azonenberg changed the topic of #scopehal to: libscopehal, libscopeprotocols, and glscopeclient development and testing |,, | Logs:
Degi has quit [Ping timeout: 240 seconds]
Degi has joined #scopehal
<azonenberg> So i'm confuzzled
<azonenberg> (ping lain)
<azonenberg> I set up an 8-layer sim in sonnet pro with a diffpair using the MAXWELL geometry
<azonenberg> that changed from layer 1 to layer 6, so ref'd to layer 2 vs 7 ground
<azonenberg> With no stitching vias
<azonenberg> And i'm not seeing the current doing what i expected
<azonenberg> I see current in the ground planes under the traces, but i can't figure out how that current gets from one plane to the other
<azonenberg> i was hoping to see a clear difference with and without stitching vias, and i'm not
<lain> hm
<azonenberg> So i'm feeling like i set up the sim wrong
<azonenberg> you can see the tighter meshing around the stitching vias
<azonenberg> but what i *expected* to see is the return current splitting off around the antipad and flowing into the via
<azonenberg> then out the via and back along the edges of the plane until it was under the traec on the othe rlayer
<azonenberg> I'm not seeing that
<lain> it's true that the signal will couple into any nearby metal and use it as a reference plane, but... yeah I'd expect to see... what you just described
<azonenberg> so i'm confused as to how current can be coming in port 1 and out port 2, but not obviously traveling between them
<azonenberg> i.e. there appears to be current in layer 2 and 7 with no path between them
<azonenberg> for this simplified test layers 3-5 are blank but that shouldnt change anything?
<azonenberg> i see l1's return current on l2 and l7's return current on l6 as expected
<azonenberg> you can see in the screenshot the pair comes in from the left then the return plane current is ~nonexistent to the right of the vias
<azonenberg> which is where the pair is ref'd to the other plane
<azonenberg> one possibility is that if you look right at the point between the two diffpairs (9 o'clock on the via antipad)
<azonenberg> you see current flow there
<azonenberg> i wonder if we're seeing differential mode current flow under the pair? i.e. there is no flow between the layers because instead it's flowing in under one leg and out under the other
<azonenberg> the ports are configured in differential push-pull mode so one is always equal and opposite to the other
<azonenberg> hmmm i am seeing some interesting current flows on the edges of the ground plane at 20 GHz
<azonenberg> This bears further investigation. maybe i need to make the sim box bigger to really see the effects
electronic_eel has quit [Ping timeout: 256 seconds]
electronic_eel_ has joined #scopehal
electronic_eel_ has quit [Ping timeout: 240 seconds]
electronic_eel has joined #scopehal
_whitelogger has joined #scopehal
Nero_ has joined #scopehal
Nero_ is now known as NeroTHz
<azonenberg> lain, monochroma, noopwafel, NeroTHz, degi: how's this pretty-final manual draft look?
<azonenberg> might add a mechanical drawing still
<azonenberg> but i think it's good otherwise
<NeroTHz> 404
<azonenberg> oops
<azonenberg> Dont have a good way of importing the pcb layout into my cad tool so i wont be doing a full mech drawing
<azonenberg> but i am adding a "mechanical specifications" section with dimensions
<NeroTHz> My go-to way for doing the PCB layout in cad is simply using a PDF export as image layer
<NeroTHz> I don´t know if I first have to convert the PDF export to PNG with illustrator or something
<azonenberg> I mean i can do the layout but i cant do pcb + components + enclosure in one render easily
<NeroTHz> ah I see
<NeroTHz> 3 Theory of operation: I would avoid saying things like ´you may be familiar with´. I personally would re-write the first sentance as ´The AKL-PT1 probe is a transmission line probe. The principle of operation of this probe is different from the more common high-impedance passive probe, or active probes.¨
<NeroTHz> but I´m maybe too used to writing academic stuff :p
<azonenberg> "many engineers are familiar with"
<NeroTHz> words like ´very´, ´carefully´, etc make little orange strobe lights go off in my head :p
<azonenberg> and yeah even when writing academic stuff i try to avoid being too obfuscatory in my tone
<NeroTHz> I would also suggest, if not to hard, to add more images for figure 2 and figure 3
<azonenberg> Reuploaded with a slight clarification there, as well as the mechanical characteristics section
<NeroTHz> like put 3 images next to each other with different ´stages´ of the removal/insertion
<azonenberg> I might get my wife to help me with that
<azonenberg> its a 2 handed process
<azonenberg> cant snap photos too
<NeroTHz> another thing (I know this is more a formating things but still): I would make the tables narrower. Like the table for 7 and 8.1 and 8.2 can easily be made narrower to just fit the text. It makes it easier to read i think?
<azonenberg> Fixed
<NeroTHz> okay. I´ll look at it again in a bit, but now I need to actually pretend I´m doing work at the office :p
<azonenberg> Also did some more accurate analysis based on the risetime curves etc and concluded my input capacitance is probably closer to 1.1 pF vs the 1.4 i had there
<azonenberg> but that's still a rough estimate which is why i didn't put a tolerance range on that spec
<azonenberg> Without a full 3d field solver model of the actual tip + socket geometry it will be hard to get better
promach3 has quit [Quit: killed]
maartenBE has quit [Ping timeout: 264 seconds]
juli965 has joined #scopehal
maartenBE has joined #scopehal
promach3 has joined #scopehal
juli965 has quit [Quit: Nettalk6 -]
<azonenberg> woop, first ten probes are boxed up and ready to ship
<azonenberg> I'm including hardcopy cal reports with them but just a little insert with a link to the URL for downloading the manual. So there's room to make last minute edits before they reach backers
<miek> nice!
<azonenberg> I'm doing the domestic ones first since customs forms are painful to fill out when you're doing this many
<azonenberg> And i'm still missing shipping addresses from several backers so those are obviously last priority
<miek> i ordered parts to make one of those sma soldering heatsinks, though i couldn't find the right threading die locally so that'll take a while from china
<azonenberg> Oh cool
<azonenberg> I've never observed problems with my SMA soldering but i tend to work pretty fast. Also the ones I use these days have a skinny center pin that heats up in a fraction of a second
<azonenberg> the body gets hot but not THAT hot and i've never observed the insulator moving
<azonenberg> It's possible i've had tiny movements that i havent noticed, though. That's one of the reasons i wanted to pick up a gage set
<azonenberg> once i have one, gaging all cables and adapters will likely become an annual task, as well as doing each newly assembled board before using it
<azonenberg> maybe every 6 months or so, but likely not more than that. i'm a one man shop so i doubt the cables get a ton of wear in terms of matings per unit time
<miek> yeah. though i also do a clean & check right before connecting anything particularly expensive, just in case
<azonenberg> i do a visual inspection under the microscope of stuff perioically, and clean occasionally with an IPA soaked lint free swab to remove any metal shavings or other debris that might have got in threads etc
<azonenberg> Once i get the gage set it will definitely become a routine PM task
<miek> i don't think i've had problems with board-mounted connectors moving (though i haven't had the tools to check before :p). the problem i'm having is with making cables - i'm pushing a bit of pressure on the cable to seat it nicely while soldering, and i think that can push the dielectric forward
<azonenberg> ah ok
<azonenberg> yeah i am not making cables
<azonenberg> if i was i'd be a lot more worried
<miek> did you see my gauging of a random cheap cable i had btw?
<miek> the pin was protruding over 40thou D:
<azonenberg> was that the one that wrapped around?
<azonenberg> lol
<miek> yeah
<azonenberg> and yeah i am not making cables, i'm not buying really cheap cables, and i'm also not letting randos play with them in a teaching lab etc
<azonenberg> So i havent been tooooo worried yet
<azonenberg> but i think it's good practice as my collection of RF gear grows
<azonenberg> same reason i got a torque wrench
<azonenberg> Pretty much all of my coax came from either minicircuits, crystek microwave, or CD International
<azonenberg> while cdint is definitely a cut below minicircuits and crystek they're not cheap cables either, i paid $20 a pop for many of them
<azonenberg> So i expected decent quality
bvernoux has joined #scopehal
juli965 has joined #scopehal
bvernoux1 has joined #scopehal
lukego_ has joined #scopehal
StM_ has joined #scopehal
StM has quit [*.net *.split]
lukego has quit [*.net *.split]
_whitenotifier-b has quit [*.net *.split]
lukego_ is now known as lukego
promach3 has quit [Write error: Connection reset by peer]
promach3 has joined #scopehal
NeroTHz has quit [Quit: Leaving]
maartenBE has quit [Ping timeout: 256 seconds]
maartenBE has joined #scopehal
bvernoux1 has quit [Quit: Leaving]
bvernoux has quit [Read error: Connection reset by peer]
juli965 has quit [Quit: Nettalk6 -]