<azonenberg>
13499Em may terminate with internal error if circuit includes CPW vias to ground that are within 1 cell of a box wall.
<azonenberg>
hmmmm i think i was responsible for that bug :p
Bird|ghosted has quit [Remote host closed the connection]
Bird|ghosted has joined #scopehal
someone-else has joined #scopehal
someone-else has quit [Quit: Connection closed]
someone-else has joined #scopehal
Degi has quit [Ping timeout: 240 seconds]
Degi has joined #scopehal
<azonenberg>
Guess what I just scored?
<azonenberg>
Teledyne LeCroy PP066 for $600-something. Missing a few accessories but includes the 10x tip resistor which is the part i care most about
<azonenberg>
very excited to get it on the VNA
<azonenberg>
(this is their 7.5 GHz transmission line probe)
<monochroma>
ooo!
<GenTooMan>
azonenberg, new tools are great, especially if you have a use for them. :D
<azonenberg>
GenTooMan: my main use case is as a test subject
<azonenberg>
i want to get some rising edge and VNA measurements on it compared to the AKL-PT1 / PT2
<azonenberg>
Also i think i'm gonna build an s-parameter viewer in glscopeclient soon
<azonenberg>
It's been a to-do item for quite a while
<azonenberg>
And there will be some changes to import semantics
<azonenberg>
in particular, importing won't close the existing session, it will add stuff to the session
<electronic_eel>
ah, so you'll be able to import for example a reference waveform, for example from simulation, so you can compare it to the measurements
<azonenberg>
Yes, that's the kind of thing i want to do
<electronic_eel>
or import some measurements you did a few weeks before with the older version of the prototype you are working on
<azonenberg>
as part of this work I also want to finally allow more than one analog waveform to occupy a single WaveformArea
<azonenberg>
And start thinking about how to implement things like memory waveforms and displaying old history waveforms
<azonenberg>
i want to be able to do things like visualizing old waveform data vs the current one
<azonenberg>
or even having filters that take inputs across time
<azonenberg>
like subtracting a historical waveform from the current one
<azonenberg>
or averaging multiple waveofrms
<azonenberg>
the data is there but right now it's kinda sequestered and can't actually be accessed
<electronic_eel>
especially with the simulation waveforms i have worked with before, you often need a way to scale them and move them around a bit, to account for offset or scaling mismatches
<azonenberg>
Yeah there is going to be some fine tuning like that to work on too
<azonenberg>
but one thing at a time :)
<electronic_eel>
but working with reference waveforms is a major pita with my rigol, really looking forward getting that in glscopeclient
<electronic_eel>
the rigol just stores them from the video buffer, every time you change the timebase, your whole reference data is bust
<azonenberg>
yeah lol
<azonenberg>
heck, even lecroy's handling of memory waveforms is pretty awkward
<electronic_eel>
hmm, if you make imported ref data available as regular data source that you can run filters and math upon, then the scaling thing i mentioned above shouldn't be too hard. it could be just a regular filter that does the scaling or offset correction
<azonenberg>
That's the idea
<azonenberg>
There's already filters to do that afaik
<azonenberg>
i know there's one for dc offset, i forget if there's one for multiply by constant
<azonenberg>
if there's not it would be trivial to implement
<electronic_eel>
you'd then also need some way to shift the ref waveform in time, to align it to the stuff you are measuring now
<electronic_eel>
that could be implemented as a delay filter, but i guess you want some gui to move it with the mouse to make it convenient
<azonenberg>
Yeah we'll worry about that kind of fine tuning later
<azonenberg>
Another thing i want to do is add support to the s-parameter loader for parsing >2 port touchstone files
<azonenberg>
and once we can load .s4p's i want to be able to convert 4-port s-parameters to mixed mode s-parameters
<azonenberg>
i.e. displaying mode conversion or differential insertion/return loss
<azonenberg>
Also add some gui shortcuts to go directly from a frequency domain s-param plot to a time domain transform
<azonenberg>
as well as some filters for things like calculating group delay
<electronic_eel>
what usecase do you have in mind for the touchstone file feature? just viewing the data in different modes or preparing it for deembedding and similar within scopehal?
<azonenberg>
mostly single-shop viewing of time and frequency domain data from simulations or vna measurements
<azonenberg>
rather than having to use sonnet for the frequency domain view and glscopeclient for time domain transforms or eye patterns
<electronic_eel>
some of that conversion stuff is usually part of a better vna gui
<azonenberg>
Sonnet has a very nice s-param viewer
<azonenberg>
but it lacks any time domain features whatsoever
<azonenberg>
and thats the thing, a lot of this data is from simulations and not vna data (and pico's vna gui sucks :P)
<electronic_eel>
what does your pico vna offer in that regard?
<azonenberg>
It has a time domain view but its terrible, and it only works with power of two harmonic sampling intervals
<azonenberg>
there's no resampling support
<azonenberg>
the plotting interface is just a pain, it's written in vb6 and only runs on windows
<electronic_eel>
hrmpf, not what you expect from a quality vna
<azonenberg>
No
<azonenberg>
I do not recommend the picovna at all
<azonenberg>
software has historically been pico's weak point
<azonenberg>
Which is why I have a good relationship with them :p
<electronic_eel>
i played with a agilent vna once, you could easily load in sim data, do transforms on it, compare it to the measured data and so on
<electronic_eel>
also i have a vnwa3. it is just up to 1.2 GHz and the software doesn't look shiny. but the software is quite capable
<azonenberg>
I'm considering two main options down the road
<azonenberg>
the copper mountain 20 GHz 4 port vna, in a... 3U? headless rackmount form factor
<azonenberg>
and the teledyne lecroy wavepulser 40ix, a 40 GHz 4 port TDR
<azonenberg>
in a headless USB form factor
<azonenberg>
both are about the same price, both are 4 port. the copper mountain is i think N and the wavepulser is 2.92
<azonenberg>
copper mountain software is cross platform iirc, wavepulser is windows only but the software looks *really* nice
<azonenberg>
the main choice is whether i want to get double the bandwidth but less dynamic range, or vice versa
<electronic_eel>
what kind of price range are those?
<azonenberg>
O($50K)
<azonenberg>
high 40s iirc?
<azonenberg>
that's MSRP for a new one
<azonenberg>
i would be targeting used/refurb and expect to pay mid to high 30s if i am able to find one refurb at the time I've got budget
<electronic_eel>
ouch
<electronic_eel>
should rename them from copper mountain to gold mountain
<azonenberg>
lol
<azonenberg>
for a 20 GHz 4 port vna that's not unreasonable at all
<azonenberg>
iirc it's pretty cheap compared to the likes of R&S
<electronic_eel>
yeah, but it means that this kind of freq range is out of reach for the vast majority of hobbyists, if you want somewhat decent gear and not old boat anchors
<electronic_eel>
same with scopes. you can very well build pcie gear or use ddr3 ram with a hobby budget. but you can't buy a scope to do any kind of si debugging on it
<azonenberg>
Yeah that's why i've been offering out my lab to do characterization for people
<azonenberg>
Because it's likely the only way they'll ever be able to get that kind of data
<azonenberg>
And the only reason *I* was able to afford the 16 GHz scope is because it's the 2009 model year lol
<electronic_eel>
another nice thing about the rigol rack scope i posted a few days back is the price: 3500 EUR for the entry model, with the likeliness of it being hackable to 2 GHz being very high
<electronic_eel>
that is my current plan to get at least to 2 GHz on a hobby budget
<azonenberg>
Nice
<azonenberg>
If we had a solid performer on the low cost end of things that would definitely help
<azonenberg>
hopefully its not 2 GHz at 0.5 WFM/s :p
* azonenberg
glares at MSO5000 seires
<electronic_eel>
they need the SFP+ slot on the rack scope to increase that to 0.6 WFM/s ;)
<azonenberg>
lol
<electronic_eel>
but i heard rigol are planning to release a new firmware for the MSO5000 to increase scpi data extraction speed
<azonenberg>
Yeah thats been rumored for quite some time
<azonenberg>
looking forward to results
<electronic_eel>
the ds8000-r doesn't seem to be completely finished yet. they claim in the datasheet that there is an "open source" sdk to extract the data. but i couldn't find that yet
<electronic_eel>
also i want them to release a firmware update before i get one, so that i can first have a look at the binaries to check how hard it'll be to hack it
<electronic_eel>
from what i've seen so far, the whole current generation of rigol scopes uses nearly the same architecture. so adapting the existing hack for the 5000 should be straightforward