> :o

03:51 <azonenberg> It's not the tip after all

03:51 <azonenberg> it was another variable i didn't control for

03:51 <sorear> humidity?

03:52 <azonenberg> no, the way i connected my "control"

03:52 <azonenberg> gimme a sec, uploading

03:55 <azonenberg> So, updated conclusion: new probe tip is basically a match for the old one in RF performance, definitely not the cause of my SI problems

03:55 <azonenberg> But the bronze wire has superior mechanical properties and is definitely the best choice moving forward

03:56 <azonenberg> way more solderable than the needles, easier to cut (although it's not pre-sharpened), and slightly springy

03:57 <azonenberg> https://twitter.com/azonenberg/status/1131601434481385472

03:57 <azonenberg> monochroma: ^

03:57 <azonenberg> So, that's the test i did before

03:57 <azonenberg> old pcb rev, SMA instead of the tip

03:57 <azonenberg> it gave good results in pass-through mode with 0R resistors, which leads me to believe the PCB and MMCX connectors are OK

03:58 <azonenberg> agreed so far? note the specifics of how it's hooked up

03:59 <azonenberg> i have my existing test board, a short piece of cable, then the modified probe in-line terminated with 50 ohsm at the scope

03:59 <azonenberg> so the goal of that test was to basically confirm that the body of the probe and cabling acted like ideal 50-ohm transmission line - which it basically does, there's a tiny bit of distortion but not enough for me to really be worried

04:04 <azonenberg> And https://twitter.com/azonenberg/status/1136120944118779906 is today's test

04:05 <azonenberg> The same "step" discontinuity is still there

04:39 <azonenberg> You know what changed? the resistor

04:39 <azonenberg> it went from a 0R (with termination only at the scope) to 1K (with termination right at the point of probing)

04:39 <monochroma> waaat

04:42 <azonenberg> So, it looks like we have a couple things goign on

04:42 <azonenberg> i'm playing around more in qucs

04:42 <azonenberg> If i remove all the parasitics from the resistor, there's a resonance building up in the unterminated stub between the resistor and the DUT

04:43 <azonenberg> I can notch that out by adding a quarter-wave stub of equal length after the resistor

04:45 <azonenberg> The inductive component of the resistor is pretty negligible

04:45 <azonenberg> Added that back in and looks OK-ish

04:46 <azonenberg> The capacitance of the resistor is what's killing me

04:51 <lain> woo

04:55 <azonenberg> lain: thoughts on the test results?

08:14 <_whitenotifier-3> [scopehal] clh27 opened issue #73: Add Tektronix TDS2000 support via SCPI - https://git.io/fjuo7

08:14 <_whitenotifier-3> [scopehal] clh27 opened issue #74: Add Agilent MSO6104A support via SCPI - https://git.io/fjuod

19:32 <bvernoux> hello