<azonenberg>
My stereo uses bare wires and binding posts
<azonenberg>
ooook so back to on topic
<azonenberg>
as of now we have basically all config saved
<azonenberg>
I think it's time to implement config loading
<azonenberg>
And then i'll shift focus to waveform data
<azonenberg>
hmmm seems i have a regression in the lecroy driver
<miek>
can scopehal decode a parallel bus & plot a trace with the analog value?
<azonenberg>
aaand yes there is a regression between scopehal:5f98ea3 and scopehal:ca74b58
<azonenberg>
miek: Not yet, but planned; scopehal:#39 is the ticket for that
<miek>
aha, thanks
<azonenberg>
there's also one for the opposite
<azonenberg>
given an analog value, get a digital vector
<azonenberg>
not sure how useful that will be but it seems reasonable to allow going both ways
<_whitenotifier-3>
[scopehal] azonenberg pushed 1 commit to master [+0/-0/±1] https://git.io/JvK6R
<_whitenotifier-3>
[scopehal] azonenberg 7988562 - Fixed regression from ca74b58
<azonenberg>
miek: basically the plan is to specify signed/unsigned, gain, and offset coefficients
<azonenberg>
then it will convert to volts
<azonenberg>
(scaling to other units will come later)
<azonenberg>
Do you have an immediate need for this or are you just wondering?
<miek>
it would be handy for something i'm working on, but no strong need (and i'd need to get logic channel support implemented for my scope first anyway)
<azonenberg>
Ok
<azonenberg>
And before you put too much time into that i want to do the trigger revamp
* zigggggy
coughs
<monochroma>
D:
* azonenberg
quarantines zigggggy before the whole channel gets sick
<zigggggy>
:(
<zigggggy>
azonenberg why is this so contagious? it seems way more contagious than say a cold or flu
<azonenberg>
zigggggy: No clue. You should ask someone who studies human viruses, not computer viruses
<azonenberg>
:p
<monochroma>
XD
<zigggggy>
i thought you read up on it
<azonenberg>
I have a basic level of familiarity, but not down to the details of exactly how it spreads
<azonenberg>
The "common cold" is actually caused by a multitude of different viruses: some coronaviruses and some rhinoviruses, possibly others
<azonenberg>
so i would naively expect the mechanism of spread to be about the same as coronavirus-based colds
<awygle>
adenovirus
<zigggggy>
the mechanism is the same
<zigggggy>
but it seems the virulity is much higher
<azonenberg>
yeah i have no idea why
<zigggggy>
virility*
<zigggggy>
as in, anyone who comes in close contact with an infected person is very likely to get sick
<azonenberg>
Yeah
<zigggggy>
ordinarily you'd stay away from sick people or sick people would be at home resting but people who are sick are asymptomatic for days!
<zigggggy>
nature is cruel azonenberg
<azonenberg>
yes i think thats actually the biggest thing
<azonenberg>
the asymptomatic spread
<azonenberg>
most colds arent contagiuous until you feel sick
<awygle>
and the diversity of severity of symptoms
<monochroma>
idk, i would expect most people who come into that same kind of contact with a flu/cold to be at similar levels of risk. but this is just getting reported on heavily so it seems worse?
<awygle>
between "zero" and "death"
<azonenberg>
awygle: yeah lol
<zigggggy>
monochroma so maybe its just that people are asymptomatic
<azonenberg>
and yes the flu spreads quite aggressively too
<zigggggy>
so more people are getting it?
<awygle>
covid-19 seems to have higher mortality rates
<monochroma>
zigggggy: yeah i think that is likely a big factor, the delay
* zigggggy
sighs
<awygle>
i've heard people say stuff like "if you don't worry about the flu don't worry about covid19" which is imo a poor argument
<zigggggy>
my father's bday is this weekend i told my mother to cancel any celebration
<zigggggy>
i dont think anyone should be out
<zigggggy>
at a restaurant
<zigggggy>
he's turning 70
<awygle>
my darling father is flying to vegas in like two days -_-
<zigggggy>
oof
<zigggggy>
its already in vegas
<zigggggy>
and its already here in NJ and increasing
<zigggggy>
exponentially
<awygle>
is it really in vegas? hadn't heard that
<awygle>
if you have a source i'll send it to dad
<zigggggy>
that's what i read the other day
<azonenberg>
i'm not sure where ELSE it is, i know it's *here* which is enough for me to care :p
<zigggggy>
its also highly suspected that its many places in much larger numbers than reported... most people are not going to get tested unless they have more severe symptoms
<zigggggy>
i even read several articles saying today even if you think you have it the recommendation is to stay home and rest unless its bad enough you need medical attention
<zigggggy>
and that testing is not necessary
<azonenberg>
i mean, a significant majority of people do recover fine on their own. It just leads to underreporting of the true level of infection if you're not tested
<zigggggy>
azonenberg but its being reported as only scary if you are of older age but in bergen county there is a 31 year old from fort lee who has it bad and is hospitalized but not in critical condition or anything
<zigggggy>
but bad enough that he needed extra care
<zigggggy>
says he has no medical issues or anything
<azonenberg>
Yeah, mostly only the older people are dying but folks of any age can get it bad enough to end up in the hospital
<azonenberg>
Which may not be a problem NOW, but if the hospitals don't have enough beds for everyone...
<monochroma>
azonenberg: why blur out your hentai collection directories?
<azonenberg>
monochroma: lol. it's mostly names of client projects
<azonenberg>
monochroma: any other options you think are necessary? And should they all be checked by default?
<awygle>
"mostly"
<awygle>
yeah it looks fine except being gtk
<zigggggy>
azonenberg i think a lot of major hospitals have provisions for setting up make shift beds as part of emergency planning
<awygle>
trololol
<azonenberg>
also is it clear that "reconnect to instrument" means "blow away any existing channel configuration etc and replace that with the settings in the save file"?
<zigggggy>
i did read a little bit about it
<monochroma>
azonenberg: hmmm i wouldn't not have thought about that until you mentioned that :P
<zigggggy>
azonenberg im next to hackensack hospital which is level 1 trauma center it hink
<azonenberg>
Would you change the wording?
<zigggggy>
azonenberg you familiar with it?
<azonenberg>
zigggggy: i know hackensack but am not familiar with the hospital in particular
<azonenberg>
i've heard camden is the place to go if you get shot in NJ
<zigggggy>
heh
<azonenberg>
because the docs have more practice than anywhere else in the area :p
<zigggggy>
hackensack is very qualified to diagnose rare diseases and stuff
<azonenberg>
I dont think this qualifies as 'rare' anymore
<monochroma>
azonenberg: maybeeee "reconnect to instrument with saved configuration" ?
<zigggggy>
no
<zigggggy>
i was just saying what i know of hackensack
<zigggggy>
i dont know if that's part of being a level 1 trauma center or not
<zigggggy>
but you need to be able to treat almost anything to get that certification
<azonenberg>
trauma normally implies physical injuries, not diseases
<azonenberg>
although generally if your trauma unit is that good, your other groups will be too
<zigggggy>
well i think the level 1 designation means you also have all kinds of surgeons and drs on standby for emergencies
<azonenberg>
monochroma: "Reconnect to Instrument (using saved settings)"
<azonenberg>
?
<zigggggy>
doenst have to be physical injury
<monochroma>
azonenberg: hmmmmmm maybe (and configure it using saved settings)
<azonenberg>
(reconfigure using saved settings) i think is more clear
<monochroma>
yeah
<monochroma>
i like that
<monochroma>
brbrbrbrb
<zigggggy>
oh nvm azonenberg hackensack is level 2
<zigggggy>
you're right azonenberg.. the one in camden is level 1
<azonenberg>
lain: btw, crazy idea: not that i am going to need it right now
<azonenberg>
But for coplanar waveguide plating modeling
<azonenberg>
i wonder how close you could get by ONLY modeling edge plating?
<azonenberg>
so much of the field travels on the sides of the traces, does the top plating even matter?
<azonenberg>
with microstrip the top surface carries a lot too, but in CPW i think it's almost all on the edges
<zigggggy>
nj only has 3 level 1 only
<azonenberg>
this would avoid needing an extra metal layer, which for those of us who can't afford sonnet pro is a big improvement
<azonenberg>
That's a navy SAR helicopter out of Whidbey Island at the bremerton airport with six of my SAR team members carrying a simulated patient onto it as part of a joint training exercise
<zigggggy>
ah
<zigggggy>
you're emt?
<zigggggy>
national guard?
<azonenberg>
WFR, which is about the first half to 2/3 of the EMT curriculum plus a bunch of stuff specific to being out in the woods far from help
<azonenberg>
We're a private nonprofit that's deployed on request by the county sheriff's office
<zigggggy>
pfft and you tell me you dont know how viruses work
<azonenberg>
but we work closely with other law enforcement and military units, including the navy sar pilots
<azonenberg>
incidentally the civilian medevac helo for the region is based at the same airport we were training at
<azonenberg>
navy is quite a bit up north but flew in for the training
<zigggggy>
ah
<azonenberg>
well ok navy is all over the place :p
<azonenberg>
but the navy airport that unit is based out of is a ways north
<zigggggy>
thats a big helo
<zigggggy>
prob carries a lot of equip
<azonenberg>
carrier based F-18s aren't going to be much help for SAR, those are a lot closer :p
<zigggggy>
and i guess you need to carry a couple emts and the patient and the pilot
<azonenberg>
it's a Jayhawk which is the navy's variant of the Blackhawk. They have a single navy corpsman who's paramedic qualified plus the pilot, copilot, and a crew chief in the back to run the winch and stuff
<azonenberg>
i'm sure the crew chief doubles as a backup medical guy when needed too
<zigggggy>
it needs a copilot?
<azonenberg>
Pretty much all sar aircraft have one, even though they can fly without one. Second set of eyes is super useful
<azonenberg>
especially when you're trying to land in really tight spots etc
<zigggggy>
ah
<sorear>
N+1 redundancy
<azonenberg>
being able to see out the other side and see if your rotor is about to hit something is nice
<azonenberg>
or even switch to flying on the other side if you have a better view
<zigggggy>
azonenberg my car knows everything around it.. i would think helos would have that by now too :P
<azonenberg>
Or take turns if one guy is getting tired. Long hovers are exhausting, i'm sure
<zigggggy>
azonenberg well this is a med evac.. i imagine its a quit sprint to the hospital
<zigggggy>
quick
<azonenberg>
These guys' primary function is rescue support for all of the navy bases in the area flying training flights
<sorear>
how good are modern helos at landing without killing people on the ground if they have 0 pilots over a populated area
<azonenberg>
if one of the crews goes down in the ocean, or the mountains, these guys are the first ones that get called
<azonenberg>
So they almost always have one or two birds on standby supporting that
<azonenberg>
If they have anyone not busy on that, they're available to support civilian rescue operations as needed
<zigggggy>
ah
<azonenberg>
whiel i guess they could do normal medevacs from urban areas in a pinch, it's not their specialty
<azonenberg>
they go places nobody else will
<azonenberg>
They flew one four of my teammates onto the side of a mountain on a slope too steep to land on
<azonenberg>
so they did a "one wheel landing" where they hover with one wheel touching the side
<azonenberg>
you get out, duck down, and it takes off over you
<zigggggy>
sounds dangerous
<azonenberg>
It is, it's not something you do if there's a good alternative
<zigggggy>
helicopters accidents are abundant
<zigggggy>
its a bit scary
<azonenberg>
But if we have to do it, i'd rather have these guys be flying us than any other pilots in the country
<azonenberg>
They're as good as they get
<azonenberg>
And when lives are on the line some risks are usually unavoidable
<azonenberg>
You take reasonable steps to mitigate them, of course
<zigggggy>
well there are usually policies in place not to perform certain rescues
<zigggggy>
there's no point if you're going to end up killing yourself and others
<azonenberg>
Yeah. But vertical hoists, one wheel landings, low hovers, etc are just another day at the office for these guys
<zigggggy>
ah
<azonenberg>
Most civilian medevac crews will NOPE out of that sort of stuff
<azonenberg>
But when we have a patient we can't get out any other way, we call them
<azonenberg>
Or, in some cases, the coast guard. Although if i had to pick i'd call navy
<azonenberg>
nothing against the coast guard pilots but the navy helos are bigger and better equipped for shore-side rescues especially in the mountains
<zigggggy>
i imagine in bad weather conditions you cant do some rescues either?
<azonenberg>
Yes, we routinely deploy in situations where aircraft are grounded
<zigggggy>
oh
<azonenberg>
either too windy or they just can't see
<azonenberg>
So we're on our own, and the only way to get someone out is to carry them on foot
<zigggggy>
there are no situations that ground even your crew?
<zigggggy>
as in, it just cant be done?
<azonenberg>
I mean, vertical technical stuff we'd call mountain rescue
<azonenberg>
we don't do straight up cliffs etc
<zigggggy>
ah
<azonenberg>
And there have been times when we've decided searching an area was ill advised because either the person was very unlikely to have gone there, or because it was at the bottom of a steep cliff that they couldn't have survived going over etc
<zigggggy>
yeah i imagine there are simply physical limits to a helo rescue
<azonenberg>
but in my four? years with the unit i can't recall ever being told the weather was too bad for us to go out
<zigggggy>
interesting
<azonenberg>
and we certainly have never been in a situation where someone was alive, we knew where they were, and couldn't get to them
<sorear>
it helps that puget sound rarely gets high-tier hurricanes
<zigggggy>
i could see a hurricane woud maybe prevent rescues in residential areas
<zigggggy>
wind + debris
<zigggggy>
and soaking rain
<azonenberg>
If we can't get a helo in, we'll either carry them out, or go into long term care mode and stabilize the patient on scene until an aircraft is able to reach us
<azonenberg>
Yeah we don't really get hurricanes. We did respond to the aftermath of an EF-2 tornado last year though
<azonenberg>
going door to door checking for trapped people, carrying food and baby supplies in on foot to people who were unable to reach the store, etc
<sorear>
not sure if zigggggy would consider a firestorm "bad weather conditions"
<azonenberg>
(roads were completely impassable and there was no power to half the city)
<azonenberg>
that was an extreme rarity, the last natural disaster we got called to was a mudslide which is far more typical for this region
<zigggggy>
i was considering bad weather condition as something that would greatly lead to a crash
<azonenberg>
Yes. It's not uncommon that the weather is too bad to get a medevac
<azonenberg>
But we normally hike into searches anyway, and we train for ground extractions all the time
<zigggggy>
but i imagine that helo can perform much better than civilian helos
<zigggggy>
so fewer events will stop it
<azonenberg>
We've been getting air support more frequently lately, especially from the sheriffs dept over in king county who has a very nice surveillance helo with FLIR etc
<azonenberg>
(but no winch or rescue capabilities, if they see somebody we still have to go in on foot to get them out)
<azonenberg>
its a combination of better gear, thermal imaging, the crew all have night vision goggles, etc
<azonenberg>
and better training
<azonenberg>
So yes they will fly in places many civilian pilots wont
<azonenberg>
but even they can't thread the needle hoisting someone from between 200 foot pine trees in heavy gusty winds
<zigggggy>
yeah
<azonenberg>
I did not make that up btw, we did actually have someone medevaced from the bottom of some very tall trees
<zigggggy>
on a similar note, i hear airline tickets are very cheap now azonenberg
<azonenberg>
The coast guard helo in question had a 250 foot cable on their winch and used almost all of it
<zigggggy>
if you can find a place to go
<azonenberg>
Lol, no thanks :p
<zigggggy>
:(
<zigggggy>
is there no island we can isolate ourselves for a month?
<sorear>
given the exponential spread of the disease in 45 days everyone not dead will be immune
<zigggggy>
:(
<zigggggy>
give the darwin award to this virus
<zigggggy>
started potentially from a bat in china and looks to be able to infect most of the glove
<zigggggy>
globe
<zigggggy>
that's survival
<_whitenotifier-3>
[scopehal] azonenberg pushed 1 commit to master [+0/-0/±2] https://git.io/JvKiM
<_whitenotifier-3>
[scopehal] azonenberg effea8a - Formatting fixes to save files
<_whitenotifier-3>
[scopehal-apps] azonenberg pushed 1 commit to master [+0/-0/±4] https://git.io/JvKiy
<azonenberg>
Hmm, so before i can add the reconnect functionality - rather critical - i need to unify the connection string parsing so an Oscilloscope can be queried for the current connection string, then be created from that string again later on
<azonenberg>
Yay for dependency graphs that don't start at the thing you're trying to do :p
<azonenberg>
So I guess the *actual* next step is to make a VICP SCPITransport and make the lecroy driver use it, so it fits into the same object model as everything else
<awygle>
are you still canonically building all this from splash?
<monochroma>
splash isn't used at all afaik
<monochroma>
cmake and make
<awygle>
oh good
<awygle>
(no 'fense)
<azonenberg>
no i think i've removed all of the splash stuff from this
<awygle>
Scopehal feature request - PSU support :-P
m4ssi has joined #scopehal
_whitelogger has joined #scopehal
<azonenberg>
awygle: actually
<azonenberg>
the APIs in the library are already there, at least for basic voltage/current control (not all of the details like soft start necessarily)
<azonenberg>
And i have a driver for my R&S PSUs
<azonenberg>
And there is a "psuclient" binary in scopehal-apps that can connect to the PSU and display realtime graphs of current and voltage as well as turn channels on and off
<azonenberg>
there are text boxes for entering new current/voltage but i haven't added or implemented the "apply changes" button
<azonenberg>
so for the most part it's just a data viewer right now. But all of the APIs for changing settings are present and it would be easy to add the UI to change config, i've just been putting it off to work on other stuff
<azonenberg>
I also have preliminary APIs for controlling function generators, as well as partial driver support for the integrated waveform generator in the lecroy wavesurfer 3000 series
<azonenberg>
but there's not currently any UI around that
<azonenberg>
as i do not currently own a function generator, my ability to work on that is limited
<_whitenotifier-3>
[scopehal] azonenberg pushed 1 commit to master [+2/-0/±5] https://git.io/JvKd2
<_whitenotifier-3>
[scopehal] azonenberg pushed 1 commit to master [+0/-0/±6] https://git.io/JvKdD
<_whitenotifier-3>
[scopehal] azonenberg 66ed64f - Removed LeCroyOscilloscope::ReadMultiBlockString() as it's not used
<azonenberg>
wow the lecroy driver was in need of a lot of cleanup lol. Dead code that wasnt even being called
<azonenberg>
This is the first time in years i've got deep in the VICP protocol portion of that stack
<azonenberg>
The new VICP code should be cleaner and more efficient too
<azonenberg>
also VICP is literally GPIB tunneled over TCP lol - it has an EOI bit in the header
<_whitenotifier-3>
[scopehal] azonenberg pushed 1 commit to master [+0/-0/±6] https://git.io/JvKFW
<_whitenotifier-3>
[scopehal] azonenberg 74a0408 - Moved a lot more VICP processing into VICPSocketTransport
<azonenberg>
Well that's a lot cleaner than I had before
<_whitenotifier-3>
[scopehal] azonenberg pushed 1 commit to master [+0/-0/±3] https://git.io/JvKFX
<_whitenotifier-3>
[scopehal] azonenberg 669432c - Continued cleanup of VICP driver
<_whitenotifier-3>
[scopehal] azonenberg pushed 1 commit to master [+0/-2/±5] https://git.io/JvKbF
<_whitenotifier-3>
[scopehal] azonenberg abb9978 - Siglent and LeCroy drivers now use SCPITransport. Deleted LeCroyVICPOscilloscope class. Fixes #71. Fixes #68.
<_whitenotifier-3>
[scopehal] azonenberg closed issue #68: Refactor LeCroy scope driver to use SCPITransport - https://git.io/fjElE
<_whitenotifier-3>
[scopehal] azonenberg closed issue #71: Refactor Siglent driver to use SCPITransport - https://git.io/fjE8i
<azonenberg>
Error_404: please check i didn't break anything in the Siglent driver with this
<azonenberg>
i deleted your analog channel checking code because the new LeCroyOscilloscope DetectAnalogChannels() is a lot more generic and handles suffixes like WAVERUNNER8104-MS correctly
<azonenberg>
so i think it should parse siglent model names too
<_whitenotifier-3>
[scopehal-apps] azonenberg pushed 1 commit to master [+0/-0/±1] https://git.io/JvKNv
<_whitenotifier-3>
[scopehal-apps] azonenberg b4afb2a - glscopeclient: now create LeCroy / Siglent drivers via SCPITransport
<_whitenotifier-3>
[scopehal-docs] azonenberg pushed 1 commit to master [+0/-0/±2] https://git.io/JvKNf
<_whitenotifier-3>
[scopehal-docs] azonenberg 61a325a - Added some notes on third party libraries
<_whitenotifier-3>
[scopehal-cmake] azonenberg pushed 1 commit to master [+0/-0/±3] https://git.io/JvKNU
<_whitenotifier-3>
[scopehal-cmake] azonenberg 51d9321 - Updated submodules with VICP refactoring
<azonenberg>
Next up is #99, proper dynamic creation. That will take a bit of work but should be doable
<_whitenotifier-3>
[scopehal-apps] azonenberg pushed 1 commit to master [+0/-0/±1] https://git.io/JvKNQ
<_whitenotifier-3>
[scopehal-apps] azonenberg ec09bc1 - Refactoring of scope creation to use separate transport/driver stuff (see #99)
<azonenberg>
Also i'm making a non-backward-compatible change to connection string format
<azonenberg>
instead of myscope:lecroy_vicp:127.0.0.1 it's myscope:lecroy:vicp:127.0.0.1
<azonenberg>
i.e. driver and transport separated by colon
<_whitenotifier-3>
[scopehal-apps] azonenberg pushed 1 commit to master [+0/-0/±1] https://git.io/JvKNj
<_whitenotifier-3>
[scopehal-apps] azonenberg 42f16d3 - glscopeclient: driver and transport are now separated by colons
<_whitenotifier-3>
[scopehal-docs] azonenberg pushed 1 commit to master [+1/-0/±3] https://git.io/JvKAy
<_whitenotifier-3>
[scopehal-docs] azonenberg 9faea3c - Updated documentation with new driver model
<_whitenotifier-3>
[scopehal] azonenberg pushed 2 commits to master [+0/-2/±22] https://git.io/JvKxE
<_whitenotifier-3>
[scopehal] azonenberg 232aa77 - Added GetDriverName() and GetTransportConnectionString()
<_whitenotifier-3>
[scopehal] azonenberg 2cd392c - Removed vestigial RED TIN driver
<azonenberg>
ok, so at this point save files have the connection string in them
<azonenberg>
now i just have to do the dynamic creation logic to instantiate drivers and transports from a connection string rather than hard coding the table in glscopeclient/main.cpp
<azonenberg>
Yay for popping the yak stack
<azonenberg>
This was long overdue
<azonenberg>
Also, once this dynamic creation table is set up, we'll be able to register new scpi transports and scope drivers at run time too
<azonenberg>
Which will enable plugins as soon as we create logic to load so's / dlls and call the register-stuff function
balrog has joined #scopehal
<_whitenotifier-3>
[scopehal] azonenberg pushed 1 commit to master [+0/-0/±9] https://git.io/JvKjd
<_whitenotifier-3>
[scopehal] azonenberg 60c4eab - Added dynamic creation for SCPITransport's. See #99.
<_whitenotifier-3>
[scopehal-apps] azonenberg pushed 1 commit to master [+0/-0/±1] https://git.io/JvKjb
<_whitenotifier-3>
[scopehal-apps] azonenberg c53ef0c - glscopeclient: now create SCPITransport's dynamically
<_whitenotifier-3>
[scopehal] azonenberg pushed 1 commit to master [+0/-0/±9] https://git.io/Jv6ee
<_whitenotifier-3>
[scopehal] azonenberg d09c151 - Split GetTransportName() and GetTransportConnectionString() out to be separate
m4ssi has quit [Remote host closed the connection]
_whitelogger has joined #scopehal
bvernoux has joined #scopehal
<bvernoux>
hi
<azonenberg>
o/ bvernoux
<bvernoux>
I have just received HydraNFC v2 Antenna ;)
<bvernoux>
Quality is nice
<bvernoux>
from PCBWay
<bvernoux>
it is a shame but PCBWay is a lot faster than MacroFab ...
<bvernoux>
and MacroFab tell me that there is delay because of CoronaVirus ...
<bvernoux>
I love there online site to do everything but it is too slow and a bit too expensive and in addition PCB are not done in USA ... but Taiwan ...
<bvernoux>
So I was expecting doing HydraNFC v2 100% in USA will be very nice & fast even if it is more than 2 or 3 time more expensive than in China
<bvernoux>
It is very hard to find PCB+Assembly as competitive as China
<bvernoux>
I do not speak in EU as there is nothing ...
<bvernoux>
Anyway I will probably use again MacroFab only for expensive product and sold in USA
<bvernoux>
A good surprise is using DirtyPCBs Laser cut service
<bvernoux>
the result is very nice and price is correct (I have not found anything cheaper which is fast enough and 100% automated online)
<azonenberg>
Still waiting for quote from my chinese fab on the probe test board. Maybe virus delays on their end
<bvernoux>
On my side all things done in china are come back to normal
<bvernoux>
DHL sent from China to France take 3 days
<bvernoux>
I'm waiting prod of 200HydraBus
<bvernoux>
using using shenzhen2u
<bvernoux>
So far they are very good to answer quickly with good price
<bvernoux>
it seems the quality is also very good I will confirm that when I will receive the boards
<bvernoux>
I need to test 200boards ;)
* zigggggy
coughs
<bvernoux>
Thanks to semi automated test
<azonenberg>
Yeah i am going to need to set up some kind of semi automated characterization setup for my probes when the time comes
<bvernoux>
The best is the delay for everything 16 open days
<bvernoux>
I have not found any other PCB+Components+Assembly fab as fast as that
<azonenberg>
hook up the SMA, stick the probe into the socket, hit a button, collect s-parameters, compared to known good
<bvernoux>
Do you plan to provide test plan to do that in China ?
<bvernoux>
In paste it was quite hard to let do them tests ;)
<bvernoux>
I have full Test Plan ultra detailed
<bvernoux>
but they often do that in hurry and they do not provide the logs
<azonenberg>
at least for small volume my plan is to do it in house. Order bare boards, have my technician do the assembly and testing on site in my lab,
<bvernoux>
I still need to find serious guys to do that correctly
<azonenberg>
The probe is three resistors, two tip contacts, and a SMA. Not unreasonable for hand assembly
<bvernoux>
Issue to do that with VNA is you need to change the connector each 500 connections ;)
<bvernoux>
it shall be taken into account
<bvernoux>
on my VNA each connector is > 150USD ;)
<bvernoux>
the same for cable
<azonenberg>
If i'm doing this kind of thing at that kind of volume, i'd have a short sacrificial adapter on the front of the VNA
<azonenberg>
m-f coupler or similar
<bvernoux>
yes
<azonenberg>
that way 250K mating cycles instead of 500
<bvernoux>
Type N to SMA are cheaper on normal VNA/SA but on mine it is 7mm to SMA ;)
<bvernoux>
it is why I will qualify the NanoVNA 2 to check the accuracy to use it for simple tests ;)
<bvernoux>
I will like to build my own NanoVNA 2 with some improvment ;)
<bvernoux>
first change the SMA connector through hole is awful for signals vs SMT SMA
<azonenberg>
wait it has a through hole sma??
<bvernoux>
yes ;)
<azonenberg>
those exist?
<bvernoux>
Nano VNA 2 only
<bvernoux>
it is available
<azonenberg>
i've seen through hole mechanical pins and then smt center conductor
<bvernoux>
yes edge launch sma is a must have even for something < 1GHz ;)
<bvernoux>
VSWR is also heavily impacted
<bvernoux>
by through hole ...
<awygle>
idk, ime there's two kinds of people. "SI? What's that? Who needs a ground plane?", and "Must have 30$ edge launch SMAs with full 3D EM simulations for my UART".
<bvernoux>
also microUSB could be replaced by USBC more robust and better for everything also for noise ;)
<azonenberg>
awygle: lol
<awygle>
usbc requires better PCB fab than micro
* azonenberg
still has some mini usb connectors in inventory...
<bvernoux>
awygle, anyway for anything I'm using 4 layers ENIG ;)
<bvernoux>
min ;)
<awygle>
sure, barely any reason to try to use 2 these days
<awygle>
unless you're _very_ cost constrained
<bvernoux>
not for antenna or very basic stuff of course ;)
<awygle>
the last two-layer board i did was an LNA
<bvernoux>
but for NanoVNA 2 a good 4 Layers PCB + ENIG is mandatory
<bvernoux>
especially it go up to 3GHz
<bvernoux>
ENIG or Immersion Silver ;)
<awygle>
did azonenberg say immersion silver was not more expensive than enig the other day? find that hard to believe
<bvernoux>
What seems interesting is they have tons of RF/Special PCB
<bvernoux>
awygle, here they have closed some company for 1 or 2 case of COVID-19
<awygle>
Anyway. I outsource all assembly these days. Totally worth it.
<bvernoux>
Today all countries have so huge dependencies with China ...
<bvernoux>
It is quite simple they do everything and they have all the raw materials ...
<Error_404>
azonenberg: re: detection regression - will do today...
<Error_404>
hmm, looks fine. I'll test it on hw in a bit. Only remark is that by using the scpi transport I'm going to need to rethink how my workaround for their non-responses are
<Error_404>
works*
<Error_404>
I cooullld just monkey-patch it in the constructor :P
<Error_404>
actually I guess the sane alternative is just to add the timeout code to the transport and have it inactive by default. I'll do that.
<azonenberg>
Yeah that seems to make a lot more sense. You can then use RTTI to check if you have a SCPISocketTransport and enable it if needed
<azonenberg>
This new architecture is a lot cleaner wrt supporting usbtmc, rs232, gpib, etc in the future though
<azonenberg>
or even pcie :p
<azonenberg>
Error_404: note that driver and transport are now separated by a colon not an underscore
<azonenberg>
so myscope:lecroy:lan:1.2.3.4
<azonenberg>
anyway there is going to be one more API change to all of the Oscilloscope classes in the near future to enable dynamic creation of scopes given the driver name. Will throw that together tonight after work
<azonenberg>
at that point i think all of the yaks i need to shave before implementing file-load will have been dealt with
<azonenberg>
(note that this is still ui config only + reconnecting to instruments pending, there's no serialization of waveforms yet - that should be easy once all of the metadata is taken care of though)
<azonenberg>
oh also, as of last night if you don't have yaml-cpp installed, you'll need it to build glscopeclient
<azonenberg>
(not currently a dependency of scopehal, but i expect it will be as soon as i implement file loading)
<azonenberg>
libyaml-cpp-dev on debian/ubuntu
<Error_404>
community/yaml-cpp on arch
<azonenberg>
In the very long term i want to start thinking about distro packaging etc
<azonenberg>
but we're a long way from being mature enough for that to be even a remote consideration right now
<azonenberg>
For the short term i'll just document dependencies in the manual
futarisIRCcloud has quit [Read error: Connection reset by peer]
wbraun has quit [Remote host closed the connection]
LeoBodnar has quit [Remote host closed the connection]
LeoBodnar has joined #scopehal
futarisIRCcloud has joined #scopehal
wbraun has joined #scopehal
sorear has quit [Remote host closed the connection]
sorear has joined #scopehal
elms has quit [Remote host closed the connection]
elms has joined #scopehal
<zigggggy>
azonenberg Due to Coronovirus (COVID19) all TCP Applications are being converted to UDP to avoid Handshakes.
<azonenberg>
Lol. Seen it already but still funny
bvernoux has quit [Quit: Leaving]
m4ssi has joined #scopehal
m4ssi has quit [Remote host closed the connection]