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
Degi_ has joined #scopehal
Degi has quit [Ping timeout: 272 seconds]
Degi_ is now known as Degi
d1b2 has quit [Remote host closed the connection]
d1b2 has joined #scopehal
d1b2 has quit [Remote host closed the connection]
d1b2 has joined #scopehal
d1b2 has quit [Remote host closed the connection]
d1b2 has joined #scopehal
d1b2 has quit [Remote host closed the connection]
d1b2 has joined #scopehal
d1b2 has quit [Remote host closed the connection]
juli966 has quit [Quit: Nettalk6 - www.ntalk.de]
d1b2 has joined #scopehal
d1b2 has quit [Remote host closed the connection]
d1b2 has joined #scopehal
d1b2 has quit [Remote host closed the connection]
d1b2 has joined #scopehal
d1b2 has quit [Remote host closed the connection]
d1b2 has joined #scopehal
electronic_eel_ has joined #scopehal
electronic_eel has quit [Ping timeout: 272 seconds]
<azonenberg> sooo i just realized
<azonenberg> since MAXWELL is going to be professionally assembled now, fitting inside my toaster oven is no longer a design requirement
<azonenberg> Which means i can probably add a little "tail" to the left side of the board
<azonenberg> and allow the LCD to mate directly with the board rather than needing a FFC adapter
<azonenberg> that should simplify things a bunch
<_whitenotifier-f> [scopehal-docs] azonenberg pushed 2 commits to master [+0/-0/±4] https://git.io/JkDI8
<_whitenotifier-f> [scopehal-docs] azonenberg e334ed3 - Added skeleton documentation for PCIe decodes
<_whitenotifier-f> [scopehal-docs] azonenberg 26841ef - Merge branch 'master' of github.com:azonenberg/scopehal-docs
<_whitenotifier-f> [scopehal] azonenberg pushed 1 commit to master [+0/-0/±2] https://git.io/JkDn7
<_whitenotifier-f> [scopehal] azonenberg 5bca82a - PCIeTransportDecoder: Can now decode memory reads/writes. Still have to do completions and message requests. See #35.
<azonenberg> miek: FYI i added code to the glscopeclient USB stack the other day to check CRCs
<azonenberg> that had been a pending issue for a while and i wanted to sanity check some stuff
bvernoux has joined #scopehal
<azonenberg> bvernoux: hey, any interest in writign a glscopeclient importer for the rigol data files?
<bvernoux> for bin files ;)
<bvernoux> yes it could be interesting
<bvernoux> do you have a template or something existing to see how it works ?
<bvernoux> For information on Windows when using my Rigol MSO5000 with Ethernet TCP/IP it never works
<bvernoux> glscopeclient is always frozen ...
<bvernoux> even when I set something like 10kpts on the scope with simple trigger armed
<bvernoux> I suspect it is an issue related to windows as it seems to works with Linux
<azonenberg> interesting
<azonenberg> file a bug so other people know, then investigate when you have time?
<bvernoux> SO yes importer will be nice to use it not in live anyway
<azonenberg> and a good starting point would be looking at the CSV import logic
<azonenberg> basically you need to add loading logic to OscilloscopeWindow, then the file format parser to MockOscilloscope
<bvernoux> ha yes CSV is the first importer IIRC
<bvernoux> so far I have very little time to work on that
<bvernoux> but I will document my reverse and provide 010 Editor bt to parse Rigol MSO5k bin files
<azonenberg> File a feature request ticket for it
<bvernoux> yes
<azonenberg> even if you cant work on it right away
<azonenberg> and if you give some reference files maybe i can work on it at some point
<bvernoux> the best will be wfm (waveform)
<bvernoux> but I have not reversed it so far
<bvernoux> As in bin file there is less details related to each channels ...
<bvernoux> like channel name are not part of bin file
<bvernoux> it is not very important but can be convenient ....
<bvernoux> let's fill a feature request ;)
juli966 has joined #scopehal
<_whitenotifier-f> [scopehal] azonenberg pushed 1 commit to master [+0/-0/±2] https://git.io/JkyVq
<_whitenotifier-f> [scopehal] azonenberg c674b12 - PCIeTransportDecoder: Added completion request parsing. Fixes #35.
<_whitenotifier-f> [scopehal] azonenberg closed issue #35: Add PCIe protocol decoder - https://git.io/JkyVm
<azonenberg> lain, monochroma: ^
<lain> :3
<monochroma> :D
<_whitenotifier-f> [scopehal-apps] bvernoux opened issue #287: Add support import for Rigol MSO5000 Wave *.bin file - https://git.io/Jkyw4
<bvernoux> done ;)
<_whitenotifier-f> [scopehal-apps] bvernoux edited issue #287: Rigol MSO5000 Wave *.bin file import - https://git.io/Jkyw4
<_whitenotifier-f> [scopehal-apps] bvernoux edited issue #287: Rigol MSO5000 Wave *.bin file import - https://git.io/Jkyw4
<_whitenotifier-f> [scopehal-apps] azonenberg opened issue #288: PCIe gen2 compliance testing wizard - https://git.io/JkywP
<_whitenotifier-f> [scopehal-apps] azonenberg opened issue #289: PCIe gen1 compliance testing wizard - https://git.io/JkywX
<miek> bvernoux: looks familiar... https://i.imgur.com/o1p5iRZ.png :)
<bvernoux> ha yes very interesting
<bvernoux> could you provide some bin ?
<bvernoux> maybe my parser work with your files ;)
<bvernoux> MSO6034
<bvernoux> it is Rigol DS6000 ?
<miek> agilent
<bvernoux> haha
<bvernoux> They have rebadged a Rigol Scope ;)
<miek> the keysight 1000X series are rebadged rigols afaik, and i think rigol has just kept a lot of stuff since that collaboration (i noticed the SCPI commands are very similar too)
<azonenberg> BTW folks, there is talk of forking the siglent driver off from lecroy
<azonenberg> now that i'm doing more advanced stuff in the lecroy driver with automation commands a lot of stuff doesnt work on siglent anymore
<azonenberg> even though the base command set is very similar from previous collaboration/rebadging
<azonenberg> if anybody has a siglent scope and is interested in working on that, let me know
<azonenberg> Error_404: this means you for example :P
<bvernoux> miek, what is the BW of your MSO6034 ?
<miek> bvernoux: and yeah, i'll upload some bins when i get a chance
<miek> 300MHz
<bvernoux> ha ok
<bvernoux> So importer for Rigol MSO5000 could maybe support also your Agilent MSO6k ;)
<miek> probably worth having a look at some of the agilent/keysight docs on their bin format - it'll probably all match up
<miek> indeed :)
<bvernoux> if there is only signature at start which change ;)
<bvernoux> AG vs RG ;)
<bvernoux> what a joke ;)
<miek> i used this python script the other day on mine: https://community.keysight.com/thread/38064
<bvernoux> miek, it seems you was doing some reverse on some firmware/rom in 010Editor haha
<bvernoux> on my side I have done that from scratch
<bvernoux> just analyzing lot of files with diff ;)
<bvernoux> so there is some fields unknown but nothing important
<miek> it just might help you with the unknowns
<bvernoux> ha yes great
<bvernoux> the most interesting will be to reverse the native Waveform *.wfm
<bvernoux> as they are more compact and contains probably more details
<bvernoux> I suspect they use RLE or something like that
<bvernoux> as they are smaller than bin
<bvernoux> I plan to write a native C parser/converter
<bvernoux> I hate python for that stuff ;)
<bvernoux> I prefer something native like C as it is rock solid and portable on all platforms ;)
<bvernoux> too fun InfiniiVision_binary_to_csv.py
<bvernoux> works with Rigol MSO5000 bin files ;)
<azonenberg> Well if we can make an importer that can work with rigol *and* keysight bin files
<azonenberg> even better
<monochroma> sounds like only /some/ keydight files though perhaps?
<monochroma> keysight*
<bvernoux> yes
<bvernoux> we can change the title ;)
<bvernoux> of the issue
<bvernoux> miek, it means you can fully unlock your Agilent Scope too ;)
<bvernoux> using the hack for Rigol MSO5000 ;)
<bvernoux> could you take some screenshot of Utility->System->About ?
<miek> my scope was released in 2005, it's not based on a rigol :p
<bvernoux> ha it is the reverse ;)
<bvernoux> Agilent has sold it to Rigol ;)
<bvernoux> They are using Qt5.5 ;)
<bvernoux> It is why the GUI is very not bad
<miek> i think what happened it keysight gave a spec to rigol for the 1000x series rebadges, and then rigol has just kept building off that on their own gear
<bvernoux> The only drawback on those Rigol MSO5000 is the data transfer over Ethernet... is too slow
<bvernoux> miek, ha yes they have done the work for them and they have reused for their new scope ;)
<d1b2> <TiltMeSenpai> is it only data transfer over ethernet that sucks?
<bvernoux> Does Agilent provide python script to decode *.wfm too ?
<bvernoux> yes data transfer over ethernet is too slow
<bvernoux> also the scope is not ultra reactive when there is 4 chan at same time ;)
<d1b2> <TiltMeSenpai> well I know ethernet sucks, but I have a raspberry pi lying around
<bvernoux> but with 8GSPS it is normal I think ;)
<d1b2> <TiltMeSenpai> if usb doesn't suck, I can throw the raspberry pi into the mix to proxy better speeds
<bvernoux> it is same over USB
<bvernoux> issue is mainly in how the scope send data and is probably synchronized with GUI ...
<d1b2> <TiltMeSenpai> GPIB or whatever the weird parallel thing is called?
<bvernoux> I will like to switch in PC mode without GUI if it could send waveform 10x faster ;)
<bvernoux> no it is native Gigabit Ethernet on Rigol MSO5000
<d1b2> <TiltMeSenpai> I've noticed that the ethernet transfer rates suck but I haven't had the motivation to find a faster way to send data
<bvernoux> so it is clearly not related to GigabitEthernet which in theory can push 100MBytes/s
<miek> bvernoux: my scope doesn't seem to save to *.wfm or anything similar
<bvernoux> when the scope struggle to send 1Mpoints in less than 1s ...
<bvernoux> miek, ha you have only *.bin or *.csv ?
<bvernoux> miek, do you have LA too ?
<miek> these are the options: https://i.imgur.com/6EwOlnk.png
<miek> yup
<bvernoux> miek, I have designed my own LA pod which can be pushed to the limit in theory 500MHz ;)
<bvernoux> I will test it with ultra fast signals ;)
<bvernoux> ha yes it is very ugly vs Rigol MSO5000 GUI ;)
<bvernoux> so definitely an older stuff
<bvernoux> the GUI is totally different here
<bvernoux> like you can see in screenshot I have added with the bin
<bvernoux> I have bought Fast risetime pulse generator (2.92mm output) ;)
<bvernoux> Will be nice to check limit of lot of things ;)
<bvernoux> I need to build a tool to output 16 chan DSO (0 to 3.3V)
<bvernoux> with different pattern
<bvernoux> which can go up to 1GSPS ;)
<bvernoux> anyway the LA is limited to 1GSPS
<bvernoux> so up to 500MSPS will be fine to check the limit
<bvernoux> so maybe let's keep bin waveform ;)
<bvernoux> we have a common format at least for Agilent MSO6xxxx & Rigol MSO5000
<bvernoux> I'm pretty sure other scope have same format derivated from Rigol/Agilent bin
<miek> i would expect it to extend across most of the agilent/keysight range, they seem pretty keen on keeping compatibility
<bvernoux> InfiniiVision_binary_to_csv.py does not work fine for LA channel especially the 2nd chann ;)
<bvernoux> but it recognize the 1st
<bvernoux> yes very nice if it is a common format for lot of scope
<bvernoux> maybe it is a standard behind ?
<bvernoux> Anyway your InfiniiVision_binary_to_csv.py bring me lot of details on things I have not found ;)
<bvernoux> let's try to decode some buggy bin files ;)
<bvernoux> I have found some bugs in Rigol MSO5000 cases with 1Chan 8GSPS and 16 DSO ;)
<bvernoux> also 1CH Analog 8GSPS with LA D7-D0 does not work fine to export bin
<bvernoux> the LA chan data are truncated ...
<bvernoux> I want to fill a bug report to rigol support/r&d for that
<bvernoux> but I need to do more tests to understand all the buggy cases ;)
<bvernoux> That confirm my asumption
<miek> that script might be failing because it uses the channel name to create the output filename
<miek> but the rigol doesn't have any
<bvernoux> the file size is 4.29MB and in the file it is quoted it shall be 8000164bytes ;)
<bvernoux> so there is some missing bytes ;)
<bvernoux> yes Channel Name are always empty
<bvernoux> it seems to be a bug too ;)
<bvernoux> except for LA
<bvernoux> chan name is LA ;)
<bvernoux> I have updated the 010 Editor script ;)
<bvernoux> so far the different with Rigol is only on LA Waveform
<bvernoux> which are a bit strange and specific
<bvernoux> with 16bits data ...
<_whitenotifier-f> [scopehal-apps] azonenberg pushed 1 commit to master [+0/-0/±2] https://git.io/JkybW
<_whitenotifier-f> [scopehal-apps] azonenberg 61063cd - Updated submodules
electronic_eel_ is now known as electronic_eel
deltab has quit [Ping timeout: 272 seconds]
deltab has joined #scopehal
bvernoux has quit [Quit: Leaving]