azonenberg changed the topic of #scopehal to: libscopehal, libscopeprotocols, and glscopeclient development and testing | https://github.com/glscopeclient/scopehal-apps | Logs: https://libera.irclog.whitequark.org/scopehal
<d1b2> <Code and Solder> In any case I'm reasonably optimistic about the Chinese option, I got an SMA-SMPM male adapter to do VNA measurements
<d1b2> <Code and Solder> The NanoVNA being a bit of a limiting factor but I don't currenty have any excuse to make my work buy something better or budget to get it on my own
<azonenberg> 2.92 to SMA will work reasonably well, yes. It's just that SMA tends to be made to looser tolerances
<azonenberg> and mating to a bad one can easily damage a 2.92 which is a bit more fragile being air dielectric
Degi has quit [Ping timeout: 260 seconds]
Degi has joined #scopehal
<_whitenotifier> [scopehal-apps] azonenberg closed pull request #548: Don't show impossible-to-enable channels in filter graph editor - https://github.com/glscopeclient/scopehal-apps/pull/548
<_whitenotifier> [scopehal-apps] azonenberg pushed 2 commits to master [+0/-0/±2] https://github.com/glscopeclient/scopehal-apps/compare/e335b92d6c04...3d6e000dc93d
<_whitenotifier> [scopehal-apps] 602p 30c8366 - Don't show impossible-to-enable channels in filter graph editor
<_whitenotifier> [scopehal-apps] azonenberg 3d6e000 - Merge pull request #548 from 602p/ngscope_dont_show_unavailable_in_filtergraph Don't show impossible-to-enable channels in filter graph editor
<_whitenotifier> [scopehal-apps] azonenberg closed pull request #546: Move towards unifying on-connect behaviour in glscope with ngscope - https://github.com/glscopeclient/scopehal-apps/pull/546
<_whitenotifier> [scopehal-apps] azonenberg pushed 2 commits to master [+0/-0/±2] https://github.com/glscopeclient/scopehal-apps/compare/3d6e000dc93d...ee0f90a73ada
<_whitenotifier> [scopehal-apps] 602p fee6cce - Move towards unifying on-connect behaviour with ngscope. Enable only those channels already enabled on connect, except for: - On headless scopes, enable all analog channels. Or, if no analog, all digital. - On regular scopes, if no channels enabled, enable the first.
<_whitenotifier> [scopehal-apps] azonenberg ee0f90a - Merge pull request #546 from 602p/glscope_respect_enabled_channels Move towards unifying on-connect behaviour in glscope with ngscope
<d1b2> <Code and Solder> Looking through potential power connectors it seems the smallest commonly available pitch for wire-to-board is 0.8mm, 28-32AWG, 1-2A, 3.8x3.2mm board space, for example MOLEX Zero-Hachi: https://www.molex.com/pdm_docs/sd/2147210020_sd.pdf
<d1b2> <Mac> I saw some discussion on DPI earlier, this is the answer (I wrote the section that starts with "On Windows") https://github.com/ocornut/imgui/blob/master/docs/FAQ.md#q-how-should-i-handle-dpi-in-my-application what was in louis' PR is basically correct, just need to pull the scale value from the OS instead.
<azonenberg> mac: Yeah exactly
<azonenberg> its the right idea but wrong implementation
<azonenberg> 3.8 x 3.2mm is actually a touch *smaller* than a u.fl
<azonenberg> interesting
<azonenberg> my u.fl footprint is about 4x4
<d1b2> <Code and Solder> MHF 4L is 2.5x2.4 footprint
<d1b2> <Code and Solder> But I'd prefer something that can be comfortably unplugged without a tool I think, even if a milimeter bigger
<d1b2> <Code and Solder> Other models down to 2x2mm: https://www.i-pex.com/products/filter?field_category_target_id=4
<d1b2> <Code and Solder> By the way, are you working on a power supply for the probes? I'm thinking about designing something nice and convinient powered from USB-C with USB-PD with one LT3045-1 per channel probably
<d1b2> <Code and Solder> I'm torn between making it modular for different voltages or some easy adjustment
<azonenberg> So i am definitely interested in designing a supply. one of the challenges is that currently the AV1 and AD4 need different voltages
<azonenberg> both will survive either voltage but the AD4 prefers a lower input IIRC to minimize dissipation in the LDOs, since it runs hot already
<azonenberg> (it's the same power stage - an inverting charge pump + LDO for the negative and an LDO for the positive - for both. Just different feedback resistors for different output voltages)
<d1b2> <Code and Solder> As long as everything is compatible a button/switch to cycle between presets seems like the easiest option
<d1b2> <Code and Solder> RFID tag on the probe with the preset to reconfigure by touch is probably overkill
<azonenberg> It's short term compatible
<azonenberg> without instant hardware damage
<azonenberg> If you give the AV1 the lower voltage meant for the AD4, it won't work right
<azonenberg> if you give the AD4 the higher voltage meant for the AV1, it will probably overheat because the LDOs are dropping more voltage than they were sized for
<d1b2> <Code and Solder> Profiling based on I/V curve should be possible, in theory
<d1b2> <Code and Solder> Step voltage from 0 to like 3V recording current and decide based on that
<azonenberg> Yeah thats a possibility
<azonenberg> The other is to have the two probes use incompatible-by-design connectors
<azonenberg> and apply fixed voltages to each
<d1b2> <Code and Solder> That would require swapping cables when changing probes though, and avoiding that seems neater, especially with RF cable being the same