00:01 <Bird|ghosted> azonenberg: YIKES! (at the GMR head being *that* ESD sensitive)

00:10 <azonenberg> Bird|ghosted: yeah there's a couple different constructions WRT orientation of layers or something

00:10 <azonenberg> but they're stupidly sensitive

00:10 <azonenberg> i forget exact numbers but they're in the double to triple digits HBM depending on design

04:23 <d1b2> <j4cbo> dang, that’s on par or worse than laser diodes

04:32 <azonenberg> Worse by far, afaik they're the most esd sensitive device in mass production

07:47 <d1b2> <theorbtwo> For what little it is worth, my crazy modular probe idea starts with a scope protector - input connector, output connector, some clamp diodes, and some LEDs past the clamp diodes that light up when you go over. (Hopefully with power via USB/5v for convenience, or 9v battery for not introducing noise or an unwanted ground.)

10:15 <_whitenotifier-5> [scopehal] azonenberg opened issue #405: Crash with pure virtual called during destructor when oscilloscope object is destroyed - https://git.io/JOeZZ

10:15 <_whitenotifier-5> [scopehal] azonenberg assigned issue #405: Crash with pure virtual called during destructor when oscilloscope object is destroyed - https://git.io/JOeZZ

10:15 <_whitenotifier-5> [scopehal] azonenberg labeled issue #405: Crash with pure virtual called during destructor when oscilloscope object is destroyed - https://git.io/JOeZZ

11:12 <_whitenotifier-5> [scopehal-apps] azonenberg opened issue #312: Save/restore dense pack flag in waveform metadata - https://git.io/JOeBG

11:12 <_whitenotifier-5> [scopehal-apps] azonenberg labeled issue #312: Save/restore dense pack flag in waveform metadata - https://git.io/JOeBG

11:12 <_whitenotifier-5> [scopehal-apps] azonenberg assigned issue #312: Save/restore dense pack flag in waveform metadata - https://git.io/JOeBG

11:13 <_whitenotifier-5> [scopehal-apps] azonenberg edited issue #312: Save/restore dense pack flag in waveform metadata - https://git.io/JOeBG

16:57 <_whitenotifier-5> [scopehal-pico-bridge] tdcox opened issue #1: feature request: Support for 3206MSO - https://git.io/JOvEJ

16:58 <_whitenotifier-5> [scopehal-pico-bridge] tdcox opened issue #2: feature request: Support for 4423 - https://git.io/JOvEq

20:29 <_whitenotifier-5> [scopehal] azonenberg pushed 4 commits to master [+2/-0/±9] https://git.io/JOvNy

20:29 <_whitenotifier-5> [scopehal] azonenberg 54ad0df - EyePattern: fixed crashes if clock has no toggles during the entire waveform

20:29 <_whitenotifier-5> [scopehal] azonenberg b6f40fd - Initial version of DramClockFilter

20:29 <_whitenotifier-5> [scopehal] azonenberg b2a7962 - EyePattern: improved handling of irregularly clocked data

20:29 <_whitenotifier-5> [scopehal] azonenberg b4142b4 - DramClockFilter: removed unused 'skip last' feature

20:40 <Degi> theorbtwo: Why would you need to boost to 300 V?

20:41 <Degi> If you want to clamp the voltage I'd suggest making an attenuator with HV caps / resistors and a spark gap

20:41 <Degi> (The spark gap needs to fire faster than the cap / resistor)

21:11 <d1b2> <theorbtwo> @degi, the idea is not to attenuate, but to clip at near the edges of the input range. I think I see where you are going, but I need a design that would also look for more sensitive scopes that would die at the merest mention of 300v.

21:21 <Degi> Hm, then at lower voltages you can just clip to two sink capable rails theorbtwo

21:21 <Degi> Otherwise for a rigol or so I'd add an attenuator since then you need less clipping rail voltage (or use GDTs) and can have higher input voltage

21:21 <azonenberg> yeah that's how we did things in the BLONDEL front end

21:21 <azonenberg> (clipping to sink capable rails)

21:22 <Degi> Some kinda attenuator either capable of high voltage or somehow has a series resistor or so which limits the max current would be nice for measuring capacitors

21:23 <Degi> (Such that the oscilloscope AFE doesn't disappear when suddenly the capacitor voltage appears across the scope input)

21:52 <d1b2> <theorbtwo> I'm thinking diff probe via ethernet transformer for that sort of thing, but I need to figure out what low frequency behavior I need to figure out what low frequency response looks like on those. Ideally, I want galvanic isolation which is flat except at actual DC, where it has no response. Clearly, however, that is completely impossible.

21:52 <Degi> You could de-embed it

21:53 <Degi> And use optocouplers or ADC/DAC for DC