00:01
<
azonenberg >
We do not have hislip support
00:01
<
azonenberg >
that is a pending ticket
00:01
<
azonenberg >
none of the currently supported instruments use it
00:01
<
balrog >
it seems that hislip is the preferred "future" standard?
00:02
<
azonenberg >
Lecroy gets 400+ Mbps of data with the right depths (tens of WFM/s) easily over scpi
00:02
<
azonenberg >
The demo scope is fast because it's spitting out data as fast as it can
00:02
<
azonenberg >
i think i actually have sleeps in there to cap it to no more than like 60 WFM/s or something
00:02
<
azonenberg >
so it doesnt peg your CPU
00:02
<
azonenberg >
but yeah the siglent spends most of its time waiting for data
00:02
<
azonenberg >
try wiresharking the traffic
00:03
<
azonenberg >
there's a lot of time where it sends a query and is twiddling its thumbs waiting for the scope to reply
00:03
<
azonenberg >
the highest performance scopes so far are the picoscope and other USB attached stuff
00:03
<
azonenberg >
Our bridge servers convert from that to SCPI control plane plus a second socket for data plane
00:03
<
azonenberg >
With that config and a picoscope, and a 10G NIC, I can push 3 Gbps of waveform data across the LAN to another machine
00:05
<
azonenberg >
Also the new picoscope GUI is night and day vs the previous rev
00:06
<
azonenberg >
their v7 software (which i just tried for the first time) is actually usable lol
00:06
<
azonenberg >
Still doesnt have all the fancy analysis of glscopeclient
00:06
<
azonenberg >
but it's not bad either
00:38
<
balrog >
how is the performance of the Digilent stuff?
00:39
<
balrog >
(though the sample rate is kinda low)
00:56
<
azonenberg >
WFM/s rate is excellent
00:56
<
azonenberg >
high double digits iirc
00:57
<
azonenberg >
maybe 40 or so?
00:57
Degi has quit [Ping timeout: 248 seconds]
00:57
<
azonenberg >
balrog: in general i've found that headless scopes perform very well
00:57
<
azonenberg >
because they were designed for that use case
00:57
<
azonenberg >
while streaming performance on stuff with a gui is an afterthought
00:57
<
balrog >
Yeah that makes sense...
00:58
<
balrog >
it would be nice if the vendors of the scopes with GUIs worked on improving that, once the devices have already shipped :)
00:58
<
azonenberg >
Siglent is trying
00:58
<
azonenberg >
i just don't know how successful they will be
00:59
Degi has joined #scopehal
00:59
<
azonenberg >
most of the big names are not, because glscopeclient competes with their $$$$ software options
01:19
<
azonenberg >
ook so tonight i think i am going to work on figuring out func gen support in glscopeclient and start thinking about AWG support
01:20
<
azonenberg >
ultimately what i want is to be able to generate an analog waveform in a filter graph
01:20
<
azonenberg >
then push to an AWG
01:20
<
azonenberg >
as a sink in the filter graph
01:20
<
azonenberg >
but i'm not sure how i would trigger it
01:20
<
azonenberg >
like. i dont want to update the arb every time the scope triggers or as fast as the filter graph can execute
01:22
<
azonenberg >
in any case, first step is going to be basic func gen support
01:22
<
azonenberg >
probably targeting pico and siglent
04:13
nelgau_ has joined #scopehal
04:15
nelgau has quit [Ping timeout: 256 seconds]
04:45
<
_whitenotifier-9 >
[scopehal-pico-bridge] azonenberg ae418ed - Initial implementation of function generator (everything up to but not including full arb) support
05:01
<
azonenberg >
PRBS generated by PicoScope 6824E's integrated function generator, sampled by one of my lecroy scopes (to get a better view of jitter, I didn't want to use the same timebase to sample)
05:01
<
azonenberg >
as of now the scpi command set for the bridge is done, now i have to add the clientside APIs
05:07
<
XMPPwocky >
anybody happen to have a nix expression for scopehal-apps? it looks straightforwards enough to write, i just don't want to if I don't have to
08:00
<
_whitenotifier-9 >
[scopehal] azonenberg 6982500 - Pico: added preliminary APIs for function generator. No AWG support. Not tested as there's no GUI in glscopeclient to call them yet. See #593.
16:37
<
_whitenotifier-9 >
[scopehal-apps] azonenberg 7fbfad6 - Refactoring: Removed ScopeApp::m_scopes as it duplicates OscilloscopeWindow::m_scopes
16:52
bvernoux has joined #scopehal
17:11
<
_whitenotifier-9 >
[scopehal-apps] azonenberg aa04468 - Oscilloscope: find function generators that are part of our scopes
20:48
bvernoux has quit [Read error: Connection reset by peer]