<d1b2>
<timonsku> making it deeper will be icky, lemme check
<agg>
perhaps memory on lvds connector all inside case wasn't meant to be :p
<electronic_eel>
ok, this last picture is just the top part of the case and there already is a pocket in there necessary to make space for the pins of the female lvds connector on the glasgow pcb?
<agg>
pins of the female connector?
<electronic_eel>
oh, right, male connector i meant
<electronic_eel>
sorry
<d1b2>
<timonsku> there is no space for any mating connector to be left in place no
<d1b2>
<timonsku> if that is a needed feature than it would be cutout but that was deciding against previously
<electronic_eel>
ok, then i misjudged the height we had available
<electronic_eel>
so if we do a hyperram addon then it has to be used without the top part of the case
<electronic_eel>
timonsku: thanks for checking though
<d1b2>
<timonsku> how much area would you need though for such an addon?
<d1b2>
<timonsku> I just had to bump the height a tiny bit, I could try to free up some more space
<d1b2>
<icb> Samtec lists the mated stack height of the two connectors as 5.82mm, plus board thickness, plus Cypress lists the height of the HyperRAM chips including balls as 1mm, so that's what, 8.52mm above board height?
<d1b2>
<timonsku> meant more in 2D space 🙂 I added a 15.8x26mm pocket now
<d1b2>
<timonsku> you could mount components on the bottom of such an addon to not increase height further
<d1b2>
<esden> if you add a thin carrier pcb with the hyperram chips mounted facing the Glasgow PCB you should be able to get it in 6mm above glasgow
<d1b2>
<esden> (you will be able to get the mating connectors for your own addons from us, as they have a several k moq)
<d1b2>
<esden> (but they will be cheaper than samtec parts)
<sorear>
0.5 mm substrate?
<d1b2>
<esden> ok pcbway can make 0.4mm thick 4 layer PCBs ... but you can make a hyperram breakout with 2layers too... the earlier pmods were just 2 layer.
<d1b2>
<icb> What's the force required for (un)mating? I'd be worried about flexing with a board that thin
<d1b2>
<esden> this will be very small... don't yank on the board but rather pull on the connector instead
<d1b2>
<icb> I mean, sure, it's probably something you put on and never take off if it fits in the case
<d1b2>
<esden> it is a very long connector so in any case it will be bad to just yank on the pcb
<d1b2>
<esden> @icb indeed, definitely less of a problem than if it was some interface used to connect to targets... where you would have to cycle it all the time
<awygle>
marcan / esden: when you tuned the LEDs, what mcd value did you end up with?
<d1b2>
<timonsku> ah ok with 5.5mm stacking I can't make enough depth without making the top uncomfortably thin
<d1b2>
<timonsku> I could squeeze it to 6mm total room but not sure that would actually work with tolerances and all 🙂
<d1b2>
<icb> PCBWay will do a 0.2mm 2 layer
<d1b2>
<timonsku> lol
<d1b2>
<timonsku> at that point you basically got flex pcb
pg12_ has joined #glasgow
pg12 has quit [Ping timeout: 260 seconds]
<d1b2>
<icb> The top of the case is flush with the top of the shroud on the A/B sockets, right? So you wouldn't want to go any higher
Foxyloxy has quit [Ping timeout: 265 seconds]
<d1b2>
<timonsku> it was and now is +0.5mm but yea even higher is a bit meh, it would be possible though given that its releasable without actuating that latch
<d1b2>
<timonsku> if its a really important feature to have it could be considered
FFY00 has quit [Remote host closed the connection]
FFY00 has joined #glasgow
electronic_eel has quit [Ping timeout: 256 seconds]
electronic_eel has joined #glasgow
PyroPeter_ has joined #glasgow
PyroPeter has quit [Ping timeout: 272 seconds]
PyroPeter_ is now known as PyroPeter
Foxyloxy has joined #glasgow
<d1b2>
<brainstorm> Back to this, I'm getting the following (not breakpointing yet, I just wasn't seeing those errors in the editor output for some reason): https://pastebin.com/wFK52NYB .... I might switch to an OpenOCD-specific channel now, sorry to bother & thanks for all the help 😉
GNUmoon has quit [Remote host closed the connection]
GNUmoon has joined #glasgow
Stephie has quit [Read error: Connection reset by peer]
Stephie has joined #glasgow
GNUmoon2 has joined #glasgow
GNUmoon has quit [Remote host closed the connection]
DarthCloud has quit [Ping timeout: 240 seconds]
DarthCloud has joined #glasgow
thasti has quit [Remote host closed the connection]
<d1b2>
<icb> For a card on the LVDS port, fitting in the case, how about a socket header like this? https://www.antenk.com/1-27mm-female-pin-header/56952949.html It allows the pins to pass through it, and is very thin itself. I think using a standard 1.6mm PCB, the top side would only be above the top of the pins by 0.75mm (3.05mm pin height, 2.2mm thick socket, 1.6mm thick PCB)
<d1b2>
<Attie> @brainstorm ... please note that none of the discord features are mirrored to IRC... replies (in this case), reactions, edits, etc
<d1b2>
<Attie> best to reintroduce your topic / refresh memories quickly, so people don't have to dig in scroll back (or not know what you're referring to)
<d1b2>
<Attie> (actually edits in this case too!)
<_whitenotifier>
[glasgow] electroniceel commented on issue #221: Qualification tests for revC2: post INA233 - https://git.io/JLjlE
<_whitenotifier>
[glasgow] electroniceel closed issue #221: Qualification tests for revC2: post INA233 - https://git.io/JTw0R
<_whitenotifier>
[glasgow] electroniceel commented on pull request #242: Bring revC2 firmware to the same level as revC1 - https://git.io/JLjl2
<_whitenotifier>
[GlasgowEmbedded/glasgow] electroniceel created branch wip-revC3-hardware https://git.io/JLjl6
<_whitenotifier>
[glasgow] electroniceel created branch wip-revC3-hardware - https://git.io/fhhGp
thasti has joined #glasgow
<_whitenotifier>
[GlasgowEmbedded/glasgow] electroniceel pushed 1 commit to wip-revC3-hardware [+0/-0/±1] https://git.io/JLjlF
<_whitenotifier>
[GlasgowEmbedded/glasgow] electroniceel 9de3778 - revC3: change silkscreen to "owGglas revC?" to emphasise that it is a development version for now
<_whitenotifier>
[GlasgowEmbedded/glasgow] electroniceel pushed 1 commit to wip-revC3-hardware [+0/-0/±1] https://git.io/JLj8v
<_whitenotifier>
[GlasgowEmbedded/glasgow] electroniceel 1fe4b7a - revC3: remove port pinout marking on top silkscreen
<_whitenotifier>
[GlasgowEmbedded/glasgow] electroniceel pushed 1 commit to wip-revC3-hardware [+0/-0/±1] https://git.io/JLj83
<_whitenotifier>
[GlasgowEmbedded/glasgow] electroniceel 9789446 - revC3: fix internal i2c bus testpoint silkscreen labels on top side
<_whitenotifier>
[glasgow] electroniceel commented on issue #250: Silkscreen labels swapped on SCL and SDA testpoints - https://git.io/JLj8G
<_whitenotifier>
[glasgow] electroniceel closed issue #250: Silkscreen labels swapped on SCL and SDA testpoints - https://git.io/JLyec
GNUmoon2 has quit [Remote host closed the connection]
FFY00 has quit [Read error: Connection reset by peer]
FFY00 has joined #glasgow
nicoo has quit [Write error: Connection reset by peer]
DarthCloud has quit [Remote host closed the connection]
DarthCloud has joined #glasgow
nicoo has joined #glasgow
<_whitenotifier>
[GlasgowEmbedded/glasgow] electroniceel pushed 1 commit to wip-revC3-hardware [+0/-0/±5] https://git.io/JLj0d
<_whitenotifier>
[GlasgowEmbedded/glasgow] electroniceel 76ce8c8 - revC3: change the I2C address of DAC A to remove the address clash with the SMBus Alert Response Request (ARA)
<_whitenotifier>
[GlasgowEmbedded/glasgow] electroniceel pushed 1 commit to wip-revC3-hardware [+0/-0/±1] https://git.io/JLjuR
<_whitenotifier>
[GlasgowEmbedded/glasgow] electroniceel 94a49b0 - revC3: increase spacing of the 3V3 flood fill on the internal power layer around the mounting holes
<_whitenotifier>
[glasgow] electroniceel commented on issue #255: Schematics & Layout TODO for revC3 - https://git.io/JLjzA
<electronic_eel>
@icb I also thought about pass-through connectors like the one you posted. But I fear we still don't win enough height with it:
<electronic_eel>
I just measured the connector esden uses for manufacturing. the base part ends at 1.6 mm above the pcb, that is where the pins begin to stick out
<electronic_eel>
your connector then has 2.2 mm height above that, then we have 0.6 mm for a very slim pcb and 0.2 mm tolerance. then we end at 4.6 mm above pcb
<electronic_eel>
the pins of the connector esden used also end at 4.6 mm above pcb. but to accommodate them timonsku already needs milled pockets in the case
<electronic_eel>
maybe @timonsku can check again and sees a way to make it possible, but i fear it is still too tight with this kind of connector.
<electronic_eel>
and changing to another connector type this late in design will not win us esdens favor...
<d1b2>
<icb> I thought @timonsku said they could/would expand the pocket to be able to fit a board
<electronic_eel>
ok, lets see what he comes up with
<d1b2>
<icb> I agree that it's too late to entertain changing the header, even to a footprint compatible one
<d1b2>
<icb> It would be nice to fit under the case, but not the end of the world if it can't be done without making too many sacrifices
<d1b2>
<timonsku> the pocket itself is 15.8x27mm, this includes the space the connector itself occupies
<d1b2>
<timonsku> from PCB top to pocket top is currently a height of 5.85mm
<electronic_eel>
ok, so the pocket would now allow for a small pcb
<d1b2>
<timonsku> yes, that should fit at least like 2 or more of these hyperrams ICs if you bottom mount them.
<electronic_eel>
yes, they are 6x8mm, so they should fit. they'd just need some small passives, but nothing more
<electronic_eel>
cool, so a completely encased hyperram addon is possible
<electronic_eel>
now we just need to design it and write firmware for it...
<d1b2>
<timonsku> ah so that height is enough with that pass through style connector?
<electronic_eel>
do you see my calculation in the scrollback? the pcb ends at 4.6 mm above glasgow pcb
<d1b2>
<timonsku> yea but hard to follow were that now ended up 🙂 but cool then I leave it at that, then you could probably even do double sided mounting for more of them
<electronic_eel>
one could also go really crazy and put a thin downward facing hd connector on there and then fill up the whole space the glasgow pcb leaves with stuff...
<d1b2>
<icb> Is there enough space on all sides of the pins for the socket?
<sorear>
how is firmware involved?
<electronic_eel>
sorear: hmpf, i meant gateware
<electronic_eel>
icb: good question, your type of connector would need a bit of space on the side facing to the outside of the pcb, you'd need a via there to route the tracks
<electronic_eel>
@timonsku sorry to bother you again, but could you measure the space between the center of the lvds connector and the outward facing wall of the pocket?
<electronic_eel>
when i see in the picture above that the wall gets thicker there (in contrast to the wall thickness at the port connector shroud) it could get tight
<electronic_eel>
or do you have step files of your case and the glasgow pcb inside? then i could measure this stuff on my own
<d1b2>
<timonsku> center of connector to outside wall is 3.3mm
<electronic_eel>
what kind of tolerances are you comfortable with such a case?
<electronic_eel>
i think a pcb would need 3.1mm there, but without any tolerances for case/glasgow fit
<electronic_eel>
are the 0.2 remaining mm enough for tolerance?
<d1b2>
<timonsku> 0.2mm should be fine but I can certainly budge that a bit more, doesn't need to be as thick there
<electronic_eel>
ok, cool
<electronic_eel>
now we just need to figure out what to do with the stupid leds...
<d1b2>
<timonsku> its now ~4.8mm
<electronic_eel>
now that should be enough, thanks
<d1b2>
<icb> What about the other dimension? IIRC, the connector is pretty close to the mounting screw and LEDs
<electronic_eel>
the mating connector would need about the same space as the one on the glasgow pcb
<electronic_eel>
since the one on the glasgow pcb is accounted for, i think this should be fine
<d1b2>
<icb> Ok. If that dimension doesn't worry you then I withdraw my question
<electronic_eel>
if it won't fit, we could always add a small surface mount black hole that dilates the space continuum a bit ;)
jstein has joined #glasgow
<d1b2>
<icb> I'm also making the assumption that socket, or an equivalent, is buyable at a reasonable price, in a reasonable quantity
<electronic_eel>
i guess esden would be able to source these or similar ones at reasonable prices
<electronic_eel>
he'd have to order like 1000pcs or similar though. but if the addon adds significant features then i don't see an issue with that
<gruetzkopf>
i much prefer surface mount spacetime distortions without a singularity at the center
<electronic_eel>
yeah, but the singularity is easier to place in the PnP
<d1b2>
<brainstorm> In the cmdline I'm not defining it at all, I guess it's fine like that?: $ glasgow run jtag-openocd --port A --pin-tck 0 --pin-tms 3 --pin-tdi 2 --pin-tdo 1 -V 3.3 -f100 tcp:localhost:5555
GNUmoon2 has joined #glasgow
<whitequark>
if you do not have a TRST# pin this is OK and does not impact ~any functionality