This forum uses cookies
This forum makes use of cookies to store your login information if you are registered, and your last visit if you are not. Cookies are small text documents stored on your computer; the cookies set by this forum can only be used on this website and pose no security risk. Cookies on this forum also track the specific topics you have read and when you last read them. Please confirm whether you accept or reject these cookies being set.

A cookie will be stored in your browser regardless of choice to prevent you being asked this question again. You will be able to change your cookie settings at any time using the link in the footer.

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
A cool product with a seeming lack of commitment.
My first post here, so hello all. I just got my first Tinkerboard and I'd like to share my thoughts so far. I am far from a Linux noob, having been involved in it since pretty much the beginning, well actually kind of before the beginning, I was involved in the open source projects that came before it and eventually got layered on top of Linus's kernel when he wrote it.

First off, I have several Pis and several Up Boards running, the other ASUS produced, well AAEON, but still ASUS, RPi form factor micro board. Both these boards are well made products, and the Tinker in particular is a very pretty board, both are kind of power pigs for such small boards, and both have some rather concerning issues with their respective Debian distributions.

However, the power sensitivity is no joke, I spent hours trying to get something to power this board up, all sources that can power a Pi without a hitch, eventually having to power it with a high output Quickcharge 3.0 power block and a super premium micro USB cable. For those of you experiencing it the symptoms of being under powered are, red LED comes on, perhaps the green one will blink once or twice then nothing, everything goes off and no video ever displays over HDMI. Honestly I do not have one of these high output chargers in the same location as an available HDMI display so I've actually not seen the video output as of yet, this requires more parts apparently. This in itself makes the Tinkerboard *not* a Pi replacement. The Pi can pretty much be plugged into just about any USB power source anywhere and boot, even if the tinker used a specialized concentric power barrel connector it would be better (this is what the Up Board uses), at least then it just becomes about being tethered to it's own special cable but could still be powered by any USB A port, even if restricted to USB 3.0 power levels, but it seems to not only be picky about cable but the USB source as well, meaning portability is not very good, this is pretty much a fail for what the board is. This "feature" right there effectively ties the Tinkerboard to something, maybe even a bit more complicated than, a power brick. They really should have made one of those jumper locations between the Micro USB and HDMI a direct connect +5V power input.

Issue number 2, and sorry but I have to mention this here, because my issue is about where "here" is. Why is the forum for the Tinkerboard here, and where is here? Why is the Tinkerboard forum not under a "Community" tab off of the ASUS Tinkerboard product page? ...and why isn't there a Wiki there, or a product details page that doesn't involve some obscure, not well described, docs you pretty much have to download blindly, zero HowTos and a pathetic FAQ. This board has through holes on it with no manufacture descriptions of their functions, no header mounted there, no description of which IPEX connector you actually connect an external antenna to despite bragging about the feature. So what? The expectation by ASUS is that they expect young hobbyists trial and error their way to getting things to work, that they expect them to take a soldering iron to the board to put headers where they were not installed? This is just amateurish, and short sighted. The Up Board BTW has full docs, a Wiki and a fairly elaborate forum, in other words the industrial division made way more effort to construct hobbyist resources for the more industrial product than the consumer division did for a product aimed at hobbyists.

There is no description whatsoever of the format of, or a basic example of using the OS images. It basically relies on some knowledge of the procedures of how to do things on a Pi and apply them to a Tinkerboard. Why on earth would you do that? Just put on your website that it's a DD image and an example or two of how to image it onto an SD card. This is like a new car manufacture not telling you which fluids to use because they just assume you'll just figure out something that will work. Also not sure why ASUS is going down the Debian Stretch 9 path when Jessie 8 is the current stable release, and the release that the latest Raspbian release is based on, that they apparently expect you to refer to to image your SD, it's all kind of nonsense. On top of that TinkerOS has a kernel compiled like someone that hasn't been on a Linux box in 20 years would compile a kernel, nothing except the WiFi is a module, it's compiled as a monolithic kernel. There is no Linux-Image-`uname -r` header source package in the repo as far as I can tell, so you may be able to download the source tree, but there is no reference tree. Again the Up Board has this, I mean come on ASUS walk across the hall and talk to the people in the other room. Not that the Up Debian distribution (UbiLinux) is perfect, it isn't, it is installed vs imaged because the Up Board uses eMMC DOC as opposed to an SD card for storage, and UbiLinux has pretty much zero install options, which for the industrial division seems pretty lacking, but it's way way ahead of the expertise that was put into TinkerOS which should have been loaded with potentially useful modules, like the spi, usb, networking, filesystem, industrialio, fbtft, video and audio modules as well as overlays like the Pi uses. ASUS went to great lengths to make the Tinkerboard a drop-in physical replacement for the Pi and then put out a distribution that takes advantage of nothing that exists on the Pi. Again the Up Board by contrast, on the ones I have running, have 162 modules loaded, the Tinkerboard has 1, the WiFi, and it doesn't work.  

So on the subject of WiFi, I am running the 1.4 version of TinkerOS, because frankly I installed the 1.8 version and struck out on all of the first half a dozen packages I tried to install, all missing from the repos. WiFi doesn't work, wpa-supplicant pukes when you try and bring it up, on the other hand with an external antenna the signal strength is really good... but you can't actually connect to it. Of all things not to work this is one that should work without "tinkering", networking is the key to tinkering, and dysfunctional networking just makes it a frustrating product. May have to try 18.1 Stretch again but really Stretch IMHO is just not mature enough for a board like this yet, Jessie is very mature, particularly for the armhf branches. Bleeding edge, not in the main branch is not a good plan for a hacker board, sorry tinkerers board. There is really not a good rational for Stretch on the Tinkerboard, it should be Jessie at this stage in the Debian cycle, and it should be complied as modular as possible -OR- as similarly as possible to Raspbian, it's neither. Because of the lack of source, lack of modules, unique release compared to other SBCs, the Tinkerboard has solidified it's position as just a little board that runs something Debian like, oh or a board that runs Android but has all kinds of features you can't actually use in Android.

So the bottom line here is, I think this board holds promise and I will likely buy a few more once I get things working right on it, however I just feel like I am going to have to hack the hell out of the OS to make it useful. Therein lies the rub, the 1.4 Jessie release is, as far as I know, the only one with a source tree. The GigE, better video and better WiFi (if it would only work and really really wish it did 5.8Ghz) addresses some key points to the shortcomings of the RPi3. The power thing is just a fail, making the Tinkerboard a non-starter for a whole lot of projects where a Pi could be used. It really is kind of a big deal. Unfortunately and honestly, unless some of these issues get addressed and soon, I feel the Tinkerboard will be a one shot product line, I don't think there will be a TinkerboardII, it will likely be eclipsed by a RPi4 or an Odroid board and fade into obscurity, a shame, because I think the board has great potential but will die because it's poorly executed. They should have powered it from one of the USB A connectors or made it PoE, or made the audio connector a power barrel and supplied a barrel to USB cable to keep RPi drop-in form factor, and they should have let the Up guys do the distro. They need to host their own forum, and a Wiki, and a real details support page.

Since this seems like the one and only resource for the Tinkerboard, I will try and post things I do that might be useful, as I have in the Up Board Forums. but I really think there needs to be a Wiki and I think this and that Wiki is really something ASUS should be doing. Reality is that many people, like myself, will come across this board and feel like they are missing the real place they should be looking. Don't get me wrong, I appreciate this resource being here but I see no evidence that anyone at ASUS is paying attention to this resource at all.
Welcome, I have 4 pi2's, 3 pi3's, 3 zero's, 1 zero w, 1 riotboard, 1 odroid, 1 beagleblack and 1 tinker, the best of them all is the pi2's, 3 are running now and have been for over a year, the worst is the riotboard that don't like debian but would like android so it can earn fees selling apps, and that is the seed of it all there, FEES FROM SELLING APPS, a lot of theses boards claim to be open source for the experimenter but make the path down the android route selling apps very easy and the debian path very hard, all things these days are controlled by the rentier class looking for a life style as a god on rent.

Now as you are clearly from an older generation like me then what you can do is help make that debian path smoother and deprive the parasites of fees.

As for power what I have found is that if you use power supplies that manufactures suggest then you will be ok, a lot of these phone chargers with many usb ports don't supply what they say on the tin. I have not had any power problems but have had problems with sd cards burning out, I have used Kingston 8 gb cards and have lost 3, I have now put the OS on a usb stick and running it from that, in other words the first partition(boot) is on the 8gb sd card and the second partition(OS) is on the 8gb usb stick, it is very slow now, I have given it a fixed ip so other servers can monitor it and we will see how long it runs
So true. I am moving back to RPI3 (until new release of TinkerOS) since last 3 weeks I am not doing anything creative, I just try to fix ASUS's bugs. It seems this is product for people with really serious background of system engineering.
- my 4k tv is not supporting via hdmi (RPI does support);
- bluetooth and wifi work really slow together (RPI ok);
- I still did not figure out how to change framebuffer resolution (raspbian has greatest support of that);
- strange bug with i2c which makes impossible to work with certain devices that I need - it does stop my work completely (RPI works good).
(05-29-2017, 12:04 AM)skorpioskorpio Wrote: Why is the forum for the Tinkerboard here, and where is here? Why is the Tinkerboard forum not under a "Community" tab off of the ASUS Tinkerboard product page?

This is currently an unofficial forum - I set it up as a normal TinkerBoard user when I couldn't find any info out there.

A huge thank you to all who have contributed to make it the good resource it has become.
ASUS are of course well aware of this forum...

(05-29-2017, 12:04 AM)skorpioskorpio Wrote: However, the power sensitivity is no joke, I spent hours trying to get something to power this board up, eventually having to power it with a high output Quickcharge 3.0 power block and a super premium micro USB cable. 
The Pi can pretty much be plugged into just about any USB power source anywhere and boot, 
The power thing is just a fail, making the Tinkerboard a non-starter for a whole lot of projects where a Pi could be used. It really is kind of a big deal.

Yes, it draws a lot of power, an official PSU would help here, even if still on microUSB. 
Most USB chargers are poor at voltage stability when you go over 1A.
On the Pi - it's ability to able to use just any USB power source stopped with the release of pi3 IME - similarly the pi3 needs care with power (partially fixed by RPi throttling back most of the time).

Also on availability - I can walk into a high street shop in the UK and buy a tinkerboard or raspberry pi.
I can't do that with any other boards like orange pi, odroid etc.
(05-29-2017, 12:04 AM)skorpioskorpio Wrote: My first post here, so hello all.

Hi welcome.

Power...i hear you... at the wrath @TKaiser, yes you need to try a few different bricks, I'm using a 2.4amp per usb port 5 Port 50w Anker USB Charger PSU with 0.5Meter usb and it's pretty stable. I don't run the tinkerboard at full bore but if I want CPU grunt I use the Quad Core Xeon box...not the Tinkerboard.

I hear you too about the lack of official forum, although there is a official blog/board on also a ASUS facebook and twitter feeds.

With absense of the Pi 4, there has been a rush to market of various RPI clones, all of which have limited support. Maybe there is some confusion from all these vendors as to audience they are advertising to and that's pretty sad. It might in end dilute the market.

The simple truth with all of the RPI clones is that they are not remotely "ready to wear", a appear to have an expectation of the buyer is some kind of hardcore developer / script kiddie. They are difficult and they won't get much easier and the lack of documentation is really to be expected and will no doubt continue, until we write it.
After I unplug micro usb power cord it was still alive taking power from HDMI cable. I did not expect that.
Honestly, I'd be perfectly happy with just working WiFi, and getting the AdaFruit 3.5 PiTFT working, which seems unnecessarily complicated. Have the driver module compiled, but doesn't create the /dev/fb1 device, and the backlight on the display strobes randomly, hmm, pretty much all the time, so I think it may be again a power issue, I think maybe it can't supply enough power to the +5V on the 4 pin to power the display. Again I have these working on Up boards no problem and they're 3+ Amp boards. The displays also work on the RPis I have around. On the Up Boards you just load the spi and the fbtft modules and done. But this board I am not sure what I am even dealing with, is it a power problem? a driver problem? a module problem?
Well after 2 weekends tinkering with the Tinkerboard, I am hanging it up. The specific things I needed to work on the board were the things that didn't work. The SPI/GPIO driven AdaFruit display HAT was one of them, I systematically went through section by section, line by line in the (not latest) kernel recompiling it to be as modular and as similar as possible to Raspbian, never changing more than one section at a time between compiles to get something workable. But in the end one thing was always true, when the board booted the display would always strobe several times and then several again when booting (and pretty much continuously through kernel compiles, whatever that was about). My guess is it's one of 2 things; first it's not possible to supply enough power though the micro USB connection to power the Tinkerboard *and* the display, this is disturbing since the display HAT draws less than pretty much any USB device, which I never had any USB devices plugged into it, so it was the sole peripheral draw on a board that was being powered by a 3A+ source and an overkill of a USB cable. Second possibility is that the dual use function of pins 18 and 22 (GPIO 24 and 25 on the PI and I have forgotten already something in the high one hundreds on the Tinkerboard). Anyway these are also used as RTS/CTS for UART1, which I don't know maybe is cool I guess for someone building RS232 controlled things to have a full implementation on the GPIO header but given that everything on the board is +5V or less and an EIA RS232 standard port is actually barely within spec at +12V I am not sure I see the point really. Anyway I spent another several hours trying to figure out how to disable hardware flow control on RS232, but the general consensus was that any serial port will assert RTS upon hardware initialization, pretty much always. On the plus side I was actually able to get the Tinkerboard to see the display and create the /dev/fb1 device for it, and return clean debugging info from it. This required modularizing spidev and spidevice in the kernel and adding the staging small display drivers, but in the end, um, not.

Issue two was a completely flakey WiFi, could do an iw scan and see every AP near me, and had really good signal strength on an external antenna, but wpa_supplicant would fail and the board would go into a loop trying to initialize it, you could ssh to it through the Ethernet port, but the HDMI output was just stuck scanning and erroring out. Also seemed once you added an /etc/network/interfaces/wlan0 file, you were stuck in this mode, even if you subsequently deleted that file, strange and, well, unacceptable. Anyway this was the secondary problem and it may have been solvable, but never got that far and frankly, I have a dozen Debian boxes with WiFi on various platforms and have never had to go digging for solutions really, so this is another fail. Also problem is the same regardless of kernel configuration, so it's not a driver issue, I don't think, it's a distribution configuration issue or perhaps something related to the Rockchip hooks to WiFi. Add to this the kicker, as I decided to just box up the Tinkerboard and considered just returning it to Amazon, so in the process of replacing the WiFi antenna jumper, carefully removing the external antenna I had on it, and the u.fl connector on the Tinkerboard popped off the board, oh WTF!

Also as my last ditch I decided to try to install Android on it, see how that went, it didn't really (This is before I rendered the WiFi permanently useless). Lots of effort to get Google Play installed only to have it immediately update and kill itself. Got a few things to work from side-loading, was interesting, but networking is again an issue. Seems WiFi and Ethernet can't coexist, if you bring up the wired Ethernet the WiFi drops, and, well an awful lot of things just fail to install, which I don't really see much point to Android if you have to go to great lengths to force load things. I mean Android is all about the PlayStore, and if that doesn't work and apps require all kinds of manual intervention, and it's all wrapped in an interface that is specifically designed to hide you from such things, what is that really? I do think there is a real market for a Google Play registered Android hobby board, but this isn't it, despite being produced by a company that probably has the means to make that happen.

So in the end, I think the Tinkerboard is aptly named, it gave me 2 weekends worth of tinkering for $60, like an unsolvable puzzle. Didn't get anywhere on the project I bought it for, by contrast the Raspberry Pis I had it to potentially replace I had 2 setup of them in an evening to far beyond where I would probably ever get to with this board, given the resources available. Sure I could get it to function as a general purpose Debian board with wired networking from the official distro pretty easily, but to get it to function beyond that? Not so much, and even at that I was dedicating a power source and cable that together cost as much as a Pi3 to power it, and like a Pi, you still need an SD card. At that point the Tinkerboard is in the same price point and base functionality as an Up Board which is a way more capable board and things do actually work on it, it doesn't have WiFi, of coarse from my perspective, neither does the Tinkeboard, and even had I gotten it to work it's only 2.4Ghz 1x1 and it seemed like it was likely to be limited to 20Mhz channels. My Up Boards are running a much better RT5572 chip based 2x2n WiFi waferboard from SparkLAN that supports 80Mhz channels and connects at 150+ Mb/s on 5.8Ghz, something a Tinkerboard will never do without the same investment as I have in the Up Boards. The limiting factor with the Up Boards is simply heat and while the same form factor as a Pi, the connector layout is different enough to make it not really compatible with Pi cases, also the 40 pin connector is 3-4mm higher than a Pi. The plus on the Up is that it has 2 extra USBs on a header on the board, so on mine I have an SSD and the WiFi connected to "internal" cableing which is kinda cool.

I considered the Odroid-C2 as well, but it lacks WiFi and doesn't have a board header for USB, so it ends up being another Frankenstein project with a bunch of external cableing or dongles, which I am trying to avoid. It also looks like it's probably a heat monster. So back to the world of Pi, a friend of mine has everything needed, including the skill level, to hack a u.fl connector onto a Pi3 for an external WiFi antenna, which is an important part of the project, since the goal is to put it into an extruded aluminum case. The traces and pads are already there on the Pi3 to do this (and BTW are there on the Pi Zero W as well). Yes it's slower, yes it's limited to 1080p and yes it has less capable sound than the Tinkerboard, but there are endless examples of Pis doing everything I need the board to do.

The project BTW, is a dual screen media center, a clock and status displayed on the little 3.5" TFT display while the main HDMI out displays on a TV, housed in a mildly hacked extruded aluminum Raspberry Pi case, preferably fanless and silent. You see over the last year or so, I eliminated my cable boxes and replaced then with Windows Media Center boxes pulling from cable card Silicon Dust tuners, which work well but Microsoft end of lifed WMC at Windows 8.1, these run on little i5 machines not much bigger than a VHS tape. Also I used to have atomic synced analog clocks hanging over all the doorways, in a six month period, after running flawlessly for years and years, they all croaked. So between the dead atomic clocks and returning my cable boxes, I no longer have a clock anywhere, except for one alarm clock in the bedroom that I have to reset twice a year. So essentially I've become timeless. The media aspect is a bonus, but really it's more about a digital display clock I can see from across the room, next to the TVs that sync over NTP. I need at least 4 of these, I also want to be able to feed them what channel the WMC boxes are tuned to when I change channels, and maybe send me reminders and such. I've already figured out how to use 2 independent displays on Debian that have nothing really to do with each other. I've never really had particularly high expectations as far as performance from these boxes, I have i5 boxes for that, that are upgradable to i7s if needed, they have 16GB of RAM, 3 NICs and are capable of 1080p on 2 displays at the same time. 95% of content is upscaled to 2K anyway, and that's all streaming and the TVs (Android) do that natively from just about any source that I need it from. 2.4Ghz 1x1 WiFi is never going to stream 2K anyway, as I am using the internal wired NIC, on what ever box it is, for access to my backend network where my NAS lives.

The Tinkerboard has been a big disappointment, and for me past tense, it's just not stable enough to be what I need it to be. That said, compared to the Up Boards, even the Pi is a toy, but requires far less stuff to accomplish the goal. Currently I have Pi2s which lack the 2nd network interface for the project, the Pi3s have that and the Tinkerboard had that. I have had Pi2s die, but they're cheap. I'll be modding 5 of them to get the 4 I need so I have a spare on hand. I already have 2 of the 4 displays I need, I had one for the Pi2 test system and one for the Tinkerboard test system, but they're easy to get, and so are Pi3s. I don't have to buy any kind of power supplies, I already know the TVs are capable of powering them on their own, so long as they follow suit with the Pi2s. I would have liked my NAS connection to be GigE, and decent sound would have been nice, but I may possibly be able to hack in a DAC and still have it fit into the tiny case, we'll see. The DACs do not use the same pins as the displays do, and I have a good 10mm or more between the display (top) and Pi (bottom) to sandwich a DAC in-between. It won't be audio over HDMI though, but I'm not sure the Tinkerboard did that either, there was nothing in the little documentation there was to suggest it did, the Pi3 does though.

Anyway, I hope all of you on this forum get out of your Tinkerboards what you hope to get out of them, for me it just isn't the solution, and I don't foresee it ever getting to the point it would. It's very far off, even though it specs well on paper. It's a very complicated board with a very casual backing. Too much doubling up on primary functions without the drivers to make them distinct. Not enough effort or commitment in OS development, by a company with the clout to do so. It could be a great Android platform, it isn't, it could be formidable alternative to the Pi, and really, it isn't. I am not a die hard Pi fan by any means, but in this case it seems like it's the (now) obvious solution.
Skorpio.........I sent you a PM!
It seems you are struggling to deal with the software, but i would says it actually have much fun.

Big Grin  I think Rockchip SOC are more playable than the other SOC, since it get a good opensource support from vendor, .
If you think it's complicated to modify kernel, i would says it won't be easy in other SOC too (like sunxi, amlogic board) (except RPI, because there are many guides for rpi).

Forum Jump:

Users browsing this thread: 1 Guest(s)