Saturday, 2014-06-21

*** tpb has joined #timvideos00:00
mithroayush3504: ping?00:24
ayush3504mitho: yes00:24
mithroayush3504: where are we at? Do you need me to do anything at the moment?00:25
ayush3504mithro: ya, we need to figure out connectors for IR daughterboard00:25
mithroayush3504: is the schematic started yet?00:25
ayush3504mithro: and a few case specific dimensions00:26
ayush3504mithro: not much, for the schematic part, spent time looking for microcontrollers00:26
ayush3504mithro: also, the parallel port behaviour. it has 8 I/O pins, what to do with rest of 25-8 pins?00:27
mithroayush3504: we have 8 I/O pins per daughterboard right? If it is spanning two daughterboards that would give you 16?00:27
ayush3504mithro: got it00:28
mithrohow many pins does that leave?00:28
ayush3504mithro: 9?00:29
mithroayush3504: btw, how are you updating the images in the spreadsheet when you do a change to the drawing?00:32
ayush3504mithro: save to png then insert. is there a direct way?00:33
mithroayush3504: kinda of, you set the drawing to "start publishing" then you "replace image" and give it the URL00:33
ayush3504mithro: ah.. thx00:34
mithroIE -->
ayush3504mithro: yeah. got it. thx :)00:35
mithrooneday I'll write a Google AppScript / Macro to do it automatically00:35
mithroayush3504: okay the spreadsheet is looking much better00:36
mithroayush3504: it's surprising what some pictures will do00:36
ayush3504mithro: thanks. You mean the diagrams pasted?00:36
mithroayush3504: you should take a copy of all the Engineering diagrams and save them in a Google Docs folder00:37
mithroayush3504: so the links don't break00:37
mithroayush3504: the pictures + eng diagrams in the first sheet, the diagrams on the second+third sheet00:37
ayush3504mithro: yeah, i have saved them in the local repo. will update the links00:37
mithroayush3504: great!00:38
mithroayush3504: you should make the height/width/depth machine readable00:38
mithro(IE split it into two columns, the width and then error)00:38
ayush3504mithro: also, take a look a the column J of Daughterboard sheet.00:38
ayush3504mithro: machine readable?00:39
mithroayush3504: so you can do "=max(C:C)" in the spreadsheet00:39
ayush3504mithro: yes, but for what exactly?00:40
mithroayush3504: so you can do "=max(C:C)" so you can show the maximum width of the connectors, maximum depth, etc without having to update it manually00:40
ayush3504mithro: ok00:41
ayush3504mithro: the connector picture is already there in column J links01:10
ayush3504mithro: sry col L not J01:10
mithroayush3504: ?01:11
tpbTitle: RS-232 DB9 daughterboard - Google Zeichnungen (at
ayush3504mithro: column L has these drawings for all daughterboards01:12
mithroayush3504: take a look at -> under the "Inserting an image inside of a cell"01:13
tpbTitle: Insert images into a spreadsheet - Docs editors Help (at
ayush3504mithro: ya, but i wanted to link to full-sized drawing01:14
ayush3504mithro: and for the side view, there's a panel cutout column (not filled yet), which will have drawings with side-view01:15
mithrowhat does the "<" in pin out mean?01:16
ayush3504mithro: it means it's there in the link it's pointing to01:17
mithroayush3504: if you convert all those images to a =images("") the whole spreadsheet won't break when I try to sort it01:22
ayush3504mithro: ok01:23
mithroayush3504: I also changed the one in chassis specification so now it should auto-update!01:23
mithroI think this is a new feature, I haven't seen it before01:23
ayush3504mithro: auto update with =image() ?01:23
mithroayush3504: the =image() full stop01:24
ayush3504mithro: ok01:24
ayush3504mithro: in the connector footprint placement drawings I need the value for "g"01:25
mithroayush3504: I think "g" depends on your connectors, no?01:26
ayush3504mithro: no it's independent, it depends on how far would you like the pcb to be from the front plate01:27
mithroand maybe the Atlys board01:28
mithroayush3504: how far you want the PCB to be away from the front panel depends on how far the connectors overhang the PCB01:29
ayush3504mithro: how far the connectors should overhang would depend on the desired gap :p01:30
ayush3504mithro: from my side it's controllable01:30
mithroayush3504: well, the Atlys board already has some connectors which overhang on the common edge you share right?01:31
ayush3504mithro: yes, but they expect a cutout on the faceplate to fit in01:31
ayush3504mithro: the the gap could be almost 0 if you want to. but 0 is difficult to achieve, we can keep something like 1 mm which is controllable.01:33
mithroayush3504: so the front face plate is going to be shared between the Atlys board and your serial board, right?01:37
ayush3504mithro: yes01:37
mithroayush3504: so if the two front edges of the Atlys PCB and your PCB line up, what does that mean for "g" ?01:38
ayush3504mithro: the gap between front face plate and the shared front edge of the two pcbs01:39
mithroayush3504: then it is dictated via the connectors already on the Atlys right?01:39
ayush3504mithro: it shouldn't be 0 because the PCB edges may not be perfectly smooth, and the daughter board could be *very* slightly tilted.01:39
ayush3504mithro: not dictated by Atlys, if you allow the overhanging connectors to pass through a cutout.01:40
mithroayush3504: we do have the option of setting your PCB back a little from the Atlys PCB edge thus giving a different "g" for your section of the board the Atlys board01:40
mithroayush3504: if we choose a g of 20mm, then none of the Atlys connectors are going to be externally visible01:41
ayush3504mithro: you mean you wish to hide the connectors on Atlys?01:42
mithroayush3504: no, the opposite01:42
ayush3504mithro: so have a g like 1 mm?01:42
ayush3504mithro:  you're going to have cutouts for the atlys board connectors right?01:43
mithroayush3504: what exactly do you mean by "cutouts"01:43
ayush3504mithro: a hole in the faceplate for exposing a connector. For overhanging connectors it's a hole in then faceplate that allow connectors to overhang through the face plate01:44
ayush3504mithro: like the motherboard shields, the PS/2 connectors lie inside just exposed. and the Audio jacks overhang through the shield.01:45
mithroayush3504: yes, all the connectors on the Atlys board will be exposed via cut-outs01:47
mithroayush3504: the case front plate will most likely be made out of either 3mm or 5mm thick acrylic / plastic01:47
ayush3504mithro: ok. how would g depend on the connectors if you allow overhanging connectors to pass through the holes?01:48
mithroayush3504: because, ideally the connector edges for things like the HDMI connectors would be flush with the faceplate front01:50
ayush3504mithro: got it01:51
mithroayush3504: IIRC for the Atlys board, we need to mill an area into the 5mm acrylic to make that work01:53
ayush3504mithro: 5 mm acrylic for which side?01:54
ayush3504mithro: front plate?01:54
mithroayush3504: yeah01:54
ayush3504mithro: but that's too thick. can't we have 2 or 3 mm?01:54
mithroayush3504: 3mm is an option01:55
mithro3mm and 5mm are the most common thickness of acrylic you find in hacker spaces01:55
ayush3504mithro: and 2 mm?01:56
ayush3504mithro: umm okay01:56
mithroand generally have tolerances of +- 5% (ISO7823-2) - see the addition in the Chassie sheet01:56
ayush3504mithro: 2 mm thickness would expose the connectors just right01:57
mithroayush3504: we want people to be able to laser cut / mill cases in there local hacker space without needing to order exotic stuff :)01:57
mithroayush3504: so it sounds like g is actually dependent on the thickness of the material uses for the front panel?01:57
ayush3504mithro: ah ok :)01:57
ayush3504mithro: g has to be tending to 001:58
mithroayush3504: so, if for something like 5mm it was 0, for 3mm it would be 2mm?02:00
mithroayush3504: or is 5mm just not doable?02:00
ayush3504mithro: nope i assumed you were going to use metal plate, not 5 mm acrylic :P02:01
mithroayush3504: metal plate is much harder to cut in a hackerspace :P02:01
ayush3504mithro: yeah02:01
ayush3504mithro: with acrylic being used it's best to have g minimum, otherwise the connectors won't expose properly and DB9/15/25 connector lock screws would not reach the thread. i'll keep g = 0 then.02:03
mithroayush3504: with g=0 can we use 5mm acrylic and still have the lock screws reachable?02:04
ayush3504mithro: no, i think the thread length is less than or similar to 5 mm.02:05
ayush3504mithro: less than 502:05
mithroayush3504: no "i think", you should know from looking at the mechanical02:05
ayush3504mithro: yes, 5 is not a option. the lock screws would not reach02:07
ayush3504mithro: and also, if we use TRS connectors they usually don't overhand more than 2-3 mm. So the a headphone plug won't go in fully even if the PCB sticks to the 5 mm front plate02:08
mithroayush3504: regarding the lock screws, just looking at the connectors I'm not sure what you said makes sense02:09
mithroayush3504: looking at something like
ayush3504mithro: that is an option, i wanted to use the lock screws through the plate to give better support02:14
ayush3504mithro: this just passes the lock screws02:14
ayush3504mithro: and what about the TRS jack problem02:14
ayush3504mithro: 2 mm , 3mm acrylic sheets are easy to find and order02:14
mithroayush3504: ahh, the lock screws through the plate was the bit I was missing02:15
ayush3504mithro: this one doesn't pass the lock screws, this will cause problem02:16
mithroayush3504: so with 3mm is the lock screws through the plate going to work?02:17
ayush3504mithro: yes02:21
mithroayush3504: do you need special lock screws? 3mm would be a lot too lose from pin depth?02:22
ayush3504mithro: another thing, on connecting my USB cable to the atlys i see there's a 3 mm mating distance, so with a 5 mm plate the usb cable is not going to plug properly02:23
ayush3504mithro: no the standards ones are like this:
ayush3504mithro: so from the picture you can make out it will work for 3 mm thick plate02:23
ayush3504mithro: err, the usb on other side.02:24
ayush3504mithro: nevermind the same problem is with DC jack02:26
mithroayush3504: I believe that is why we had to "mill into" the 5mm plate (basically make the 5mm, 3mm thick on the area where the PCB board was02:27
ayush3504mithro: if milling into is an option, we can go as thin as we want right?02:28
ayush3504mithro: the DC jack plugs in completely, can't have more than 1 mm between DC plug and DC jack. the DC jack isn;t overhanging either02:29
mithroayush3504: well, mills have their own margin of error02:29
mithroayush3504: you can just cut a hole big enough for the DC plug, right?02:30
ayush3504mithro: oh ok. yes that'll work02:30
mithroayush3504: you'd probably use a square the same shape as the DC plug front?02:31
ayush3504mithro: so how thin can we go with milling? I guess 2 mm is reasonable?02:31
ayush3504mithro: no the DC plug on official adapter is wider than the rectangle shape02:32
ayush3504mithro: DC plug isn't an issue, make a circular cutout for it. how thin can we go with milling?02:35
ayush3504mithro: i'll make a circular cutout for it*02:36
mithroayush3504: if we can do it without any mill thinning, the better02:39
ayush3504mithro: 3 mm then, assuming 2 mm acrylic isn't common.02:40
mithroayush3504: Well 3mm is equivalent to 2mm with a g=1mm right?02:42
ayush3504mithro: yes02:43
ayush3504mithro: it's practically better to have g = 0 as long as our panel thickness is not less than 2 mm.02:44
mithroayush3504: but that means that the front won't be flush any longer? (Assuming we designing for 3mm...)02:44
ayush3504mithro: the connectors overhang about 2 mm, so if we have 2 mm panel thickness and g = 0, it would flush nicely, if we use 3 mm panel thickness, the connectors would sink in slightly02:45
ayush3504by connectors i mean the atlys connectors02:46
mithroayush3504: okay02:46
mithroayush3504: that goes for all of them?02:46
mithroayush3504: so maybe we should design for 2mm, and just accept with 3mm they would be sunk back by 1mm?02:46
ayush3504mithro: yes, but then the 3 TRS connectors need to flush with the faceplate, because headphones don't work properly until we fully insert them.02:48
ayush3504mithro: and the TRS overhang just 2 mm02:49
ayush3504mithro: i'm concerned if 1 mm sinking of TRS jacks could cause a problem02:49
mithroayush3504: do a diagram and find out?02:51
ayush3504mithro: just tested this with my headphone on laptop, the sound distorts if i take out the plug 1 mm.02:51
ayush3504mithro: i don't know how a diagram would help with that. the connector is designed in a way that you have to plug in the jack completely otherwise it  doesn't work right. And I practically tested just now that 1 mm gap distorts sound with laptop and headphones02:53
mithroayush3504: why is there a 1mm gap? the hole can be cut big enough for the connector, right?02:54
ayush3504mithro: the holes needed will merge since the three TRS jacks are pretty close02:55
mithroayush3504: you talking about on the Atlys?02:57
ayush3504mithro: yea02:57
ayush3504mithro: i can get my stuff to work with 3 mm02:58
ayush3504mithro: also, the width of the motherboard with atlys board attached comes out to be near 40 cms for 6 daughterboard slots. that looks like a lot02:59
mithroayush3504: yeah, 40cm is quite a bit03:14
ayush3504mithro: yeah03:14
mithroso "h" the front-plate height should be 1ru03:15
tpbTitle: Rack unit - Wikipedia, the free encyclopedia (at
mithrowidth wise, there are two options03:16
mithrostandard 19-inch (482.6 mm) rack, or "A/V half rack" 9.5-inch or 241 mm03:17
ayush3504mithro: so this fits in the 19" rack?03:17
mithroayush3504: I think that is a good guide03:23
ayush3504mithro: yeah. thanks for the help03:23
ayush3504mithro: got to eat, brb in 20 min?03:24
mithroayush3504: see how I'm using the spreadsheet to calculate the values for me?03:24
ayush3504mithro: yes, it is interesting!03:24
mithroayush3504: this is how you make flexible design decisions03:24
mithroayush3504: you put all your data into a spreadsheet with correct formulas03:25
mithroayush3504: then you change the base values and see how it effects the outcomes03:25
ayush3504mithro: yeah, got to use them..!03:25
ayush3504mithro: also need to subtract 8.0 and n to get the slot space03:28
mithroayush3504: feel free to change the spreadsheet03:49
ayush3504mithro: yes. thanks a lot! :)03:56
ayush3504mithro: is case clearance what i call g?04:01
mithroayush3504: yes, except on the sides rather than the front04:02
ayush3504mithro: ok04:02
ayush3504mithro: 3mm just for faceplate and other sides going to be 5mm thick right?04:04
ayush3504mithro: or 3mm overall?04:05
mithroayush3504: 5mm on the other sides probably make it stronger?04:06
ayush3504yes, since it's long04:06
ayush3504mithro: what do you say?04:11
mithroayush3504: I think that makes sense05:59
ayush3504mithro: ok06:06
*** rohitksingh has joined #timvideos08:25
mparuszewskimithro: you can now see on and try to edit a file, you will see nice preview :)09:03
aps-sidsmparuszewski: looks cool. But I think it should open in same page, if possible.09:07
mparuszewskiaps-sids: I will check if that's possible to have ability to edit, preview and commit changes to GitHub on one page.09:09
*** tija has joined #timvideos09:24
*** puck has quit IRC10:05
*** puck has joined #timvideos10:05
*** techman83 has quit IRC10:06
*** techman83 has joined #timvideos10:08
*** ChanServ sets mode: +v techman8310:08
*** rohitksingh1 has joined #timvideos10:22
*** rohitksingh has quit IRC10:24
*** rohitksingh1 is now known as rohitksingh10:24
*** rohitksingh has quit IRC10:58
*** rohitksingh has joined #timvideos11:10
*** slomo has joined #timvideos14:03
ayush3504mparuszewski: looks good. just an opinion - I feel that the font for body text isn't as legible as the regular ones, looks nice though.14:29
ayush3504mparuszewski: ok. i just found out that it doesn't render as nicely on Windows. Check the screenshots comparing rendering on ubuntu (left) and windows (right)  -
tpbTitle: mcj.png - Google Drive (at
ayush3504mparuszewski: make sure you actually download the file and see it, you won't see the difference in the google drive preview14:38
mparuszewskiayush3504: thanks, I will take a look on that. It seems that Chrome on Windows doesn't use anti-alliasing for fonts :)14:39
ayush3504mparuszewski: yeah, i guess so.14:40
*** xobs has quit IRC15:03
*** xobs has joined #timvideos15:08
*** CarlFK has quit IRC16:19
*** rohitksingh1 has joined #timvideos16:24
*** rohitksingh has quit IRC16:25
*** xobs has quit IRC16:44
*** xobs has joined #timvideos16:49
*** sarwarc has joined #timvideos17:10
*** mparuszewski has quit IRC17:11
*** rohitksingh1 has quit IRC19:07
*** rohitksingh has joined #timvideos19:11
rohitksinghtija: ping!19:16
tijarohitksingh: pong!19:16
rohitksinghHi! I wanted to ask if I can use USB-Serial code you had written for Atlys. Actually, I'm kind of stuck in removing flickerings in VGA capture so thought maybe some debug output can be useful. What say? Is it stable?19:18
tijarohitksingh: the problem is the driver for linux is outdated. I have tested it on windows.19:21
rohitksinghokay that means i can use it on windows. right? let me try it! thanks! Any info u would like to convey?19:22
tijarohitksingh: you can try it. I have documented it one of my blog post.19:23
rohitksinghok! great! going to boot windows and try it. :)  I desperately need some debugging info19:24
tpbTitle: TimVideo GSoC 2014: MJPEG Optimisation: UART Documentation (at
rohitksinghtija: thanks! :)19:25
*** rohitksingh has quit IRC19:26
*** rohitksingh has joined #timvideos19:34
*** CarlFK has joined #timvideos20:00
*** ChanServ sets mode: +v CarlFK20:00
*** slomo_ has joined #timvideos20:03
*** slomo_ has joined #timvideos20:03
*** slomo has quit IRC20:03
*** sarwarc has quit IRC20:34
*** slomo_ has quit IRC20:46
tijarohitksingh: Is it working?20:48
rohitksinghtija: just now added it to the project. will report it in some time!20:49
tijarohitksingh: I am going to sleep. If there is any issue, mail it to me I will look at it when I wake up. Thanks!20:54
rohitksinghtija: sure! :) have a nice sleep!20:54
*** tija has quit IRC20:58
ayush3504tija / rohitksingh: can we configure the 4 clock pins on VHDCI for regular I/O ?22:24
*** rohitksingh1 has joined #timvideos22:25
*** rohitksingh has quit IRC22:25
rohitksingh1\nickserv regain rohitksingh guddurks22:33
*** rohitksingh1 has quit IRC22:35
*** rohitksingh has joined #timvideos22:36
rohitksinghoops...wrong command in hurry...facepalm :P22:37
ayush3504rohitksingh: happens. Be sure to change the password ;)22:38
rohitksinghayush3504: yeah we can absolutely use them. They are only special because they provided highspeed low-skew jitter free routing for clocks!22:38
rohitksinghayush3504: yeah! changed it...sigh! :P22:38
ayush3504rohitksingh: great! So I get up to 44 I/O pins right?22:39
rohitksinghayush3504: yes all the pins apart except GND, VU and VCC22:40
ayush3504rohitksingh: thanks22:41
ayush3504rohitksingh: 40 pins included clock pins. counted it wrong.22:53
rohitksinghohh ya...its my VGA board they are 10X3 RGB pins, 2-I2C, 3-SYNC, 1-DATACK, 1-PWRDWN, 1-OE_Field, 1-CLAMP, 1-COAST = Total 4022:57
ayush3504rohitksingh: even i have two pins left, 6 ports, 6 pins per port, 2-I2C23:10
*** rohitksingh has quit IRC23:39

Generated by 2.12.1 by Marius Gedminas - find it at!