peterm6881 has joined #Speedsaver
<peterm6881> hey Xogium
<peterm6881> are you there?
<Xogium> yes
<peterm6881> something must have gone wrong with your test earlier, the R3c can be powered from the otg port
<peterm6881> alone
<Xogium> hmm how do you know that ?
<peterm6881> I powered it from it via OTG on a wall wart, logged in via the NEC, then disconnected the TTL usb, so it was again just running on OTG
<peterm6881> when I reconnected the NEC and connected via puTTy again, it was still logged in to Buildroot
<peterm6881> excuse the bad grammar
<Xogium> hmm
<peterm6881> I dunno what happened when you tried it
<Xogium> well at least on my computer, it doesn't seem to work
<peterm6881> but it definitely stayed logged on despite running on purely OTG power
<peterm6881> now, im using a wall wart for OTG power, to replicate a power bank
<peterm6881> let me try it again with an actual power bank
<peterm6881> i mean, as soon as I plugged in the OTG, the green power LED came on
<Xogium> hmm I wonder if
<peterm6881> i knew id read somewhere you can power from either
<Xogium> the thing I did was to connect the usb with uart first
<peterm6881> interesting, i connected otg first
<peterm6881> by the way they are labelled ttl and otg
<peterm6881> let me replicate it how you described earlier
<peterm6881> gimme a sec
<peterm6881> i'll even use screen
<peterm6881> ok its booting
<peterm6881> ok im at the prompt, wait
<peterm6881> ok now ive connected another NEC usb to otg
<peterm6881> screen still ok so far, still at the login prompt. wait
<peterm6881> now ive disconnected the ttl
<peterm6881> wait
<Xogium> for me, as soon as I unplug uart, the green light goes out
<peterm6881> thats bizarre, sounds like you have a dodgy micro usb cable
<peterm6881> as soon as i reconnected the ttl and relaunched screen, it was still at the mangopi login prompt
<peterm6881> it can definitely be powered from OTG, no question. So I think you have a dodgy cable or something weird going on with the power on your usb ports
<Xogium> well here's the thing. I can swap the cables just fine and get exactly the same result
<peterm6881> something peculiar about the power on your usb
<Xogium> yeah, maybe. Whatever lol
<peterm6881> meh, dont worry about it. The important thing is I can keep it alive
<Xogium> heh
<Xogium> will try and port our stuff to it
<peterm6881> so, i can boot via laptop, and just keep it powered up all the time. I emailed mangobuge, did you read it?
<Xogium> yeah I saw
<peterm6881> exciting times
<peterm6881> it'll blow my mind if the software just runs
<Xogium> navit and such probably will
<Xogium> but I need to trim down the device tree
<Xogium> although… How many i2c buses do we get ?
<Xogium> the big problem is that we can't even trust the freaking schematics
<peterm6881> its bizarre their image boots ok, but the external tree doesn't. My problem is i have no idea what the difference is when it comes to booting up. Am I correct in saying the u-boot and kernel is untouched?
<peterm6881> let me check that
<Xogium> yeah its all the official stuff from them. I haven't modified anything related to that
<peterm6881> what Multiplexing Function did we decide they were using?
<Xogium> uh… let me check your email
<peterm6881> they're using uart1 for the bridge right?
<Xogium> multiplexing function 3
<peterm6881> hmm... this could be quite bad
<peterm6881> of course mangobuge never answered to confirm or correct my analysis
<peterm6881> it 100% hinges on if its uart1 in the bridge
<peterm6881> on, ffs
<peterm6881> question, can you tell which or how many i2c bus's are enabled currently?
<Xogium> only one
<Xogium> enabled in the device tree
<peterm6881> can you tell which it is? tw something
<Xogium> currently connected to a er… lcd panel we don't have
<Xogium> its just called i2c0
<peterm6881> this is too hard, because they have dedicated clock and data pins, labelled as such, so that is assumed to be an i2c connection. However i have no information what GPIO on the Allwinner chip they are physically connected to. The LCD panel referred to is probably connected via one of the funky ZIF sockets
<peterm6881> so, let me check the chip
<peterm6881> 3 TWI controllers on board
<peterm6881> so, enable i2c 1 and 2, in addition to 0
<peterm6881> the challenge for us is to reverse engineer this fucking board and deconstruct whatever the hell way they designed it, because none of it is documented
<peterm6881> we just have to figure out what works by trial and error
<Xogium> yeah… erk
<peterm6881> lol
<peterm6881> i just emailed you the basic version of the F1C200s, are you able to read the tables on pages 14 and 15
<peterm6881> f1c200s datasheet i meant
<peterm6881> it gives you the GPIO depending on what Multiplex Function is selected
<peterm6881> Now, the following is exposed to pins, what much is absolutely certain, ok?
<peterm6881> THAT much, i meant
<peterm6881> PA0 to PA3, and PE0 to PE11
<peterm6881> what im not 100% certain on is what GPIO they are currently using elsewhere, and what Multiplexing Function they have selected. I THINK its 3, out of a possible 6
<Xogium> hmm
<Xogium> will try
<peterm6881> I think its mode 3, because I believe thats the only one that gives you UART1 on unexposed pins
<peterm6881> specifically, PD1 to PD4
<peterm6881> the bottom line is, the chip supports 3 TWI interfaces, so enable i2c 1 and 2
<peterm6881> and we'll work out which works for our oled, either the dedicated Clock and Data pins, or TWI0 clock and data on PE11 and PE12 respectively
<peterm6881> I think the pins labelled clock and data are TW10, connected to GPIO PD0 and PD12
<peterm6881> although why would they provide pins, label them clock and data, but not enable them in their build.....
<peterm6881> its a head fuck, but if we decode their board, it brings people to our party
<peterm6881> 100 mA on a 2600 mAh power bank, 26 hours
<peterm6881> be interesting to see what it draws with a speedsaver image
<peterm6881> night
peterm6881 has quit [Remote host closed the connection]