azonenberg changed the topic of #scopehal to: ngscopeclient, libscopehal, and libscopeprotocols development and testing | https://github.com/ngscopeclient/scopehal-apps | Logs: https://libera.irclog.whitequark.org/scopehal
_whitelogger has joined #scopehal
Degi has quit [Ping timeout: 248 seconds]
Degi has joined #scopehal
ablu has quit [Ping timeout: 260 seconds]
ablu has joined #scopehal
skipwich has quit [Remote host closed the connection]
skipwich has joined #scopehal
_whitelogger has joined #scopehal
ablu has quit [Ping timeout: 258 seconds]
ablu has joined #scopehal
<_whitenotifier-a> [scopehal] azonenberg pushed 1 commit to master [+0/-0/±3] https://github.com/ngscopeclient/scopehal/compare/7a4b35dc376a...1559b10b7964
<_whitenotifier-a> [scopehal] azonenberg 1559b10 - LeCroy: improvid robustness of 8b10b trigger handling and avoid sending malformed control characters when configuring a new trigger object
eightdot has quit [Server closed connection]
eightdot has joined #scopehal
lethalbit has quit [Remote host closed the connection]
lethalbit has joined #scopehal
d1b2 has joined #scopehal
<d1b2> <esden> test
<esden> test
<d1b2> <esden> ok should be back
<dingwat> mildly offtopic but someone in here might know, what's the absolute thinnest _and_ bendiest coax anyone has worked with? Specifically it should be at least thinner than 1.2mm and as flexible as possible
<azonenberg> dingwat: let me look up something h/o
<azonenberg> https://koaxis.com/kf047 this is 1.52mm and is among the thinnest i've worked with but its super noodly
<azonenberg> 1.19mm plus the FEP jacket
<azonenberg> Samtec ARC6 cabling is twinax, let me see how thin that cable is...
<dingwat> ok super noodly is good
<azonenberg> 0.94 x 1.27mm twinax
<dingwat> for context, our techs are complaining that the keysight power probe coax tips can break ceramic caps if you look at them wrong. And I want to build some Z0 probes for rail ripple measurements for use with our 8ch scope
<azonenberg> i dont know if you can buy just the cable but not as part of a harness
<azonenberg> dingwat: I use KF047 on my solder in probe tips
<azonenberg> specifically because of how noodly it is
<azonenberg> i normally terminate in a SMPM but have experimented with (still in progress) direct solder terminations to stripped pigtails as well
<d1b2> <azonenberg> thanks
<azonenberg> dingwat: in particular, KF047 has pretty high insertion loss in terms of dB/meter especially at higher frequencies compared to e.g. KF086 or KF120P
<azonenberg> but it's excellent at not putting stresses on tiny solder joints
<azonenberg> For complex solder in probe setups where i want low loss and good flexibility, koaxis has a custom in house solderable splice assembly for bridging from 0.047 to 0.086 cables
<azonenberg> one of my favorite probe cables is 30 inches of KF086 as a low loss home run spliced to 6 inches of KF047 as a strain relief leading up to the business end
<azonenberg> get some kapton tape, secure the KF047 to the DUT, tape down the splice or the very end of the KF086, and you're good to go
<azonenberg> This is going to be my standard test cable for most of my solder in probes once they hit production
<dingwat> ok that's very good info thank you
<azonenberg> (return loss of the splice is excellent, I order all of my cables with serial numbers and factory touchstone files and the data they've sent me has been great)
<azonenberg> in general koaxis is a custom cable shop, they will make you pretty much any combination of cable stock, length, markings, and terminations you want as long as it's physically possible
<dingwat> i'm gonna guess the answer to "how expensive" is "if you have to ask you can't afford it" hehe
<azonenberg> Actually no they're shockingly cheap as custom high end cables go. Those 30/6 KF086/KF047 hybrids i wanna say were on the order of $100-120 with MOQ of 2 units and that included serialization, vna testing, and serialized touchstone files
<azonenberg> and phase matching
<dingwat> holy moly. Ok that's nice. I dunno exactly how we'd do it, but probably just ask for a single connector on one end. Maybe just straight to BNC-M and put the scope close to the board :)
<azonenberg> You can order custom pigtails with stripped ends and whatever you want on the other end, assuming a compatible connector is available (e.g. BNC on a .047 probably isn't going to be possible for mechanical reasons)
<dingwat> ah I didn't think of that. Maybe some kind of press-fit, SMB or something. And then buy a pile of BNC to SMB cables
<azonenberg> Back in October 2024 I quoted six inches of KF047 with cut/stripped ends to SMA
<azonenberg> $40.25 @ single units, $26.50 each @ qty 5, down to $16 @ qty 500
<azonenberg> that included strip to shield/center conductor exposed for specified lengths
<azonenberg> dingwat: dm'd you my sales contact
<azonenberg> wrt z0 probes for rail measurements are you looking for 500 ohm resistive probes? or more like a 1x?
<dingwat> azonenberg: I'm undecided. Generally everything is gonna be <=3V3, so a 1x would be ok wrt to scope input voltage
<azonenberg> yes, but heavy loading can cause problems wrt power supply behavior
<azonenberg> by adding dc load
<dingwat> yeah
<azonenberg> i love my lecroy rp4030s and would love to make an open hardware competing design eventually
<azonenberg> anyway, if for power rails or otherwise you want some of my AKL-PT5s I can probably hook you up pretty soon even if i'm not officially taking volume production orders yet
<dingwat> so I think 500R probes is the better option, and should be at least better than what we're doing now (passive 10M probes)
<azonenberg> i have a production sample on my desk
<azonenberg> you've seen those right?
<dingwat> maybe? which one is that again?
<azonenberg> better than +/- 1 dB flatness from DC to 6 GHz and then slowly peaks, excellent performance out to 16+ GHz if you de-embed the response curve
<dingwat> azonenberg: 450R?
<azonenberg> yep
<dingwat> nifty. What flavor is the connector?
<azonenberg> SMPM
<azonenberg> the standard cable for it will be that 30/6 inch KF086/KF047 cable i mentioned
<azonenberg> dingwat: https://www.antikernel.net/temp/tline-probes.png S21 of some Z0 probes i benchmarked against
<azonenberg> red = pico ta061
<azonenberg> green = picoconnect 921 (20x, all the others 10x)
<azonenberg> blue = lecroy pp066
<azonenberg> orange = PT5 prototype
<azonenberg> measured with a demo of a copper mountain 18 GHz VNA with automatic fixture removal soldered/probed across a 2x thru
<azonenberg> +/- 1 dB to 6 GHz and <3 dB peaking out to almost 12 GHz
<azonenberg> then it starts to climb more but is still usable with a de-embed
Darius has quit [Server closed connection]
Darius has joined #scopehal
<_whitenotifier-a> [scopehal] azonenberg pushed 1 commit to master [+0/-0/±1] https://github.com/ngscopeclient/scopehal/compare/1559b10b7964...75c455aea587
<_whitenotifier-a> [scopehal] azonenberg 75c455a - PCIeTransportDecoder: make unrecognized or aborted request TLPs show up as error color
sgstair has quit [Read error: Connection reset by peer]
octorian has quit [Server closed connection]
octorian has joined #scopehal