azonenberg changed the topic of #scopehal to: libscopehal, libscopeprotocols, and glscopeclient development and testing | https://github.com/azonenberg/scopehal-apps | Logs: https://libera.irclog.whitequark.org/scopehal
Degi_ has joined #scopehal
Degi has quit [Ping timeout: 240 seconds]
Degi_ is now known as Degi
<azonenberg> Took a bit of time off from the filter graph refactoring to work on the Digilent WaveForms bridge which some folks were bugging me about
<azonenberg> I just successfully dumped a sinewave to CSV from the bridge
<azonenberg> next step is to spin up a bit more of a scpi stack and write the scopehal side client code
Bird|otherbox has quit [Remote host closed the connection]
Bird|otherbox has joined #scopehal
bvernoux has joined #scopehal
<GyrosGeier> oof
<GyrosGeier> I need to get librevisa working
<mxshift> azonenberg: work has some used 8GHz solder-in probes that you're welcome to if you want to look at their construction
<mxshift> Tektronix P77STFLXA for TDP7000 probes
<azonenberg> mxshift: a few photos, both overview and (if you can get them) microscope would be cool
<azonenberg> no need to send them physically
<mxshift> What areas do you want under microscope?
<azonenberg> The tip where it mates with the DUT
<azonenberg> ah yes these are the ones that have a little clip style mounting so you dont need a coax connector on the tip
<mxshift> The blocky piece near the tip appears to be ferrite. I've got a few so I can try to remove to get details underneath
<azonenberg> i suspect its an EM shield over an active amplifier
<azonenberg> i'm also interested in whatever is between that block and the actual solder points
<azonenberg> if there's any SMT or thru hole damping resistors
<azonenberg> geometry of the solder points
<azonenberg> etc
bvernoux has quit [Quit: Leaving]
<mxshift> had to figure out how to use the amscope camera
<azonenberg> Trying to figure out what i'm looking at at the end
<azonenberg> was this cut off the DUT?
<azonenberg> or are these complete length leads?
<mxshift> this was used a few times and then cut
<azonenberg> Ok so there's normally stuff coming off the end
<azonenberg> are they just wires? resistors like the lecroy ones?
<mxshift> it comes with just through holes
<azonenberg> how long?
<azonenberg> ah ok
<azonenberg> so you have to run wire to it
<mxshift> yes
<azonenberg> innnteresting
<mxshift> folks here say please don't do this
<azonenberg> yeah i would expect the length to be tightly controlled because every little bit between the DUT and the probe input is going to contribute to reflections
<azonenberg> longer wire = more phase shift
<mxshift> At 3:42 in this training video for these probes, they go to trim the lead and knock the resistor off with the cutters
<azonenberg> wow that is not a very good video
<azonenberg> its super blurry half the time
<azonenberg> also lol i see the very carefully timed cut so you don't see the resistor fly off
<azonenberg> it looks like they delaminate a bunch of stuff with that cut
<azonenberg> at 4:15 you can see ceramic fragments and what looks like some glue and delaminated coverlay
<azonenberg> also the soldering iron tip he's using is humongous. looks like soldering with a tree trunk lol
<azonenberg> or a telephone pole
<azonenberg> also using grossly excessive volumes of solder
<Degi> That is an official video... wow
<azonenberg> lol yes
<azonenberg> it looks like something you'd find on the eevblog forums
<Degi> A bit before they mentioned to not overheat it because that might cause the resistor solder to melt and move around (also the part where they cut off a part of the resistor looks very jittery and hasty)
<azonenberg> lol yes
<azonenberg> they did glue down the resistors though
<azonenberg> so melting shouldnt be a huge deal
<mxshift> There are many reasons we have a growing pile of these
<azonenberg> lol
<azonenberg> Any interest in an engineering sample AKL-PT5 soonish?
<Degi> Also the solder in wires at the tip seems weird, especially in terms of consistency... I guess you could put wires there and never desolder them and just calibrate it with that...
<azonenberg> yeah it does seem hard to get good repeatable results with
<azonenberg> mxshift: so you're not a fan of these tek probes?
<azonenberg> is it the ergonomics? performance? both?
<mxshift> I'm relaying my coworkers' experience: "these are horrible"
<mxshift> Tedious to use. Easy to damage
<azonenberg> Would love to hear specific complaints. especially about how other probes they like more beat them
<azonenberg> Most differential probe tips are going to be easily damaged if you're careless
<azonenberg> requiring you to custom cut wire is awkward
<azonenberg> The LeCroy solder-in tips, as well as the AKL-PT5, have pre-cut wire/resistors already attached that is intended to stay part of the probe and not be replaced unless damaged (although it can be if damaged)
<azonenberg> and you desolder it from the DUT after use
<azonenberg> Looks like underfilled WLCSP amplifier of some sort
<mxshift> Gap filler up to the shield/heatsink/ferrite thing
<mxshift> so at least partially for thermal
<azonenberg> Yeah i imagine. so it looks like preamp in the tip, then i guess the main amplifier in the probe body
<azonenberg> All of the LeCroy tips I've worked with to date have been entirely passive
<azonenberg> however the newer DH series - which i've seen marketing materials and docs for, but never actually used
<azonenberg> add an EEPROM to the tip to store per-tip response curves
<azonenberg> the D-series probes have calibration data in the amplifier, but not each tip
<azonenberg> i think it's combined tip/amp response
<azonenberg> because you have to tell the software what tip you're using (no autodetection)
<azonenberg> and when you cal, you have to send back the amplifier and all your tips
<azonenberg> the DH series presumably avoid this by storing amplifier and tip response separately
<azonenberg> and it just concatenates the two to flatten it
balrog has quit [Quit: Bye]
<Degi> The solder mask and components look a bit uneven
<azonenberg> looking at amscope pics after having had the Leica for a while really shows off the difference in optical quality lol
<azonenberg> When you've only used cheaper optics you don't even really see what you're missing
<Degi> Is amscope the one used in the Tektronix video?
<azonenberg> no that's an olympus. but an older one and a not-great lighting setup
<azonenberg> mxshift said he was using an amscope though
<Degi> I see
<Degi> It would be nice to have an image with polarization filters
<azonenberg> (you can see the olympus markings earlieri n the video i think)
<azonenberg> and yes that's exactly what i was thinking
<azonenberg> so much glare
<azonenberg> you get used to it, until you start seeing images without it :p
<Degi> I wonder if modern raytracing CGI tools consider polarization when multiple specular reflections are involved
<azonenberg> i'm sure it could be done
<azonenberg> but i dont think it's commonplace
<azonenberg> Also, AKL-PT5 v0.4 PCBs shipped from oshpark, should be here by thursday
<azonenberg> this is the same as the current v0.3 except with a solder pad for a mounting foot
<Degi> I wonder if the temperature mentioned at 0:34 is what they actually used or if they set it lower
<Degi> Nice
<azonenberg> so i expect no change in electrical properties
<tnt> AFAIK the vast majority is using the "Disney Principled Shader" and I don't recall anything about polarization in there.
<Degi> At 5:39 the PCBs also seem to be pretty thick, not just a flex board
<azonenberg> Would not surprise me if it was flex-rigid or stiffened flex
balrog has joined #scopehal
Stephie- has joined #scopehal
Stephie has quit [Ping timeout: 256 seconds]