arm-netbook

arm-netbook@lists.phcomp.co.uk

1082 discussions

On (potentially not) 3D printing laptop case parts
by Forest Crossman 25 May '17

25 May '17

Hi, all, The recent update on Crowd Supply (regarding the difficulty in mass-producing parts with 3D printers) reminded me of some RepRap/RepStrap innovations from a few years back: - https://web.archive.org/web/20150318000301/http://justindunh am.net/making-cast-resin-3d-printer-parts/ <https://web.archive.org/web/20150318000301/http://justindunham.net/making-c…> - http://reprap.org/wiki/Clonedel - http://reprap.org/wiki/Moldmaking_Tutorial This was before cheap, mass-produced 3D printers became widely available, so many people were bootstrapping their RepRaps by purchasing 3D-printed parts from others. At some point, someone realized that it was a lot faster to 3D print a set of parts, make molds from them, and then use those molds to make additional part sets, than it was to separately print each set of parts. I'm not sure how complicated the parts of the laptop case are, but if the parts could be re-designed to be optimized for resin casting, it might cut down on the production time. Of course, I've never done any resin casting before so I don't know how feasible this would be in practice. Also, I have no idea how fast modern 3D printers can run now, so the speed advantage might not even exist any more. Anyways, I just though I'd mention this on the off chance it might help speed things along. Keep up the good work! -- Forest Crossman

3 3

Re: [Arm-netbook] rk3399 what full schematics does lkcl want?
by ronwirring＠Safe-mail.net 24 May '17

24 May '17

-------- Original Message -------- From: Luke Kenneth Casson Leighton <lkcl(a)lkcl.net> Apparently from: arm-netbook-bounces(a)lists.phcomp.co.uk To: Eco-Conscious Computing <arm-netbook(a)lists.phcomp.co.uk> Subject: Re: [Arm-netbook] rk3399 what full schematics does lkcl want? Date: Sat, 20 May 2017 16:27:33 +0100 > btw ron please do cut context that's not necessary, it was very hard > to find the questions you wrote. > > On Sat, May 20, 2017 at 3:47 PM, <ronwirring(a)safe-mail.net> wrote: > > >> then once that's done you then have to do the PCB design *from* the > >> schematics, and that takes even longer. and will contain mistakes. > >> > > > Is it difficult or takes a lot of time? I ask because the converting to cad files cannot be split up in numerous pieces and put on a website, where people can each solve one piece? > > so, cutting all but the relevant context, and answering just the > questions you asked: there are two aspects. > > firstly, the schematics: these just take time. it could be > parallelised in theory > > secondly, the PCB layout: this takes time *and* is difficult. > creating the footprints for compoents could be parallelised, but the > PCB layout - positioning of components, connecting components, laying > out tracks and the planes? no - for such a small PCB that is best > handled by one person with the required expertise. > > and that task alone takes months to do from scratch. > > l. Do you make your pc card a20 cad files public? > > _______________________________________________ > arm-netbook mailing list arm-netbook(a)lists.phcomp.co.uk > http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook > Send large attachments to arm-netbook(a)files.phcomp.co.uk

2 1

Re: [Arm-netbook] mali gpu reverse engineering lkcl may ignore
by ronwirring＠Safe-mail.net 24 May '17

24 May '17

-------- Original Message -------- From: Neil Jansen <njansen1(a)gmail.com> Apparently from: arm-netbook-bounces(a)lists.phcomp.co.uk To: Eco-Conscious Computing <arm-netbook(a)lists.phcomp.co.uk> Subject: Re: [Arm-netbook] mali gpu reverse engineering lkcl may ignore Date: Mon, 22 May 2017 17:28:21 -0400 If you reinterpret what I write then tell on what grounds. I think lkcl's reasons for not reverse engineering a mali are right. Weighing them up against the importance of getting a libre software gpu, I reach the conclusion that the reverse engineering should be done. There have been some remarks about the probability of a successful crowd funding. I mentioned the numbers 50000 people, each paying 5euro. Notice on this email list, people want to pay 5eu, if they get the software in question. It is safe to say more than 10 million people have gnulinux on their computer? A major part of them know about the importance of libre software and a part of them would want to act on it. I agree about the responses on the practicability of a crowd funding. I am not sure a crowd funding can be done successfully. I would like to know why? Why cannot such campaigns be coordinated centralized? We should have a known entity, which would approve libre software related crowd fundings. People would know them to not be scams. Why does an entity like the free software foundation not take on this task? I should ask them. About these fragmented crowd fundings, you yourself have to make a validation, which is difficult. When lkcl initiated his pc card crowd funding, I did not know him. I still mostly do not. On trisquel's forum people referenced thinkpenguin. Thinkpenguin endorsed lkcl. That is why I supported lkcl. Is the free software foundation not involved because they fear scandals or not have enough resources? The matter about libre software consumer electronics, I consider highly important. There have been remarks about the retaliation towards people who want to do reverse engineering. How to deal with it. One option would be to let one entity collect the funding, manage public relations and another hidden team do the reverse engineering. Everything still legal on taxes. Legalities seems to be an important issue. Are these legal concerns an american matter? I do not know everything which may happen but if the public part of the crowd funding is located in eu, I think a campaign should be fine.

2 1

Re: [Arm-netbook] arm-netbook Digest, Vol 82, Issue 47
by Ismo Väänänen 23 May '17

23 May '17

> > Date: Sun, 21 May 2017 00:45:30 +0100 > From: Luke Kenneth Casson Leighton <lkcl(a)lkcl.net> > To: Eco-Concscious Computing <arm-netbook(a)lists.phcomp.co.uk> > Subject: Re: [Arm-netbook] GR8 based EOMA68 card > Message-ID: > <CAPweEDywhR-LUBAZojgGr2Hn0WJijOuZayxvJBK4W > DH7MuhRDw(a)mail.gmail.com> > Content-Type: text/plain; charset="UTF-8" > > ismo, hi, > > it was just too tempting to have an initial go at converting the > CHIP_PRO_v1 schematic and PCB into a version of an EOMA68-GR8 card, > deliberately cutting it down to the absolute basic minimum - after > only about 6 hours i'm pretty much done connecting everything > together, having cut/paste bits and pieces over: > > http://hands.com/~lkcl/eoma/gr8/ > > those are exported from PADS into 9.0 ASCII format if you want to take a > look. > Sweet, thanks. Just imported these to Altium. Actually noticed that the Allegro .brd import in to Altium needs Allegro installed to function. The imported layout looks like this: https://i.imgur.com/uPmEsQb.png > importing from ORCAD/Allegro doesn't half make a dog's dinner of > things... *sigh*... > It so does. None of the component parameters for the passives carried over in the import. The component values imported as separate strings for the passives, at least the cases are OK. Of course, the mentioned cases don't exist. So that needs some additional work. > anyway it's the *absolute* absolute bare minimum. TV-out, gone, > HP/MIC, gone. OTG, gone. NAND, gone. SDC2 is connected to a MicroSD > slot, SDC0 is connected to EOMA68. SPI2 is connected to EOMA68 so in > theory could also be used as bootable media (factory-install purposes > only). boot is basically from on-board MicroSD, that can be > over-ridden if people want to by using a Housing MicroSD card (if the > Housing *has* one). > That's OK with me. More space to use for potential future expansion. > the layout's hilarious: the PCB is over 50% completely empty. what i > haven't done is: > > (1) complete the tracks / connections > (2) add a VCC-3V3 power plane or other power planes. god did ORCAD > make a mess... *sigh*. > (3) convert any of the 0201 components to 0402 (for goodness sake > don't use 0201 components!) > (4) sort out any tracks as a result of 0201 conversions > Ya 0402 is a lot more sane to deal with than 0201. > (5) re-add the copper pours for the power outputs from the inductors > (to the AXP209). > This and in general figure out the power stuff. > (6) do a full review to check that the dog's dinner mess made by ORCAD > hasn't split some of the NETs. PADs doesn't support multi-named nets. > i found one (and joined them) but there may be others. > (7) sync'd PADS schematic and PCB so that a proper NETLIST review could be > done. > It's gona be FUN mating the CHIP Pro schematic files with that layout. Absolute joy. > remember that this *really is* the bare minimum - it'll be amazingly > an under $10 BOM. also that if you _did_ want to add HP and/or TV-out > and/or MIC sockets, as well as OTG.... you can't: there's not enough > room on the connector, not and have MicroSD as well. so you can do 3 > out of 4 of those connectors but not all. > I'll fully decide on this later on, USB+Video would allow for standalone use. And sort-of usable while standalone is aligned with the spirit of the EOMA68 standard (or how I see it at least). also bear in mind that there's a couple of tabs on the Litkconn > casework that fit *just* either side of the MicroSD, whch make it damn > awkward to fit the MicroSD anywhere but directly in the middle of the > end of the Card. > Where can I find datasheets, drawings or 3d models for the Litckonn cases? Can I buy a few sample units from somewhere for fiddling around? > also bear in mind that finding mid-mount 2.5in and 3.5in multi-pin > jacks is a complete bitch. i *might* have one supplier (who speaks > chinese only) who *might* be able to help, there (Runde). > I'm perfectly OK with deleting the sound input/output. Just a shame to leave the sound hardware unused when it exists. > so - what ya wanna do? do you want to take over this layout/schematic > from here and go with it? if you leave it as-is it should be easy to > finish within a matter of 2-3 weeks, which means that it's potentially > possible to add to the upcoming (2nd) planned crowdfunding campaign... > or you could run your own. > More or less regardless of crowdfunding or anything I'll continue doing work on my vision of what a GR8 based EOMA68 card would be. But my progress won't be fast as I'm working full time and doing the same stuff at home as I do at work is not attractive every single day. A low cost alternative for low powered embedded applications might also be attractive to some folks and in general for testing and debugging housings. On an other note, replying to this list while it's in digest mode is a bit weird. I might have to change it from digest to the individual message style and hope that gmail is good enough at clumping the threads. -Ismo / 2ftg

3 2

GR8 based EOMA68 card
by Ismo Väänänen 23 May '17

23 May '17

Hello, The recent talk about de-blobbing R8 and thus also GR8 inspired me to do a quick write up on how I see a Next Thing Co. GR8 System-In-Package chip being used in an EOMA68 compatible card. The latest version of the write up can be found here: https://sites.google.com/site/oh2ftg/eoma68/eoma68-gr8 I named the project "EOMA68-GR8" because why not. Like Vincent I'm doing this on the side and also with Altium. I haven't done anything this complex in Altium before it's likely going to be an interesting challenge. GR8 has TTL/RGB LCD interface, USB2.0, I2C, SPI(SDMMC) and so on. All the interfaces to make a compute card compatible with EOMA68 are there. Except if I want the card to have "front facing" USB I'll need to include some USB HUB chip like TI TUSB2046, which conveniently has no firmware being a state machine. I'll have to look at how the interrupts go, at AXP209 PMIC and it's routing. And decide on if I'll layout the NAND as the talk about blobs being required for NAND support sounds worrying. That and seeing in general if even a half reasonable layout is possible on 4-layer FR4 to get cheaper rates on the pcb's. At least there's no DDR RAM to route, but the TTL/RGB fanout and length matching is gona be chore. One more good reason to learn how the automated length matching in Altium works. Sourcing connectors and housings in small quantities. All that fun stuff. If anything I have missed comes to mind please mention it, I'd rather hear it now then when I have prototype pcb's at hand or layout nearly done. Cheers, Ismo Väänänen de OH2FTG alias 2ftg

8 13

designing a low-cost decent 3d printing board
by Luke Kenneth Casson Leighton 23 May '17

23 May '17

hi all, ok so i've been looking around and the practice of creating a "modular" 3d printing electronics board is extremely common, thanks, many years ago, to the stupid, stupid decision to use prototyping (evaulation) plugin boards with 1.3 to 2.0 *amp* stepper ICs mounted onto micro-postage-stamp-sized PCBs. the problems these cause are endless. *some* people have actually read the datasheets associated with these driver ICs and have designed PCBs that respect the advice of the designers, by making sure that a large part of the PCB's copper is used as a heat-spreader. however the practice is not very common, and the people who make such boards are typically actual well-trained electronics engineers from the west, but who then use western PCB manufacturing and assembly plants, end result: very expensive boards ($120 to $200). as i would like to see both a low-cost board made *and* have a decent design, and having researched this for many weeks and found that what i am searching for DOES NOT EXIST i need some help. the board i believe is best started from is the Duet WIFI: https://github.com/T3P3/Duet this uses 2.5A TMC2660 "silent" steppers and importantly they're SPI-based. the idea is to cut the DuetWIFI down to the *absolute* bare minimum, turning it into the exact same thing as TRAMS (which uses the 1.3A TMC5130). so instead of an on-board ATSAM4, you use an arduino due. instead of WIFI you use a *standard arduino WIFI shield*. now, debatable is whether to split out the MOSFETs, endstops and thermistors into their own separate shield as well, which i feel might be sensible. in particular i would very much like to "stack" the steppers because i have a design where it is going to use *four* z-axes. in a first iteration those may be wired up as 2x 2-serial in parallel, however it would be very *very* nice to be able to just add a second "stack" of 4 z-axis steppers and use them to do automatic bed-levelling. so to do that it would mean having on-board jumpers that could select alternative GPIO pins... or maybe to use an I2C-based or SPI-based GPIO expander in order to reduce the amount of GPIO needed. however an alternative idea which would do just as well (without needing too much work) would be to have the expectation that there would be a "base board" into which *two* modular 4x TMC2660 stepper boards would be plugged, and an arduino due, *and* the MOSFET/endstop/thermistor board, and the *base* board would take care of GPIO expansion. anyone considering this: the layout on the DuetWIFI for its steppers is very *very* specific and should really not be touched or altered, merely cut away surrounding parts and add the necessary arduino-compatible headers. any takers for what is likely to be a quite straightforward project? l. --- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68

2 2

mali gpu reverse engineering lkcl may ignore
by ronwirring＠Safe-mail.net 22 May '17

22 May '17

Because I deleted a previously email about this subject, I start a new email. Info. Lkcl said, he is not in favor of reverse engineering a mali gpu. Because it is about 150000eu and new gpus will emerge during the reverse engineering and the outcome is uncertain. I agree on his arguments. I do not find them strong enough. Maybe lkcl assumed it was about a mali gpu on a pc card. It was not. My question was a general question about getting a broadly known mali gpu reverse engineered. In my email I referenced the https://www.kickstarter.com/projects/1771382379/firefly-rk3399-six-core-64-… because according to lkcl it is as fast as a duo core intel mainboard and only the mali gpu software is not source code. There is no new arm source code computer for sale. I think one should be provided. People must be able to get one. 150000eu is a crowd funding of 30000 people, each 5eu. I would pay an extra 5eu to be able to buy a source code computer. I do not know if 30000 people are interested or if they can agree on one board. You cannot get the mali source code faster, if you put more people on it?

6 6

"lkcl may ignore" - please do not put that in subject lines
by Luke Kenneth Casson Leighton 22 May '17

22 May '17

so this is a nice experiment, but i'd really prefer it if people did not put my initials into subject lines. the result is dozens of repetitions which gets stale very very quickly: http://lists.phcomp.co.uk/pipermail/arm-netbook/2017-May/thread.html

1 0

EOMA50
by Luke Kenneth Casson Leighton 21 May '17

21 May '17

http://elinux.org/Embedded_Open_Modular_Architecture/CompactFlash#Table_of_… if you recall, someone kindly pointed out that USB3.1 does *not* need two sets of Tx/Rx differential pairs but only one. so ironically that means that four pins on an extremely low pincount standard are now free. i thought, "ah ha! should add a 2nd USB2 to EOMA50!" ... but should it? one of the things about SoCs that are that small, they can fit onto a 43x30mm PCB is: the probability of the pincount being high enough for them to have two USB2 interfaces is... debatable. the GR8 definitely does but that's just one SoC. and if it *doesn't*... then fitting a USB hub on there to provide two USB2 interfaces is going to be a bitch-and-a-half. also, two new pins are definitely free: "by default" i would make those EINT2 and EINT3 *but* the other option is to make them SPI data lines 2 and 3 (for 4-bit DDR SPI). choices, choices... i'm leaning towards 2 extra EINTs particularly as i have an idea in mind for a modular smartphone, and the addition of extra EINT-capable GPIO would allow a sub-module standard which has an EINT line "per module". mind you a break-out GPIO device (such as oo i dunno... a $1 STM32F!) is probably going to be needed for that design concept anyway. bizarre, isn't it: you can either choose a $1.50 dedicated GPIO breakout IC which only has 12 pins... or you can choose a $1 EC which has 32 or 48 pins and has GPIO, USB, SPI, ADC, DAC, I2C, UART... so bizarre. anyway. thoughts appreciated. l. --- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68

3 11

GR8 based EOMA68 card / EOMA50
by Mike Leimon 21 May '17

21 May '17

> the layout's hilarious: the PCB is over 50% completely empty. Allow me to be the first to point out that for this very reason this SOC might be an excellent candidate for the first SOC to adapt to the EOMA50 standard. > remember that this *really is* the bare minimum - it'll be amazingly an under $10 BOM. And here is another good reason to use it as a first pick... -Mike

2 1

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

arm-netbook