azonenberg changed the topic of #scopehal to: libscopehal, libscopeprotocols, and glscopeclient development and testing | https://github.com/azonenberg/scopehal-apps, https://github.com/azonenberg/scopehal, https://github.com/azonenberg/scopehal-docs | Logs: https://freenode.irclog.whitequark.org/scopehal
<azonenberg> tentacle probe holder, lol
<monochroma> lewd.
<azonenberg> monochroma: ...
Degi has quit [Ping timeout: 260 seconds]
Degi has joined #scopehal
<azonenberg> monochroma: https://www.amabileelectronictest.com/pdf/aet_pdf_1064.pdf page 2 check out that graphic
<azonenberg> early wavelink differential probe, open frame without all the plastic trim to make it look pretty
<azonenberg> looks like metal tweezer-style handpiece, flex pcb for the active stuff w/ rigid at the probe tips
<monochroma> ohhh yeah, i've seen that before
<monochroma> reminds me of a drafting compass
<azonenberg> Yeah. in fact, i bet we could turn one into a probe this way without too much work lol
<azonenberg> just take out the needles and mount a flex pcb to it
<azonenberg> But i'm not doing any new probe designs any time soon. I already have 3 different ones in progress
<monochroma> hehehe
<azonenberg> Random interesting data point: I think the SDA 8Zi-B uses a Stratix V
<azonenberg> The hardware serial trigger is 6.5 Gbps and can be unlocked to 14.1 Gbps
<azonenberg> that is a very unique and unusual SERDES line rate
<azonenberg> stratix v is the only fpga i know of that has an upper limit of exactly 14.1 Gbps
electronic_eel has quit [Ping timeout: 260 seconds]
electronic_eel has joined #scopehal
<_whitenotifier-f> [scopehal] azonenberg pushed 2 commits to master [+0/-0/±5] https://git.io/JTfaZ
<_whitenotifier-f> [scopehal] azonenberg ef086e7 - Removed temporary debug message
<_whitenotifier-f> [scopehal] azonenberg 7425fd2 - Added Oscilloscope::GetChannelBandwidthLimiters() and implemented for LeCroy. Fixes #268.
<_whitenotifier-f> [scopehal] azonenberg closed issue #268: Add APIs for querying available bandwidth limits on an instrument - https://git.io/JUal4
<_whitenotifier-f> [scopehal] azonenberg pushed 1 commit to master [+0/-0/±1] https://git.io/JTfwA
<_whitenotifier-f> [scopehal] azonenberg 214b29e - LeCroyOscilloscope: fixed bug when setting bandwidth limits >= 1 GHz
<_whitenotifier-f> [scopehal-apps] azonenberg pushed 2 commits to master [+0/-0/±6] https://git.io/JTfr4
<_whitenotifier-f> [scopehal-apps] azonenberg 6335a1a - ChannelPropertiesDialog: added control for bandwidth limiters
<_whitenotifier-f> [scopehal-apps] azonenberg 71e3f22 - Removed bandwidth limiter controls from context menu to reduce clutter, since they're already in the properties dialog. Fixes #193.
<_whitenotifier-f> [scopehal-apps] azonenberg closed issue #193: Remove bandwidth limit from context menu and move to channel properties dialog - https://git.io/JUalG
<_whitenotifier-f> [scopehal] azonenberg pushed 1 commit to master [+0/-0/±1] https://git.io/JTf1m
<_whitenotifier-f> [scopehal] azonenberg 67c4b11 - LeCroyOscilloscope: ensure proper left alignment when changing memory depth
<_whitenotifier-f> [scopehal] azonenberg pushed 1 commit to master [+0/-0/±1] https://git.io/JTfDm
<_whitenotifier-f> [scopehal] azonenberg 619998e - LeCroyOscilloscope: fixed bug in HDO9K sample rate table
<_whitenotifier-f> [scopehal] azonenberg pushed 2 commits to master [+0/-0/±4] https://git.io/JTfd2
<_whitenotifier-f> [scopehal] azonenberg 51a688e - MDIODecoder: Added colorization, some improvements to decoding, and merging of MMD register transactions
<_whitenotifier-f> [scopehal] azonenberg 5ef7254 - MDIODecoder: added PHY type info. Added decodes for most common KSZ9031 MMD registers plus a lot more standard registers. Fixes #82.
<_whitenotifier-f> [scopehal] azonenberg closed issue #82: MDIO decoder: allow specification of one or more (phy address, PHY ID) mappings - https://git.io/JeHFg
<azonenberg> monochroma: you might find this helpful with some of your projects https://www.antikernel.net/temp/mdio-ksz9031.png
<azonenberg> (also lukego)
<Degi> Huh, the diff probe is interesting
<Degi> https://imgur.com/2vYYTSv.png Are those vias in the BGA and whats up with the BGA pads
<azonenberg> those are un-tented dogbone vias
<azonenberg> being probed from the back of the board
<azonenberg> it looks like the board is flex-rigid, fr4 section for mounting the probe tips then flex the rest of the way
<Degi> Ah, they're the backside, that makes sense
<azonenberg> what do you want to use that for?
<Degi> Idk looks fon
<Degi> *fun
<Degi> From 62.5 to 32000 MHz
bvernoux has quit [Quit: Leaving]
<azonenberg> Looks like the experimental TPU probe shell will be here tuesday or so
azonenberg_work has quit [Ping timeout: 256 seconds]
<monochroma> azonenberg: oh nice @ MDIO decode
juli965 has joined #scopehal
azonenberg_work has joined #scopehal
katharina has joined #scopehal
<katharina> azonenberg: I am going to develop the prefs system in two phases, initially i want the tree to work (it does), and have the rudimentary GUI working for bools, reals and strings.
<katharina> Afterwards, im going to implement support for units, colors and enums.
<katharina> I leave it up to you whether you want to merge after phase 1, or after everything is done, i just need the feedback on that
juli965 has quit [Quit: Nettalk6 - www.ntalk.de]
bvernoux has joined #scopehal
<azonenberg> Katharina: great
<azonenberg> Do you have an estimate for when phase 1 will be done?
<azonenberg> phase 1 completeness is enough that we can start adding some of the actual prefs there are open tickets for
<katharina> azonenberg: im planning for sunday evening, but ofc cant give a guarantee
<azonenberg> Of course, none expected. Just wanted some idea of how close you were
<azonenberg> In case you didn't see BTW, https://www.antikernel.net/temp/mdio-ksz9031.png
<azonenberg> MDIO protocol analyzer got a big facelift
<azonenberg> Katharina: also i'm working on something right now where the serial redaction pref would have been handy lol
<azonenberg> Doing dev on a borrowed instrument over VPN, the donor would prefer to be anonymous
<azonenberg> right now i have to manually censor each screenshot i share in gimp
<azonenberg> so almost perfect timing for you on having the system close to usable now lol
<katharina> azonenberg the facelift is excellent, nice job
<azonenberg> The new "color coding and collapsing of related packets" has been around for some time but i've been going back and retrofitting it into older decodes when i had to do work on them for other reasons
<katharina> BTW, do we have a CAN bus decoder yet?
<azonenberg> The ticket is still open. There is a decode but it's bare bones and the guy who contributed it wanted to add more stuff before closing the ticket, i forget what
<azonenberg> i havent heard from him in months and don't have any hardware handy that speaks CAN
<azonenberg> In general i've been filing tickets any time i think of a protocol that would be nice to have, and implementing stuff any time i had a board handy that used said protocol
<katharina> Maybe after the UI work, Ill work on that. Would be handy to have. Im currently building a simulator cockpit for flight sim, and my components are communicating via can bus
<katharina> and my scopes decode is super bad
<azonenberg> I havent been working on CAN myself due to lack of test data
<azonenberg> Speaking of which...
<_whitenotifier-f> [scopehal] azonenberg opened issue #300: SAE J1939 protocol decode - https://git.io/JTUqd
<_whitenotifier-f> [scopehal] azonenberg labeled issue #300: SAE J1939 protocol decode - https://git.io/JTUqd
<azonenberg> But yeah, any contributions from you are most welcome :)
<azonenberg> also as a reminder the "windows portability fixes" ticket is still open, the last activity was me commenting for you to close it once you feel the windows build is generally on par with the linux one
<_whitenotifier-f> [scopehal] azonenberg opened issue #301: Add Windows support for USBTMC transport - https://git.io/JTUme
<_whitenotifier-f> [scopehal] azonenberg labeled issue #301: Add Windows support for USBTMC transport - https://git.io/JTUme
<katharina> good catch
<azonenberg> just broke #301 out to a separate ticket so lack of usbtmc support should not be considered a blocker for it
<katharina> Ill take a look at it later, we should have most features working (on a conceptual level, ofc there might be bugs)
<azonenberg> Yeah but small stuff should be separate tickets
<azonenberg> that one was intended to be "doesnt compile or work at all on windows"
<azonenberg> and i think between you and bvernoux that's done
<katharina> then ill close it
<katharina> it definitely reliably compiles and runs
<_whitenotifier-f> [scopehal-apps] nshcat closed issue #113: Windows portability fixes - https://git.io/Jfbz4
<azonenberg> (there's also one for libscopehal)
<azonenberg> #38 i think?
<_whitenotifier-f> [scopehal-apps] nshcat commented on issue #113: Windows portability fixes - https://git.io/JTUmu
<azonenberg> also re your flight sim setup why not go all the way and hook up components with ARINC 429 or MIL-STD-1553? *hint hint*
<azonenberg> (in all seriousness i want to add support for them and dont have any avionics hardware to get test signals from right now...)
<katharina> haha
<smkz> isnt arinc 420 or whatever just boneless CAN
<katharina> i mean technically it would be possible, i'd need to use my RISCV Softcore CPU, implement MIL-STD-1553 hardware, and each instrument panel would need to use an FPGA
<azonenberg> smkz: it's much higher voltage, both are like 20+ V signals in order to have better toleranec to EMI
<azonenberg> i dont know much about the protocol itself
<azonenberg> i havent looked
<katharina> but im happy i can use the STM32 chips i use, they come with CAN and USB.
<smkz> oic
<azonenberg> I did actually work with 429 on some $BIG_AIRLINER_MANUFACTURER hardware ages ago, i was doing a pentest of a particular aircraft component made by a subcontractor
<azonenberg> But my testing was not targeting the 429 side, but the $OTHER_BUS interface
<azonenberg> so i pretty much ignored the 429
<electronic_eel> didn't the aerospace industry move to a modified variant of ethernet some time ago?
<electronic_eel> at least for newer stuff
<azonenberg> that was the other portability ticket
<azonenberg> electronic_eel: would not surprise me, especially if fiber since that's emi-proof
<_whitenotifier-f> [scopehal] nshcat closed issue #153: Windows portability fixes - https://git.io/Jfbz8
<azonenberg> The gizmo i worked with actually had 429 and ethernet on it
<azonenberg> ethernet was the other bus i was targeting
<_whitenotifier-f> [scopehal] nshcat commented on issue #153: Windows portability fixes - https://git.io/JTUmA
<azonenberg> but i dont remember (and couldnt say if i did) what that ethernet interface connected to
<azonenberg> if it was avionics or IFE or just for ground level debug or what
<katharina> how about a SpaceWire decoder? ;P
<katharina> okay that one might be secret, im not sure tho
* monochroma worked on a spacewire debug interface
<_whitenotifier-f> [scopehal] azonenberg opened issue #302: SpaceWire protocol decode - https://git.io/JTUYG
<_whitenotifier-f> [scopehal] azonenberg labeled issue #302: SpaceWire protocol decode - https://git.io/JTUYG
<katharina> hahaha nice
<azonenberg> lecroy has a decode for it, cant be too secret
<katharina> i guess getting waveform dumps of spacewire might be next to impossible
<azonenberg> bluecmd[m]: hey, so were you interested in implementing CSV import?
<azonenberg> and/or VCD?
<azonenberg> Katharina: that is the blocker for several of my current decode tickets
<katharina> electronic_eel i am not sure if its modified ethernet, but it does use fiber: https://de.wikipedia.org/wiki/STANAG_3910
<azonenberg> 429, 1553, SPDIF, J1939, FlexRay, at least are blocked on lack of test data
<katharina> it uses MIL-STD-1553 to negotiate
<katharina> sorry for the german link, oops
<azonenberg> 1-Wire and I2S have had test data contributed to us in Saleae(?) CSV import, so we will need to write an importer for that before we can make a decode
<electronic_eel> I'm german, no problems
<azonenberg> Katharina: yeah we probably have almost as many germans as americans in this channel :P maybe not quite, but it seems to be more than any other nationality besides US
<electronic_eel> but this STANAG thing seems to be mostly military. I thought the commercial airliners had some ethernet thing
<katharina> azonenberg: thats very interesting
<azonenberg> and then several others very close by, like bluecmd[m] i think is Swiss
<monochroma> spacewire is an ESA standard, so iirc it's pretty open
<azonenberg> monochroma: yeah the challenge will be getting a satellite or part thereof to get signals from :p
<sorear> does it have to be a real assembled satellite? can't just set up a dummy source with the grlib code or whatever?
<electronic_eel> AFDX seems to be what I was remembering: https://en.wikipedia.org/wiki/Avionics_Full-Duplex_Switched_Ethernet
<monochroma> azonenberg: that's not too hard, iirc there is enough ground equipment that uses it :P
<azonenberg> sorear: anything that you're reasonably confident speaks spacewire would work. But of course this isnt a high priority as i have no immediate use for it
<azonenberg> the decodes i spend the most time on are ones that i intend/expect to use at work or on personal projects
<azonenberg> also i just found a SATA-to-usb3 adapter that i might try and see if i can get signals off of somehow
<azonenberg> there's a ticket for SATA decode that might be fun to work on at some point. if nothing else getting test data will let somebody else work on it
<sorear> and you don't have stuff at home that uses 1355 huh
<katharina> azonenberg hows your opensource scope project getting along btw?
<electronic_eel> maybe one day ktemkin gets too annoyed of her usb3 analyzer and she adds some advanced usb decoders to scopehal
<azonenberg> electronic_eel: lol
<azonenberg> Katharina: the scope stuff is on hold for a while. the 96 channel LA is slowly moving along
<azonenberg> i spent a while working on probes and just received a prototype run of 25 active 8-bit probe podss
<azonenberg> (MEAD)
<azonenberg> Waiting on the uni i'm collaborating with to pay me for their half of those before i finish and order the main logic board
<azonenberg> because i can't afford to front that much cash out of pocket
<sorear> 96 channels? is that for pcix or something?
<azonenberg> sorear: their immediate use case is sniffing DDR RAM. like, the entire bus
<sorear> 96 channels and multiple GT/s?
<azonenberg> they have an interposer set up with probe points on each DQ/A/control pin but need a LA with enough pins to crunch the data
<azonenberg> The inputs are in theory capable of 5 Gsps sample rates
<azonenberg> although i may not be able to run all of them that fast due to internal RAM write bandwidth limitations
<azonenberg> the rightmost four are 10 Gsps
<azonenberg> There's a reason this beast has a 40GbE pipe to the outside world, lol
<sorear> 40 Gb/s is low compared to modern DDR
<azonenberg> Yes, of course. You wont get realtime streaming
<azonenberg> data coming off all of the probes at full speed is about half a Tbps
<azonenberg> it's going to get compressed down to around 100 Gbps and buffered in the internal DDR3
<azonenberg> then pulled off at 40 Gbps to glscopeclient
<_whitenotifier-f> [scopehal-apps] azonenberg labeled issue #223: FilterDialog: allow "no file name" to be specified (add "x" button to clear the current filename) - https://git.io/JTUC4
<_whitenotifier-f> [scopehal-apps] azonenberg opened issue #223: FilterDialog: allow "no file name" to be specified (add "x" button to clear the current filename) - https://git.io/JTUC4
<_whitenotifier-f> [scopehal] azonenberg opened issue #303: FilterParameter: add hint to allow filename type to be specified as input or output - https://git.io/JTUC0
<_whitenotifier-f> [scopehal] azonenberg labeled issue #303: FilterParameter: add hint to allow filename type to be specified as input or output - https://git.io/JTUC0
<_whitenotifier-f> [scopehal] azonenberg pushed 1 commit to master [+0/-0/±2] https://git.io/JTU8i
<_whitenotifier-f> [scopehal] azonenberg 71a3304 - EthernetProtocolDecoder: now optionally write packets to a PCAP file/pipe. Fixes #57.
<_whitenotifier-f> [scopehal] azonenberg closed issue #57: Ethernet: add PCAP export - https://git.io/JTU8P
<azonenberg> monochroma, lain, lukego: https://www.antikernel.net/temp/glscopeclient-pcap.png
<monochroma> :D
<lain> nice!
<azonenberg> Live streaming via pipe too, not just files
<monochroma> oooo!
<azonenberg> As of now it's unidirectional, there's no support for merging flows from tx and rx decodes
<azonenberg> that can come later
<sorear> glscopeclient is generating pcap and streaming it to wireshark?
<azonenberg> sorear: yes. Live, every trigger it flushes the pipe and you get new packets in wireshark
<azonenberg> This currently works for all supported PHYs: 10baseT, 100baseTX, 1000base-X, 10Gbase-R
<azonenberg> as well as RGMII
<azonenberg> and GMII
<azonenberg> Although of course instead of a pipe you can also write to a normal pcap file and do offline analysis
<azonenberg> but the intent here is to use wireshark as a protocol decode on steroids
<azonenberg> while doing only the PHY level decode in glscopeclient
<katharina> thats very neat
<katharina> azonenberg: do you think preferences should store what their min/max values are?
<katharina> and maybe something like stepping for reals?
<azonenberg> Katharina: you mean legal ranges? hmm, that would be a useful thing to add
<azonenberg> right now filter parameters have no support for that but it might be handy to add there too
Katharina_ has joined #scopehal
katharina has quit [Ping timeout: 260 seconds]