ChanServ changed the topic of #glasgow to: glasgow interface explorer · code https://github.com/GlasgowEmbedded/glasgow · logs https://freenode.irclog.whitequark.org/glasgow · discord https://1bitsquared.com/pages/chat · production https://www.crowdsupply.com/1bitsquared/glasgow · no ETAs at the moment
Stormwind_mobile has joined #glasgow
samlittlewood has quit [Quit: samlittlewood]
<_whitenotifier-3> [GlasgowEmbedded/glasgow] esden pushed 3 commits to wip-revC2 [+0/-0/±8] https://git.io/JUfgX
<_whitenotifier-3> [GlasgowEmbedded/glasgow] esden 2bd0a49 - revC2: Switched eeprom packages to TSSOP-8 to gain more space.
<_whitenotifier-3> [GlasgowEmbedded/glasgow] esden c96b88d - revC2: Switched bulk cap footprint, Added system reset button circuit.
<_whitenotifier-3> [GlasgowEmbedded/glasgow] esden 3cd8961 - revC2: Bigger footprint bulk cap and reset button.
<_whitenotifier-3> [glasgow] esden synchronize pull request #196: WIP revC2 - https://git.io/JfXxO
electronic_eel has quit [Ping timeout: 240 seconds]
electronic_eel has joined #glasgow
PyroPeter_ has joined #glasgow
PyroPeter has quit [Ping timeout: 246 seconds]
PyroPeter_ is now known as PyroPeter
d1b23 has joined #glasgow
d1b2 has quit [Read error: Connection reset by peer]
d1b23 is now known as d1b2
<d1b2> <TomKeddie> @Attie I found that out too, no damage though thankfully. https://github.com/GlasgowEmbedded/glasgow/issues/170
<_whitenotifier-3> [GlasgowEmbedded/glasgow] esden pushed 3 commits to wip-revC2 [+2/-1/±3] https://git.io/JUfoQ
<_whitenotifier-3> [GlasgowEmbedded/glasgow] esden c8a5567 - revC2: Moved sync connector to increase separation.
<_whitenotifier-3> [GlasgowEmbedded/glasgow] esden 4768d28 - revC2: Minor silkscreen work. Added tact switch 3d model.
<_whitenotifier-3> [GlasgowEmbedded/glasgow] esden 82df0d9 - revC2: Updated Molex KK 3D Model.
<_whitenotifier-3> [glasgow] esden synchronize pull request #196: WIP revC2 - https://git.io/JfXxO
Exec1N has joined #glasgow
FFY00 has quit [Remote host closed the connection]
FFY00 has joined #glasgow
m4ssi has joined #glasgow
Stephie has quit [Quit: Fuck this shit, I'm out!]
Stephie has joined #glasgow
bvernoux has quit [Ping timeout: 240 seconds]
m4ssi has quit [Remote host closed the connection]
Exec1N has quit [Ping timeout: 264 seconds]
FFY00 has quit [Read error: Connection reset by peer]
FFY00 has joined #glasgow
samlittlewood has joined #glasgow
nmolo has quit [Ping timeout: 240 seconds]
FFY00 has quit [Ping timeout: 240 seconds]
PyroPeter has quit [Ping timeout: 240 seconds]
electronic_eel has quit [Ping timeout: 240 seconds]
samlittlewood has quit [Ping timeout: 240 seconds]
lxdr has quit [*.net *.split]
simukis_ has quit [*.net *.split]
levi has quit [*.net *.split]
Sellerie has quit [*.net *.split]
uovo has quit [*.net *.split]
TiltMeSenpai has quit [*.net *.split]
MadHacker has quit [*.net *.split]
oeuf has quit [*.net *.split]
nmz787 has quit [*.net *.split]
danilonc has quit [*.net *.split]
^^makoto_ has quit [*.net *.split]
smeding has quit [*.net *.split]
Guest39888 has quit [*.net *.split]
noopwafel has quit [*.net *.split]
pie_ has quit [*.net *.split]
Shiz has quit [*.net *.split]
affix has quit [*.net *.split]
d1b2 has quit [*.net *.split]
Stormwind_mobile has quit [*.net *.split]
ali_as has quit [*.net *.split]
bsmt has quit [*.net *.split]
FireFly has quit [*.net *.split]
marcan has quit [*.net *.split]
_whitelogger has joined #glasgow
sven has joined #glasgow
FFY00 has quit [Remote host closed the connection]
sven has quit [Quit: ZNC 1.7.4 - https://znc.in]
sven has joined #glasgow
FFY00 has joined #glasgow
<_whitenotifier-3> [GlasgowEmbedded/glasgow] electroniceel pushed 1 commit to wip-revC2 [+0/-0/±1] https://git.io/JUJtz
<_whitenotifier-3> [GlasgowEmbedded/glasgow] electroniceel d7fe6a3 - revC2: fix distance between port headers to multiple of 2.54 mm
<_whitenotifier-3> [glasgow] electroniceel synchronize pull request #196: WIP revC2 - https://git.io/JfXxO
<whitequark> electronic_eel: hm, i'm somewhat concerned about all the connector moves
<whitequark> they are all so minute as to be within mechanical tolerances of any existing adapters, right?
<electronic_eel> they are minute, so existing addons should still fit
<electronic_eel> but now it should be easier for other users to create new addons
<whitequark> great. and the big connectors are also now on a proper grid. that's both of the concerns
<whitequark> yes. excellent. thanks, i completely missed that
<whitequark> (i was worried about breaking existing addons *and* ending up in an arduino situation)
<electronic_eel> also the dimensions just add up, you can just enter the numbers from the dimension layer into your cad and you should be done, without having to add or subtract, multiply or whatever
<electronic_eel> the distance in the dimension layer is now the distance between pin 1 of connector a and pin 1 of connector b
<whitequark> nice
<electronic_eel> esden wants to finish revC2 very soon and order prototypes. one issue still open is updating the led descriptions. it should be finished for the prototypes because of photos for the campaign
<whitequark> i don't like "APP STATUS" at all
<whitequark> bleh, i'm going to have to review revC2 soon, then
<electronic_eel> so how about "SYS STATUS" instead of "FX2 STATUS" and "USER LED"?
<whitequark> can you screenshot the current state of revC2?
<electronic_eel> the user leds are labeled U1 to U5, so I think "USER LED" makes sense
<electronic_eel> https://imgur.com/3u2px2j
<_whitenotifier-3> [GlasgowEmbedded/glasgow] electroniceel pushed 1 commit to wip-revC2 [+0/-0/±1] https://git.io/JUJqu
<_whitenotifier-3> [GlasgowEmbedded/glasgow] electroniceel 5ab166e - revC2: center the vias to GND for the TVS diode arrays
<_whitenotifier-3> [glasgow] electroniceel synchronize pull request #196: WIP revC2 - https://git.io/JfXxO
<whitequark> USER LED is fine with me
<whitequark> maybe even USER LEDS
<electronic_eel> or "USER LEDs"?
<whitequark> while you're at it, ESD Prot. should be all in uppercase, or it's inconsistent with all other functional block labels
<electronic_eel> yeah, right
<whitequark> also: damn that's a lot of tiny text
<whitequark> i expect every fab that isn't expensive as heck to make a mess out of it
<electronic_eel> esden has designed a nice arrow symbol for pullups/pulldowns for one of his pmods
<electronic_eel> he should try to use it instead of the text
<electronic_eel> but I guess he didn't have time to finish that part yet
<electronic_eel> re text size: revC1 has a lot of even smaller text, all the component labels
<whitequark> point
<electronic_eel> and it worked out quite well for the one that greg made for me or the photos I have seen from esden
<electronic_eel> so I wouldn't worry too much about it
<whitequark> tbh i just find the text noisy
<whitequark> lots of redundant labels too, like the pull ones
<electronic_eel> right, it is between the ics and right of them, this is redundant
<whitequark> i don't like how verbose it is in general, though i am not opposed to it because i assume esden has some good reason to go through this effort that i might not understand
<electronic_eel> I think the idea is to give the user an overview about which section does what. also it works a bit as a feature list for the people that prefer to look at the board instead of text
<whitequark> does anyone actually do that?
<whitequark> the former i mean
<whitequark> hm
<electronic_eel> look at for example at what hackaday posted recently. it was the picture of the board.
<whitequark> does this mean we don't actually need the labels to be on the actual physical PCBs? we just need them on the screenshots etc?
<whitequark> right
<whitequark> i'm still not sure if the conclusion makes sense
<whitequark> like, end users don't specifically care that the board has an ADC and a DAC
<electronic_eel> does it hurt adding them?
<whitequark> in fact
<whitequark> i think we should actually remove those, because they are misleading
<electronic_eel> someone familiar with the tech might care about how it is built and goes together internally
<whitequark> there's no actual analog functionality
<whitequark> hm
<whitequark> but the board doesn't even give a good overview of that
<d1b2> <TiltMeSenpai> level sense and level set?
<whitequark> what actually matters is the specific selection of shifters, the way we use the FX2 and the FPGA, and so on
<whitequark> because that's what determines the performance (or lack thereof) for the tasks the board will be used
<electronic_eel> hmm, the level shifter section is labeled as such and you can easily see that there is one shifter per in/out pin. I think this is what matters
<whitequark> what matters is that there is direction control
<whitequark> i would *gladly* use ganged level shifters with direction control if they existed
<whitequark> like, yes, you're technically pointing to the fact that we use level shifters with direction control, indirectly, by showing that there's one per pin, and relying on the external knowledge that ganged shifters with individual OE don't exist
<electronic_eel> exactly
<whitequark> this doesn't communicate the fact to anyone who didn't already spent a lot of hours building the same board.
<whitequark> and the five people who know this offhand will probably just look at schematic first, not the board
<electronic_eel> I don't know, I wouldn't underestimate the people interested in a board like this
<whitequark> i mean
<electronic_eel> and if someone doesn't get this level of detail from the silkscreeen, it also doesn't really matter. because silkscreen does not cost extra
<d1b2> <TiltMeSenpai> are there plans for a faq
<d1b2> <TiltMeSenpai> could throw it there
<whitequark> i feel like the intersection of "can pick up this implication just looking at the silk" and "studies silk instead of looking at schematic" is an empty set
<electronic_eel> TiltMeSenpai: esden has a nice overview graph on the campaign page. a faq could also be added there
<electronic_eel> whitequark: I think it depends on what you see first. if you see a picture of the pcb posted on some news site, you might look at that first. if it triggers your interests, you go and look up the github and search for the schematics
<whitequark> I guess
<whitequark> like I said, mostly I don't like the added noise
<whitequark> but
<whitequark> maybe I'll change my mind after I see revC2 in person
<whitequark> it's not like we can't remove it later
<electronic_eel> we started out with each and every resistor labeled on silk, so the noise is already reduced significantly
<whitequark> hm
<whitequark> you know what, you're objectively right
<whitequark> i'm not sure what actually bothers me about it
<whitequark> maybe just seems unfamiliar? ignore me, then
<electronic_eel> https://imgur.com/mAy5IWr
<electronic_eel> this is my proposed led silk
<electronic_eel> I also cleaned up the pull descriptions and esd protection
<electronic_eel> ok?
<noopwafel> I think the border boxes make it look quite a bit noisier to me
<electronic_eel> noopwafel: you mean around each section, or just around the leds?
<noopwafel> around each section
<noopwafel> (I am kind of used to the revC1 design so I'm not sure this is a useful opinion, likely biased :P)
<electronic_eel> hmm, but a description without showing what it is for, is a bit meaningless
<noopwafel> I adore the little 10k\ohm and ESD PROT type things thoug
<whitequark> i agree with noopwafel actually
<whitequark> in the sense that i think border boxes are what bothers me about ti
<whitequark> i'm not necessarily saying that they should be removed
<electronic_eel> hmm, strange, when I look at them I get the feeling that they give a structure to the board and I like that
<electronic_eel> like dividing a complex thing into smaller parts
Getorix_ has joined #glasgow
Getorix has quit [Ping timeout: 260 seconds]
<electronic_eel> whitequark: should I commit the silkscreen changes as in the screenshot above or should we postpone deciding this?
<whitequark> electronic_eel: sorry, i missed that message. looking now
<whitequark> shouldn't the pull label also be uppercase? like if we fixed esd prot might as well fix that
<whitequark> or would that use the symbols later?
bvernoux has joined #glasgow
<electronic_eel> https://imgur.com/GWGUYQh
<electronic_eel> I made them uppercase
<_whitenotifier-3> [glasgow] kasbah opened pull request #200: Add kitspace.yml to revC1 branch - https://git.io/JUJYV
<electronic_eel> but actually I prefer the upper / lowercase combo we had before, because it makes the very much shortened words easier to discern
<_whitenotifier-3> [glasgow] kasbah commented on pull request #199: Add "eda" field to kitspace.yml - https://git.io/JUJYw
<_whitenotifier-3> [glasgow] kasbah closed pull request #199: Add "eda" field to kitspace.yml - https://git.io/JUvmL
<whitequark> "PULLS CTRL" ?
<_whitenotifier-3> [glasgow] Error. The Travis CI build could not complete due to an error - https://travis-ci.org/github/GlasgowEmbedded/glasgow/builds/720777836?utm_source=github_status&utm_medium=notification
<_whitenotifier-3> [glasgow] whitequark closed pull request #200: Add kitspace.yml to revC1 branch - https://git.io/JUJYV
<_whitenotifier-3> [GlasgowEmbedded/glasgow] whitequark pushed 1 commit to revC1 [+1/-0/±0] https://git.io/JUJY6
<_whitenotifier-3> [GlasgowEmbedded/glasgow] kasbah d7e5a87 - Add kitspace.yml to revC1 branch
<electronic_eel> do you immediately associate "pulls" with pullup / pulldown? I think the "PUp/PDwn" is better to trigger that association
<d1b2> <Attie> PUP / PDN / HIZ ?
<_whitenotifier-3> [glasgow] Error. The Travis CI build could not complete due to an error - https://travis-ci.org/github/GlasgowEmbedded/glasgow/builds/720778540?utm_source=github_status&utm_medium=notification
<d1b2> <esden> it is not HIZ though
<d1b2> <esden> it is passive
<d1b2> <esden> we already discussed this
<d1b2> <Attie> ah, fair... could be active drive
<electronic_eel> esden: hi
<d1b2> <Attie> apologies
<d1b2> <esden> hey you are going ham on the board I see @electronic_eel 😉
<d1b2> <esden> Thanks for fixing the position of the connectors
<whitequark> electronic_eel: i can't really parse PUp/PDwn/NoP at all
m4ssi has joined #glasgow
<d1b2> <esden> I actually find it pretty clear to be honest @whitequark
<d1b2> <esden> PUp does divide into P-Up
<d1b2> <esden> so ...
<electronic_eel> whitequark: but does "PULLS" really work better for you?
<whitequark> electronic_eel: i wouldn't suggest it otherwise
<d1b2> <esden> It probably depends on the person reading it
<whitequark> maybe just "PULL"
<whitequark> "pull" doesn't have any other meaning in electronics, does it?
<d1b2> <esden> Fair
<d1b2> <esden> I would say PULLS as it describes the section
<electronic_eel> maybe we leave out the "ctrl" and just write PULL UP/DWN?
<d1b2> <esden> an that section implements the PULLS
<whitequark> PULL UP/DN also works for me
<d1b2> <esden> IO-PULLS
<whitequark> IO-PULLS also works for me
<d1b2> <esden> Yeah we can skip the ctrl I think.
<d1b2> <esden> but if we have space the full version could be "IO PULL Ctrl"?
<electronic_eel> https://imgur.com/feMkYG9
<electronic_eel> how about this? we have space for it all
<d1b2> <esden> That is fine with me.
<d1b2> <Attie> +1
<whitequark> electronic_eel: totally fine with it
<_whitenotifier-3> [GlasgowEmbedded/glasgow] electroniceel pushed 1 commit to wip-revC2 [+0/-0/±1] https://git.io/JUJOe
<_whitenotifier-3> [GlasgowEmbedded/glasgow] electroniceel cd546c8 - revC2: improve silkscreen labes for LEDs, pulls and ESD protection
<_whitenotifier-3> [glasgow] electroniceel synchronize pull request #196: WIP revC2 - https://git.io/JfXxO
<d1b2> <esden> I think we might have enough space to write "ESD PROTECT"... might be better than "PROT"?
<d1b2> <esden> But I am fine with the way it is now too.
<electronic_eel> there are some vias in the way I think. we'd need to align it left on top and right on bottom which would make it asymmetric
<d1b2> <esden> Ahh ok that makes sense.
<d1b2> <esden> Then leave it the way it is.
<electronic_eel> esden: what do you think about moving the polymer cap left + down a bit? to allow reworking the fx2
m4ssi has quit [Remote host closed the connection]
<d1b2> <esden> Yeah I agree with your comment there. We should move it.
<electronic_eel> do you want to take over or should I do it?
<d1b2> <esden> If you are on the roll already go at it. 😄
<electronic_eel> ok
<d1b2> <esden> I will deal with it later today if you dont. 😄
<electronic_eel> I have about 2 hours left till I go to sleep, you can take over then
<electronic_eel> round-the-clock layout
<noopwafel> I had to desolder that cap once to make space for fixes, so .. please
<electronic_eel> yeah, it has a nice solid connection to the ground layer, it is not designed to come off easily...
<whitequark> caps on power rails don't come off easily in general
<whitequark> esp smd caps
<noopwafel> I just destroyed it, was quite a pain even then
<noopwafel> learnt lesson, add that cap *after* reflow :)
<whitequark> i mean there's probably very little we can do about it
<electronic_eel> oh, tht caps can be a pain on a multilayer board too. especially opening the holes afterwards again so you can solder in a new one
<whitequark> i've never been able to cleanly rework these even if there's lots of spare spaces
<whitequark> electronic_eel: that's true but if the cap is small enough you don't *really* need that
<whitequark> you can just heat both leads at once until it starts sliding in, then switch sides
<whitequark> you can *sorta* apply the same principle to smd components but not to electrolytic caps
<d1b2> <esden> With the previous cap I was able to get to the FX2 pins without too much trouble, without removing the cap.
<electronic_eel> hmm, last time I tried it that didn't really work well. I used a small syringe needle (stainless) instead to open up the hole
<d1b2> <esden> So if we can get at least similar separation of the new cap to the FX2 as before we will be fine.
<d1b2> <esden> But I do have a very fine soldering iron, not everyone has access to that.
<whitequark> electronic_eel: oh that is a clever idea
<electronic_eel> if there are just one or two shorts on the fx pads, it should be enough to put in a bit of flux and do a swipe with a small tip iron
<whitequark> i've did something similar with cut off resistor leads and such
<whitequark> but a needle would work way better
<electronic_eel> the resistor leads would solder in, the stainless doesn't take solder, so you can still pull it out when the solder hardens
<d1b2> <esden> @electronic_eel that is exactly what I do when I need to fix things on QFN packages. Takes care of %90 of the cases. The rest requires removing the part completely.
<d1b2> <esden> Great tip regarding the through hole parts! I have to remember that for the future. 🙂
<whitequark> electronic_eel: i pull it out while it's hot
<electronic_eel> if it is an 6 or 8 layer board, you need good preheating for that to work. the syringe needle method often works without
<electronic_eel> but you are the blowtorch witch, so no problem for you ;)
bvernoux has quit [Quit: Leaving]
<whitequark> electronic_eel: it is actually a problem
<whitequark> i *can* use a blowtorch, and i like it *more* than tools that don't work, but i don't like it all that much
<whitequark> it's way too easy to set things on fire
<whitequark> it's not a problem because of the fire (ul94v0 takes care of that, i think) but it is a problem because i'm probably trying to not destroy something.
<whitequark> so if i do it on a PCB i actually care about, rather than just desoldering stuff, it's kinda stressful
<electronic_eel> you can easily desolder stuff that is hard to solder on again, like bgas
<electronic_eel> I had to fix an old pc mainboard with bad capacitors once. I wanted to get this exact model to work again and had just this one...
<whitequark> tbh i usually have the inverse problem
<whitequark> like i am trying to desolder a bga but i also move a few passives around
<whitequark> the bga, i can easily get back there. worst case, i just contract it out to the nearest unauthorized cellphone repair shop
<whitequark> they'll do a far better job than i could ever try
<whitequark> the passives though not so much
FFY00 has quit [Remote host closed the connection]
FFY00 has joined #glasgow
<_whitenotifier-3> [GlasgowEmbedded/glasgow] electroniceel pushed 1 commit to wip-revC2 [+0/-0/±1] https://git.io/JUJsv
<_whitenotifier-3> [GlasgowEmbedded/glasgow] electroniceel 5a711e0 - revC2: move polymer capacitor to allow reworking the pads of the FX2
<_whitenotifier-3> [glasgow] electroniceel synchronize pull request #196: WIP revC2 - https://git.io/JfXxO
<electronic_eel> sorry for bringing this up again, but today the led labels haunt me
<electronic_eel> we now have "SYS STATUS" and "USER LEDS"
<whitequark> mh true
<electronic_eel> the "SYS" in "SYS STATUS" was meant to differentiate agains "FPGA STATUS" (which we previously had instead USER LEDS)
<electronic_eel> why the "SYS"?
<electronic_eel> I'd prefer just "STATUS" now
<electronic_eel> ok?
<whitequark> dropping SYS seems fine
<whitequark> but the inconsistency between "STATUS" (no "LEDS") and "USER LEDS" bothers me
<electronic_eel> "STATUS LEDS" and "USER LEDS" - or "STATUS" and "USER"?
<whitequark> honestly both seem fine to me, maybe a mild preference for the former
<_whitenotifier-3> [GlasgowEmbedded/glasgow] electroniceel pushed 1 commit to wip-revC2 [+0/-0/±1] https://git.io/JUJsq
<_whitenotifier-3> [GlasgowEmbedded/glasgow] electroniceel 8affcad - revC2: rename led blocks once again, to "STATUS LEDS" and "USER LEDS"
<_whitenotifier-3> [glasgow] electroniceel synchronize pull request #196: WIP revC2 - https://git.io/JfXxO
<_whitenotifier-3> [GlasgowEmbedded/glasgow] electroniceel pushed 1 commit to wip-revC2 [+0/-0/±1] https://git.io/JUJsO
<_whitenotifier-3> [GlasgowEmbedded/glasgow] electroniceel 1dd1fb5 - revC2: adapt text about recovery in schematics to match current silkscreen + reset button
<_whitenotifier-3> [glasgow] electroniceel synchronize pull request #196: WIP revC2 - https://git.io/JfXxO
<_whitenotifier-3> [GlasgowEmbedded/glasgow] electroniceel pushed 1 commit to wip-revC2 [+0/-0/±1] https://git.io/JUJs6
<_whitenotifier-3> [GlasgowEmbedded/glasgow] electroniceel 5efc2a9 - revC2: add bom keys and datasheet for button + reset ic
<_whitenotifier-3> [glasgow] electroniceel synchronize pull request #196: WIP revC2 - https://git.io/JfXxO
<electronic_eel> esden: ok, I'm done for today
<electronic_eel> the ib2-bom-key for the button is missing, probably others aren't correct yet. The ones for the 4r7 resistors might be wrong.
<whitequark> 1b2-bom-key?
<whitequark> homoglyphs strike again
<electronic_eel> oh yeah 1b2
<electronic_eel> 1b2 is really hard for me, as the internal domain at work is i2n
<electronic_eel> i always mix that up
<d1b2> <esden> @electronic_eel thank you very much! 🙂 I will deal with the keys.
<electronic_eel> one thing that came to my mind was if you want to try pin-in-paste with the usb connector
<electronic_eel> like make a larger opening for the through hole parts in the stencil
<d1b2> <esden> I already did pin-in-paste with this connector on another board.
<electronic_eel> and, how did it work? was there enough solder on the tht parts?
<d1b2> <esden> All it needed is opening the stencil up for the pin pads. There is some additional tweaking that can be done, where the opening is even bigger to provide more solder to the pin. But it worked ok with the amount it got already.
<electronic_eel> so why not try to enlarge the opening a bit?
<electronic_eel> or do you want to do that at a later stage?
<d1b2> <esden> Because that is dependent on the assembler, and what stencil thickness they will use at the end.
<d1b2> <esden> But yeah we can probably enlarge it now already.
<d1b2> <esden> Whoever assembles will have to tweak it anyways.
Getorix has joined #glasgow
Getorix_ has quit [Ping timeout: 240 seconds]