<rqou> i did not realize just how easy a quick and dirty bare die decap is
<rqou> azonenberg: ^
_whitelogger has joined ##openfpga
stoopkid__ has joined ##openfpga
<azonenberg> rqou: didn't melt your face? That's always a good thing i guess
<azonenberg> And wet etches are great for uncontrolled depth, like using HF to strip to bare silicon
<azonenberg> if you have an etch stop they work well
<azonenberg> But it's very hard to control how far you're going, and get uniform depth across the whole chip
pie_ has quit [Ping timeout: 240 seconds]
ZipCPU|Laptop has joined ##openfpga
<rqou> whee, second decap and i spilled chemicals all over the floor
<rqou> nothing on my body though, so i guess that's good enough?
<rqou> azonenberg_work: nor flash is on surprisingly-modern process nodes
<azonenberg_work> like?
<azonenberg_work> i think 65
<azonenberg_work> maybe 45
<rqou> idk, maybe you can look at my recent photo and guess?
<azonenberg> Below 130 they all look about the same
<rqou> i thought they were all >90nm or something
<azonenberg> optically
<azonenberg> oh no
<rqou> i guess less mm^2 --> less money
<azonenberg> sec...
<rqou> anyways, that photo was a neo40, but otherwise it's not calibrated
<azonenberg> Basically below ~130 all you have is giant power distribution on the top layer
<azonenberg> and the fine details are sub-optical resolution
<azonenberg> so you cant really distinguish them optically
<azonenberg> the most you could do is know the memory capacity a priori
<azonenberg> and know the die area used by the array
<azonenberg> then calculate the bitcell size and extrapolate
<rqou> so i actually do know the memory size, but i'm too lazy to go and calculate it right now
<rqou> also, the TSOP package is suuper inefficient in terms of use of space
<azonenberg> Duh
<azonenberg> most leaded packages are
<azonenberg> heck, most non-WLCSP are :p
digshadow has quit [Ping timeout: 258 seconds]
<rqou> the TPS61029DRC isn't too bad
<rqou> but then that's an analog part
<rqou> btw azonenberg interesting thing
<azonenberg> so this is 65 nm
<rqou> the die ID on micron parts isn't officially anywhere in the datasheet, but it is in the filename and footer
<azonenberg> lol
<rqou> not sure if intended or not
<rqou> anyways, the flash i decapped was a MT28EW128ABA
<rqou> once the relevant parts show up, i should take better photos and put them on pr0n
<rqou> anyways, guess what
<rqou> due to reasons, i will soon have _three_ PM-ADF adapters
<rqou> rather than the desired one
<rqou> azonenberg: what's the name of those tiny vials you use to store dies?
<azonenberg_work> centrifuge tubes
<rqou> 1.5 mL or 0.5 mL?
<azonenberg_work> I think i mostly use 1.5s
<azonenberg_work> john uses 20s or 50s for larger dies like x86s etc
<rqou> wow, amazon is now totally confused just what type of thing i buy :P
<azonenberg_work> one time it saw targets and pistol cleaning supplies and thought i wanted to buy baby products
<azonenberg_work> soooo... :p
<rqou> what
<rqou> how does the AI even learn that?
<azonenberg_work> Good question
<rqou> although my housemate did at one point buy some random (but unusual) household item for the shower or whatever, and then amazon started suggesting some elastic band thing that pregnant women use to make the weight of a baby less uncomfortable
<rqou> so apparently only housewives/homemakers care about fixing the shower? :P
<azonenberg> lol
<rqou> oh btw afaict this drain cleaner i bought is almost pure sulfuric acid
<rqou> no dyes
<rqou> the power of walmart being cheap :P
<rqou> maybe it still has inhibitors, but i just put it on my shitty hotplate with basically no temp regulation and crank it to max, so it easily hits 300+ degrees C and destroys any inhibitors
* azonenberg hopes you did that outside with at least some PPE
<rqou> inside, but with PPE
* azonenberg slaps rqou with a clue stick
<rqou> i did it next to the window, and with a huge box fan blowing the air out
<azonenberg> So local exhaust
<azonenberg> Not a fume hood, but next best thing
<rqou> yeah, the whitequark method
<rqou> but even shittier because i couldn't find an inline fan at home depot and didn't want to have to go order one
<awygle> i feel like whitequark would object to your repeated characterizations of that method
<awygle> but what do i know
<rqou> i was also using a respirator+full face shield
<rqou> because i heard that acid in eyes is bad? :P
<azonenberg> Generally, yes
<rqou> oh yeah, and this decap run proves that the ebay/china-grade beakers were indeed borosilicate
<rqou> :P
<azonenberg> lol
<rqou> btw in case you haven't noticed, i actually do adjust the amount of cavalier vs careful as a function of the danger
<rqou> it's not YOLO all the time
<awygle> we mock and shame because we don't want you to die (or at least i do and don't, shouldn't speak for others)
<whitequark> rqou: so in general
<whitequark> the idea is that PPE is your last line of defense
<whitequark> when everything else goes quit
<rqou> yes, I'm aware
<whitequark> ok
<rqou> i actually got no chemicals on myself at all while doing decap
<rqou> i just make it sound more dangerous than it actually was
<awygle> rqou is an edgey edgelord. try not to melt your bones.
digshadow has joined ##openfpga
_whitelogger has joined ##openfpga
digshadow has quit [Quit: Leaving.]
digshadow has joined ##openfpga
soylentyellow has quit [Quit: Leaving]
soylentyellow has joined ##openfpga
digshadow1 has joined ##openfpga
soylentyellow has quit [Max SendQ exceeded]
soylentyellow has joined ##openfpga
digshadow has quit [Ping timeout: 248 seconds]
digshadow has joined ##openfpga
uelen has joined ##openfpga
digshadow1 has quit [Ping timeout: 240 seconds]
<rqou> azonenberg: so, for the xc95288xl i've got
<rqou> do you want a bare die decap or a live decap?
<rqou> (this is the one you told me not to mail you)
<azonenberg> Bare die
<azonenberg> i have no plans to probe it
<rqou> i'm thinking of doing another batch today because of just how quick a sulfuric bare die decap is
<azonenberg> Lol, go for it
<azonenberg> Unrelated, http://thanatos.virtual.antikernel.net/unlisted/uart-decode2.png - anyone have thoughts on this UI?
<rqou> hmm, should i try to jtag read this cpld first?
<rqou> or meh, we can land probe pads on it later if we really cared
<rqou> *land probes
<azonenberg> i'm trying to tweak my scope UI to make more use of context menus and such, so there's less fluff in the way of your waveforms
<eduardo> azonenberg: are you there?
<azonenberg> more minimalistic
<azonenberg> eduardo: yeah
<eduardo> Did you read the email from clifford about the proposal for the darpa project?
<azonenberg> Saw it
<azonenberg> Was out all day and didnt get much time to do anything
<rqou> azonenberg: you should consider scanlime's (?) suggestion about ingesting sigrok plugins :P
<azonenberg> I might at some point
<rqou> azonenberg: is that UI all custom-rendered? is it using opengl or something else?
<eduardo> azonenberg: We need you proposal until tomorrow.
<azonenberg> Custom rendered in cairo, although i wouldnt be opposed to doing a GL port if performance became an issue
<azonenberg> eduardo: Tomorrow? :o
<rqou> i want a webgl port :P
<azonenberg> what time tomorrow?
<rqou> er, what are you two scheming about? :P
<eduardo> azonenberg: yes, we are already in the budgeting process. Things move fast.
<azonenberg> rqou: grant proposal involving, among other things, funding dev of that eFPGA we've been thinking about
<rqou> hmm, i'm interested in working on this if possible
<eduardo> azonenberg: I wrote you already a placeholder to give the people an idea what it is about. But you definitely need to rewrite it as everything I wrote is a damn lie :-)
<azonenberg> eduardo: Can it wait until about 19:00 Pacific time tomorrow?
<azonenberg> its almost midnight here and i'm about to sleep, then i'll be out at a training event during the day
<eduardo> rqou: you should finish your study. Once you are done you can join. The thing is, this is going to be an admin nightmare.
<rqou> right, of course
<eduardo> azonenberg: Yes, thats ok.
<azonenberg> ok, good :) remind me tomorrow and i'll bang it out over the evening
<eduardo> rqou: and you will not miss the party as it will last 4 years anyway.
<rqou> heh nice
<azonenberg> eduardo: i kinda have my hands full this week with this http://thanatos.virtual.drawersteak.com/unlisted/books.jpg
<eduardo> rqou: I also got confirmation today that we received an additional client which will work with us on opensoruce FPGA toolchain for the coming years.
<eduardo> rqou: No ,I can not tell what company it is.
<eduardo> azonenberg: yes, I have seen your tweet, but this proposal paragraph can not wait.
<azonenberg> I'll see what i can do in the next half hour
<azonenberg> then more tomorrow afternoon
<eduardo> great
<eduardo> thanx
<azonenberg> i feel like i'm back in grad school, working on grant proposals while sitting next to a desk covered in textbooks :p
<azonenberg> Only difference is, now i have a day job to do on top of that
theMagnumOrange has quit [Ping timeout: 240 seconds]
<azonenberg> rqou: anyway, thoughts on the UI?
<azonenberg> Goal is to make something that will work well between 1080p and 4k resolution, with anything from 2-4 analog channels up to 50+ digital channels from an on-FPGA LA
<azonenberg> with complex protocol decodes
<azonenberg> i dont have a trigger system figured out yet but ti will be accessed through the channel-name buttons at the left side of the display
<azonenberg> i have to add a Y axis scale to the analog channels for voltage too
DingoSaar_ has quit [Remote host closed the connection]
DingoSaar_ has joined ##openfpga
<cr1901_modern> All this work because you don't like the license :P?
<azonenberg> eduardo: look better?
<azonenberg> cr1901_modern: well i also wanted to do things sigrok couldnt
<azonenberg> Like, last time i checked they couldn't multi-source
<azonenberg> i want to be able to hook up to multiple signal sources, cross-trigger, and align all of the captures etc to the correct offsets in time
<azonenberg> even if they're not all the same length, or phase shifted, etc
<cr1901_modern> Hmmm well I'll try it if you release it... I still mainly use gtkwave for "capture" (for some definition of "capture")
<azonenberg> gtkwave is meant for postprocessing
<azonenberg> Not interactive stuff
<eduardo> azonenberg: For me it looks good. Anyway, you will need to implement it. I will add some praising sentences about Mr. Zonenberg s competence to implement all this :-)
<azonenberg> this is very much intended to be an interactive UI
<azonenberg> So it has to include all of the logic for trigger control etc, which gtkwave lacks
<azonenberg> my very first LA tool actually captured to a VCD then launched gtkwave
<azonenberg> but it was a pain because you couldnt configure triggers in the app
<cr1901_modern> azonenberg: An interactive UI doesn't really matter to me all that much when 90% the scopes I use are FPGAs jamming a 32kb buffer down an 115200 baud UART :P
<cr1901_modern> It's not real time anyway
<azonenberg> Lol see, right now i am using it with a lecroy that pushes a multi-MB buffer down 100mbit ethernet (still slow, have to debug)
<azonenberg> And the next-gen backend, starshipraider, will be moving GB of buffer down a 10gbit ethernet link
<azonenberg> But if you use smaller captures, it will totally be interactive
<eduardo> azonenberg: If you want to learn about good interactive interfaces, I suggest http://worrydream.com/ He is one of my heroes.
<azonenberg> I want to add an eye plotting filter? decoder? to it
<azonenberg> cr1901_modern: and it's on github in the antikernel repo, but needs some refactoring and tweaking to be useful to folks other than myself
<azonenberg> this is an old project that i'm kinda reviving and cleaning up
<azonenberg> the new UI looks a lot nicer already and is only gonna get better
theMagnumOrange has joined ##openfpga
<azonenberg> eduardo: basically so far what i'm trying to do is take my existing LA viewer app
<azonenberg> Add the features i liked from LeCroy's UI, like the general cleanliness and use of context menus rather than always-present UI elements
<azonenberg> but adapt it to scale to larger screens and numbers of channels
<azonenberg> Like, a 4K display with six analog channels, 30 digital buses, and an eye diagram
<azonenberg> or something
<azonenberg> So rather than having channel names on the bottom like they do, i'm putting them on the side
<azonenberg> With each waveform in its own row, rather than all jumbled together like a typical oscilloscope does
<azonenberg> closer to how a LA viewer or logic simulator does it
<azonenberg> but for mixed analog and digital waveforms
<azonenberg> The protocol decoder rendering is kinda borrowed from saleae's "over the top of the waveform" renders, except I allow you to stack multiple decodes to do higher level protocol analysis
<azonenberg> And rather than having the decode icons all one color like saleae, i borrowed a page from lecroy and had them color-coded by functionality
<azonenberg> so like Ethernet has preamble gray, SOF yellow, MAC cyan, ethertype pink, data dark blue, then CRC green or red (although i don't actually check the CRC so it's always green for now :P)
<azonenberg> Seems like sigrok does that too
<azonenberg> Another thing i do that i havent seen other tools do is support "display compression", where long periods of inactivity (all signals have the same value) are collapsed and the timebase has a "break" in it
<azonenberg> great for looking at sparse activity on a long bus
<azonenberg> Have to figure out how to indicate inactivity on an analog signal, right now it only works for digital
<eduardo> azonenberger: That tool fits very well into our overall stragegy of providing open source tools for chip designers. (the best tools are those who are done by the users themself :-)
<azonenberg> eduardo: and unlike sigrok et al it's permissively licensed
<azonenberg> it's currently somewhat glued into my antikernel toolchain but i'm in the process of decoupling it to make it more standalone
<eduardo> azoneberg: I added a paragraph about you
<azonenberg> A bit awkward, i'll fix the phrasing tomorrow
<azonenberg> Bedtime for me :p
<eduardo> azonenberg: good night
<eduardo> azoneberger: thank you for helping out with the porposal. its ok for now.
<sn00n> le me livin in the wrong time zone :<
pie_ has joined ##openfpga
pie_ has quit [Ping timeout: 260 seconds]
eduardo_ has joined ##openfpga
eduardo has quit [Ping timeout: 240 seconds]
m_t has joined ##openfpga
digshadow has quit [Ping timeout: 240 seconds]
<rqou> xc95288xl is a failure. apparently large tqfps contain more plastic than expected, leading to unexpected reactions
pie_ has joined ##openfpga
soylentyellow has quit [Ping timeout: 248 seconds]
azonenberg_work has quit [Ping timeout: 255 seconds]
azonenberg has quit [Ping timeout: 246 seconds]
Hootch has joined ##openfpga
soylentyellow has joined ##openfpga
xdeller has quit [Quit: Leaving]
<sn00n> rqou: you mean decapping it?
azonenberg_work has joined ##openfpga
azonenberg has joined ##openfpga
<sn00n> so the bubbly layer on top of the dice of an ice40hx1k ... is it Si3N4 or is it Polyimide?
pie_ has quit [Ping timeout: 248 seconds]
eduardo_ has quit [Ping timeout: 260 seconds]
eduardo_ has joined ##openfpga
eduardo_ has quit [Ping timeout: 240 seconds]
sunxi_fan has joined ##openfpga
sunxi_fan has quit [Client Quit]
<whitequark> unlike sigrok it's not backed by a community though
eduardo_ has joined ##openfpga
eduardo_ has quit [Ping timeout: 248 seconds]
eduardo_ has joined ##openfpga
enriq has joined ##openfpga
enriq has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
pie_ has joined ##openfpga
pie_ has quit [Remote host closed the connection]
pie_ has joined ##openfpga
sunxi_fan has joined ##openfpga
pie_ has quit [Ping timeout: 248 seconds]
theMagnumOrange has quit [Ping timeout: 248 seconds]
xdeller has joined ##openfpga
eric_j has quit [Read error: Connection reset by peer]
theMagnumOrange has joined ##openfpga
digshadow has joined ##openfpga
digshadow has quit [Ping timeout: 255 seconds]
<sn00n> okay, TIL how 2 crack an ice40 dice m)
digshadow has joined ##openfpga
Hootch has quit [Ping timeout: 240 seconds]
pie_ has joined ##openfpga
xdeller has quit [Read error: Connection reset by peer]
xdeller_ has joined ##openfpga
xdeller_ has quit [Read error: Connection reset by peer]
xdeller has joined ##openfpga
<jn__> pie_: i should make a screenshot of that
<jn__> … in jpg
<pie_> heh
<awygle> This is occurring to me like, maybe a day late but rqou where did you find a wal mart in the bay?
stoopkid__ has quit [Quit: Connection closed for inactivity]
<rqou> so, i think I've discovered all of the catastrophic failure scenarios of my ghetto decap setup
<rqou> don't ask how
xdeller has quit [Ping timeout: 246 seconds]
enriq has joined ##openfpga
DingoSAL has joined ##openfpga
DingoSAL has quit [Client Quit]
DingoSaar_ has quit [Ping timeout: 248 seconds]
<awygle> Is BLE fundamentally insecure or do people just suck?
m_t has quit [Quit: Leaving]
ZipCPU|Laptop has quit [Ping timeout: 240 seconds]
ZipCPU|Laptop has joined ##openfpga
enriq has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
gnufan has quit [Quit: Leaving.]