electronic_eel has quit [Ping timeout: 260 seconds]
electronic_eel_ has joined #glasgow
PyroPeter_ has joined #glasgow
PyroPeter has quit [Ping timeout: 265 seconds]
PyroPeter_ is now known as PyroPeter
FFY00 has quit [Remote host closed the connection]
FFY00 has joined #glasgow
_whitelogger has joined #glasgow
mndza has joined #glasgow
Eli2_ has joined #glasgow
Eli2 has quit [Ping timeout: 264 seconds]
marcan has quit [Quit: Now where's my screwdriver...]
marcan has joined #glasgow
<electronic_eel_>
@rwhitby the level shifting is looking good to me
rafaelmartins_ has joined #glasgow
<rwhitby>
excellent, thanks for reviewing
<electronic_eel_>
seems like you want to go with 2 usb connectors (and not use a 3rd one as in my previous suggestion)
<electronic_eel_>
how about this then: make vusb of one of the connectors disconnectable with a jumper
<rwhitby>
what would the purpose of the 3rd one be again? I'm not sure I followed that.
<electronic_eel_>
then you could plug in a non-20v-safe usb device, like a luna, in one
<rwhitby>
Luna rev 0.3 has USB-C connectors I believe.
<electronic_eel_>
have the luna talk usb 2 on d+/d-, but command the device on the other connector to output 20v
<rwhitby>
Perhaps the second connector needs to be switchable between VBUS and 5V?
<rwhitby>
So Luna thinks it's just plugged into a Type-A.
<electronic_eel_>
yes, this would be the best solution, a 3-way jumper to select between vbus and 5v
<electronic_eel_>
you could also use a basic usb-c to mirco-usb or usb-a adapter cable to connect luna then. but you wouldn't need an exotic cable that has vbus cut
<rwhitby>
right. Maybe I can extend the VBUS test point at the top right somehow to create this option.
<rwhitby>
Since it has VBUS and 5V right there.
<electronic_eel_>
yes, good idea
<electronic_eel_>
(and there goes the space for your silkscreen ;-)
<rwhitby>
yes, I've been continually moving that around
<electronic_eel_>
what are the holes in the 4 corners of the pcb for?
emily has quit [*.net *.split]
rafaelmartins has quit [*.net *.split]
<rwhitby>
JLCPCB needs at least 2 or 3 tooling holes or assembly
<rwhitby>
They can be moved
<electronic_eel_>
I haven't used JLCSMT yet. don't they have the regular rail-like panels around the pcb and then some breakout-tabs are broken off to free the final pcbs after assembly?
<electronic_eel_>
but the user has to be a bit more aware of what pin is where to not short something than when using a traditional 3-way pin header where you put the jumper either on the right or left position
emily has joined #glasgow
marcan has joined #glasgow
jschievink has quit [Ping timeout: 240 seconds]
cyrillu[m] has quit [Ping timeout: 240 seconds]
randomplum has quit [Ping timeout: 240 seconds]
svenpeter has quit [Ping timeout: 240 seconds]
_alice has quit [Ping timeout: 240 seconds]
emily has quit [Ping timeout: 258 seconds]
whitequark[m] has quit [Ping timeout: 244 seconds]
<electronic_eel_>
do you have some space around the FUSB302 for the CC testpoints i suggested?
<electronic_eel_>
now that it is right of the usb traces running through i think there should be
<d1b2>
<rwhitby> yes, we can make space
<d1b2>
<rwhitby> what size test points do you suggest. I originally just made sure everything was accessible on vias for blue wiring, but you're talking about a real test point.
<electronic_eel_>
i thought about something you could easily probe with a scope
<d1b2>
<rwhitby> I also need to put some decoupling caps around the FUSB
<d1b2>
<rwhitby> So I'd better do those first, as they should take priority of locality.
<electronic_eel_>
yeah
<electronic_eel_>
the ideal testpoints for a scope would be plated through holes, about 0,5 to 1 mm diameter, and a similar sized gnd pad next to it
<electronic_eel_>
if there isn't space for the gnd, then you can still probe the gnd pads of the glasgow connector
<electronic_eel_>
should still be good since cc isn't that high speed
<d1b2>
<rwhitby> what do you suggest for decoupling of the FUSB302 ?
<electronic_eel_>
does it need anything special? i haven't used it before so i don't know. if it has no special needs, then a 100nF, maybe add a 1µF on the 5v rail
<d1b2>
<rwhitby> And that picture is fully routed.
<electronic_eel>
looking good now. now only the protection for vbus > 5v is missing, also for vsense for glasgow revC1 and before
<electronic_eel>
i posted links to some protection ics for d+- the other day, iirc ti and nxp have other ics in their usb-pb series that are designed for power protection and not data
<electronic_eel>
maybe they fit and are not tinypitch wlcsp
<d1b2>
<rwhitby> well, if ADC for revC1 and before is not safe, I may just remove VBUS from it. FUSB302 can report VBUS voltage instead.
<d1b2>
<rwhitby> ok, removed VBUS from ADC on Port A (replaced by a test point, so someone can always wire it back up again if they are aware of the risk)
<electronic_eel>
alternative would be a jumper between vbus and vsense. if you only have a revC2 or newer, you can safely solder in that jumper
<electronic_eel>
yeah, the solder jumper should be easy to use
<electronic_eel>
the silergy ic you posted allows only 5.5V, or am i missing something?
<electronic_eel>
what would the required features of such an ic be?
<electronic_eel>
i'd say it must survive 20v on the output and should protect from reverse power (switch on and output higher than input)
<electronic_eel>
you don't need current limiting as the glasgow will do that for you
<electronic_eel>
so the real barebones solution would be a p-fet and a diode
<electronic_eel>
not really good though because of the additional diode drop, when you set vio to 5v on glasgow it won't reach true 5v unless you power glasgow with more than 5v
<electronic_eel>
so if you add an additional schottky diode drop it might become a problem for some usb device
<electronic_eel>
so maybe one of those ideal diode controllers?
<electronic_eel>
maxim has some nice ones
<d1b2>
<rwhitby> We can measure the 1.2V supply too.
<d1b2>
<rwhitby> " In case the output is pulled higher than the input voltage under the shutdown, the SY6280 can block the current flowing from the output to the input. " from the SY6280 datasheet.
<d1b2>
<rwhitby> It does say maximum 6V on all pins though.
<electronic_eel>
yes, that won't fly with 20v
<rwhitby>
Thinking ahead, I actually want 50V protection, for when USB-PD is extended to 50V and I want to use this board to test the PD messaging to negotiate that.
<electronic_eel>
hmm, 50v will be much more tricky than 20v in my experience
<electronic_eel>
for example the ideal diode controller i mentioned are rated for around 40v iirc
<d1b2>
<rwhitby> yeah, I'm probably dreaming for that one
<electronic_eel>
when pd for 50v is finalized and first devices appear on the market, then there will probably also be specialized ics that can handle 50v
<d1b2>
<rwhitby> The FUSB302 is only rated to 21V max anyway
<electronic_eel>
uh, that is tight
<d1b2>
<rwhitby> oh, absolute is 28V
<electronic_eel>
ah, ok
<d1b2>
<rwhitby> ok, so we're looking for a load switch that can pass 20V, and block 20V reverse voltage.
<d1b2>
<rwhitby> (it doesn't actually need to pass 20V, but we're more likely to fine one that will block 20V by looking for one that will pass 20V)
<rwhitby>
I've used the NX20P5090UK before, but it's not very hobbyist-assembly friendly (WLCSP only)
fibmod_ is now known as fibmod
<d1b2>
<rwhitby> NX5P3090 would do the job too (WLCSP12)
<d1b2>
<rwhitby> I guess it's only those who want to do DFP stuff who need to install the load switch.
<electronic_eel>
although it has 0.4mm pitch, the pads are layed out in way that the middle pads have the same signal as outer ones. so you don't need complex fanout
<electronic_eel>
so you could add it without making manufacturing the pcb more expensive
<electronic_eel>
MAX16141 would be an alternative
<electronic_eel>
it is in qfn, but a bit more expensive than the NX5P3090
<electronic_eel>
and you need more supporting circuitry around it
egg|anbo|egg has joined #glasgow
<d1b2>
<rwhitby> Here's a simpler solution - add another jumper for DFP operation.
<rwhitby>
If you want to be a Source, then you insert a jumper, and make sure you're always the source (i.e. don't accept power role swap). If you want to be a Sink or Snoop, then remove the jumper.
<rwhitby>
@electronic_eel: I beleive that resolves all the issues :-)
<electronic_eel>
ok. only issue could be if you want to test fancy role swap code
<rwhitby>
next you'll be asking for fast role swap support ;-)
<electronic_eel>
oh, i'm thinking about building a source for up to 20v. so some control output for an external smps
<electronic_eel>
also for pd 3.0, so with fine current control ;)
<rwhitby>
I can do a different board with NXP PTN5110 and OnSemi NCP81239 for that.
<electronic_eel>
an addon for the addon
<rwhitby>
well, you can always connect to the 2x3 header on this board and do your own load switch implementation
<electronic_eel>
yes, some external smps and i2c dacs to set the limits
<electronic_eel>
but it shouldn't be on this board, would make it too complicated
<rwhitby>
I'm not even sure the DFP stuff will work in practice, if the sinks draw too much current during the implicit contract period it might overcurrent glasgow.
<rwhitby>
i.e. plug in a USB fan and see what happens.
<electronic_eel>
you could plug your bench supply to the vbus header
<electronic_eel>
as long as you don't want to negotiate >5v i don't see an issue
englishm_ has quit [Remote host closed the connection]
key2 has quit [Remote host closed the connection]
<d1b2>
<rwhitby> A well-behaved USB device should only draw 100mA max before enumeration.
<electronic_eel>
well-behaved and usb devices...
<d1b2>
<rwhitby> 😉
<d1b2>
<rwhitby> So, you could use the VBUS_DFP as a low-current enable signal to a high power lab supply that connects to VBUS and GND on that 2x3 header.
<d1b2>
<rwhitby> I'm starting to like this 2x3 general purpose header.
<electronic_eel>
it is not perfect, but it allows for a lot of flexibility
<d1b2>
<rwhitby> I could put a 5V zener on VBUS_DFP, and that would ensure nothing blows if someone leaves the DFP jumper in place and power role swaps to 20V.
englishm_ has joined #glasgow
englishm_ has quit [Remote host closed the connection]
<d1b2>
<rwhitby> The good thing about the jumper is that the usual case of being a sink and not snooping is all jumpers out.
<electronic_eel>
yeah, something more beefy than the small tvs on glasgow. for example SMAJ5A
englishm_ has joined #glasgow
<d1b2>
<rwhitby> anything suitable at jlcpcb ?
<electronic_eel>
don't they have SMAJ5A?
<electronic_eel>
oh, it is called SMAJ5.0A
<d1b2>
<rwhitby> well, they have 15 different types of that one. which one to pick?
key2 has joined #glasgow
<d1b2>
<rwhitby> and which part in KiCAD has the same footprint?
<electronic_eel>
it is a quite generic jellybean part, just like a bss138 small signal nfet. you can pick pretty much any of them and it will fit and work
<electronic_eel>
the sma diode footprint from the kicad lib works
<rwhitby>
wow, that's a big area footprint. not sure where I can put that.
<rwhitby>
it's pretty much the same size as the 2x3 jumper area
<electronic_eel>
but usually the usb pd device will deliver something much higher, like 9v or 12v. that should well be within the conduction region of the tvs. that will then trip the ocp of the usb pd psu, hopefully before anything is damaged
<electronic_eel>
hmm, the BZD... is a classic zener, not a tvs. don't know how well it will work for discharging a lot of power for a short time
<electronic_eel>
ok, i gotta go shopping groceries before the stores fill up with people. i can take another review look over the whole addon sometime during the weekend if you want
<d1b2>
<rwhitby> thanks for your excellent inpu
<d1b2>
<rwhitby> input
<rwhitby>
@electronic_eel: I've pushed v0.4a to github, would welcome another review over the weekend
umarcor has joined #glasgow
futarisIRCcloud has quit [Quit: Connection closed for inactivity]
emily has quit [Ping timeout: 240 seconds]
whitequark[m] has quit [Ping timeout: 240 seconds]
whitequark[m] has joined #glasgow
egg|anbo|egg_ has joined #glasgow
emily has joined #glasgow
egg|anbo|egg_ has quit [Remote host closed the connection]
samlittlewood has quit [Quit: samlittlewood]
Eli2| has joined #glasgow
d_olex_ has joined #glasgow
egg|anbo|egg_ has joined #glasgow
infinigon_ has joined #glasgow
noopwafe1 has joined #glasgow
Eli2_ has quit [*.net *.split]
PyroPeter has quit [*.net *.split]
d_olex has quit [*.net *.split]
infinigon has quit [*.net *.split]
coon has quit [*.net *.split]
noopwafel has quit [*.net *.split]
etrig has quit [*.net *.split]
d_olex has joined #glasgow
etrig has joined #glasgow
PyroPeter has joined #glasgow
infinigon has joined #glasgow
coon has joined #glasgow
d_olex has quit [Max SendQ exceeded]
infinigon has quit [Ping timeout: 254 seconds]
infinigon_ is now known as infinigon
egg|anbo|egg_ has quit [Remote host closed the connection]
egg|anbo|egg_ has joined #glasgow
emily has quit [Ping timeout: 244 seconds]
jschievink has quit [Ping timeout: 244 seconds]
emily has joined #glasgow
jschievink has joined #glasgow
bvernoux has joined #glasgow
egg|anbo|egg_ has quit [Remote host closed the connection]
samlittlewood has joined #glasgow
egg|anbo|egg_ has joined #glasgow
lxdr has quit [Ping timeout: 260 seconds]
simukis_ has quit [Remote host closed the connection]
lxdr has joined #glasgow
simukis_ has joined #glasgow
tomtastic has quit [*.net *.split]
mwk has quit [*.net *.split]
alen has quit [*.net *.split]
merry has quit [*.net *.split]
lethalbit has quit [*.net *.split]
Ultrasauce has quit [*.net *.split]
^^makoto_ has quit [*.net *.split]
interruptinuse has quit [*.net *.split]
GregNGM has quit [*.net *.split]
plaes has quit [*.net *.split]
janhenrik_ has quit [*.net *.split]
mndza has quit [Read error: Connection reset by peer]
tomtastic has joined #glasgow
merry has joined #glasgow
Ultrasauce has joined #glasgow
alen has joined #glasgow
lethalbit has joined #glasgow
GregNGM has joined #glasgow
^^makoto_ has joined #glasgow
plaes has joined #glasgow
interruptinuse has joined #glasgow
janhenrik_ has joined #glasgow
mwk has joined #glasgow
fibmod has quit [Ping timeout: 246 seconds]
fibmod has joined #glasgow
jstein has joined #glasgow
egg|anbo|egg_ has quit [Remote host closed the connection]
noopwafe1 is now known as noopwafel
egg|anbo|egg_ has joined #glasgow
egg|anbo|egg_ has quit [Remote host closed the connection]
bvernoux has quit [Quit: Leaving]
egg|anbo|egg_ has joined #glasgow
egg|anbo|egg_ has quit [Remote host closed the connection]
egg|anbo|egg_ has joined #glasgow
egg|anbo|egg_ has quit [Remote host closed the connection]