sometimes, when i'm feeling sad, i lie down with a screen of some sort and read a new specification or a datasheet for a device i think is neat or cursed
this is how i know so much about technology
sometimes, when i'm feeling sad, i lie down with a screen of some sort and read a new specification or a datasheet for a device i think is neat or cursed
this is how i know so much about technology
i have a really profitable coping mechanism
the nice thing about like specifications or datasheets or something is that they're usually very engaging, sometimes quite cursed, and only hurt me consensually if at all. this is why i like them so much
today's morning sadness reading is this thing: https://www.microchip.com/en-us/development-tool/EV06P90A
it's 10 Mbit Ethernet! not from the 90s, from the 2020s!! very cool i think
@jpm yep i have a pair of phys
@whitequark Oh is that the new 2-wire industrial ethernet stuff?
intel ass naming scheme
@whitequark @jpm it's cool and doable with simple circuits. I should try! The funniest should be power over single pair ethernet. I see DC blocking caps so this should be possible.
Failure to heed this warning
@whitequark I'm sure there's already an errata somewhere with a couple hundred writes into undocumented registers in order to unlock basic functionality. That's usually how it goes with microchip's networking products in my experience
@slaeshjag h-haha, encouraging...
this chip has a pin called "GPIO0" which cannot be arbitrarily set or read via its registers
@whitequark The GPIO0 pin has no value.
@swetland I wanted this since 2014!
@whitequark I've been meaning to check this out now that the various single-pair ethernet PHYs are in production. Especially curious how complex/expensive the PoE options are because combining single-pair PHYs with little MCUs sounds like a lot of fun especially if you can deliver power too...
@jpm nope!
@whitequark let me guess, it drives a LED doesn’t it?
@jaseg oh yeah! some soft realtime too, it seems to be geared for AVB?
@whitequark I think this stuff could be a great alternative to CAN and the like in non-realtime applications
@jaseg it is literally just a normal ethernet phy. they say nothing about the upper layers except there's special support for gPTP that makes sure the timings are right
@whitequark Since you just read up on it, do you happen to know what the upper protocol layers in this stuff are projected to be? I was kind of wondering since IPv6+TCP or whatever is cheap to do with today’s chonky MCUs, but it adds a bunch of bandwidth overhead if all you’re sending is short packets.
@jaseg i think smoltcp might have 6lowpan as a PR somewhere?
@whitequark something like 6LowPAN might be neat, but I don’t think anyone ported that yet?
@whitequark Yeah, I see industry inertia as more of a limiting factor in these applications. Technically it should be fine with proper bandwidth planning.
@Laserboy only 50%?
(psst i could reverse engineer any such DLL)
@whitequark If you buy a 200k€/$/£ laser today there’s 50% chance it will communicate to your PC with string commands on serial ports, or it will be on USB 2 and have the most cursed DLL in existence and no documentation.
@Laserboy yep sure! i will give it an overview to get a rough estimate of effort, then (unless it's somehow interesting to me personally or very easy) i can quote you for a complete RE effort
@whitequark I really meant 50% serial or 50% cursed DLL, basically either. As far as the DLL goes I am forced to work in 32b because it only exists in that version. Sirah laser controls. If you wish I am happy to get that to you.
@Laserboy oh i do enjoy that! i just can't afford to do a complete end-to-end RE effort for free, and that's what i opened with
@whitequark Oh I apologize, I have no need of such RE, I though you might just enjoy looking at it for fun. And it’s not like I can get spending for that in this funding environment 🥲
@whitequark they're all like that
@rygorous speechless
@slaeshjag the PHY sure does behave in a ... weird manner
@whitequark @slaeshjag keep us updated. A friend designed a few boards around it for his sailing boat, I think that project is still in bring-up hell :v
@slaeshjag I got it to receive like... most of the time. Like maybe 9 out of 10 times. 1 out of 10 times it does something batshit
@whitequark @slaeshjag Yeah, that's what I've heard too. Friend of mine got it to transmit but not to receive, and now he wants to re-make the boards because he doesn't know if it's a microchip thing or if the boards weren't soldered up correctly :D
@slaeshjag and by "receive" I mean I have these issues with near end loopback alone
@slaeshjag i think the near end loopback mode in this PHY is fucked
@slaeshjag I *think* I found a bug with their RMII implementation but I'm still trying to make sense of it all
@whitequark @slaeshjag I'm starting to think those PHYs are just special in all the wrong ways, and that there's nothing wrong with the boards my friend put together. At one point, the devices and datasheets suddently disappeared of Microchip's website for like 5 months while mouser kept selling the chips :v
@slaeshjag ok i finally got the PHYs working, i had a bug in the RMII TX code as well as RX
took a bit but:
@slaeshjag zero writes via MDIO
@whitequark @slaeshjag did you have to do anything weird with registers or so, or was it all just according to the docs?
GNU social JP is a social network, courtesy of GNU social JP管理人. It runs on GNU social, version 2.0.2-dev, available under the GNU Affero General Public License.
All GNU social JP content and data are available under the Creative Commons Attribution 3.0 license.