4 4
TonyMac32

ASUS Tinker Board

Recommended Posts

[update 12/2017 at bottom]

Possibly late, but I would like to put everything we know in one place for anyone who might think of buying this board.

 

DSC_0640-1038x576.jpg

 

Overview:

 

   This is a form factor and (mostly) I/O clone of the Raspberry Pi 3 with a much more powerful quad-core Cortex-A17 Rockchip rk3288.  It supports HDMI 2.0, has 2 GB RAM, Gigabit Ethernet, Wifi and BT on board, etc:  https://www.asus.com/us/Single-Board-Computer/Tinker-Board/

 

   As numerous other sites have covered all the typical performance metrics and extolled the power and so forth of this board, I'm going to go ahead and give you the less exciting information and the tradeoffs/problems.

 

Mainline:

 

Getting the mainline kernel to boot on this machine was pretty straightforward, mainline support for the hardware, including WiFi, makes for less patching and allows a lot of functionality from the mainline kernel without excessive patching.  That said, so far Bluetooth and squashing a reboot bug have not been successful (I'm under the impression the rk3288 was never truly intended to boot solely from external sdmmc devices)

 

Important Hardware Considerations:

 

          Power Solution:  This board is equipped with a micro-USB connector as it's power input.  Micro-USB is only rated for 1.8 Amps, no matter how big the numbers on your power supply are.  It is entirely possible, even likely, that you will hang this board by plugging in peripherals to the USB 2 slots.  Micro-USB is a terrible method of providing power to a single board computer, and is the most serious problem with this device.  This device should be powered via the GPIO header using a filtered supply if you wish to have any semblance of stability.

 

          Heat:  The rk3288 is not a low-power chip, and the heat sink supplied (pictured above), is not adequate for any CPU-intensive activity, quickly throttling performance when it gets too hot. 

 

          USB throughput:  I have not empirically tested this, mostly because it is unnecessary.  For some reason the 4 USB 2.0 ports on the board are all routed through a single USB Hub as on the Raspberry Pi.  Not incredibly useful, other than not having to buy an external hub to make the one exposed USB port into 4.  (unless of course those devices use power, then you need a powered hub anyway)  In case you are wondering, there are 2 USB2 ports available on the SoC, however the dev team for this board decided to dedicate one to an "HD Audio codec" instead of using the dedicated I2S/PCM output to do that job.

 

          Undocumented pins:  The 4 pin header  next to the micro-USB power serve no documented purpose.  One pair is definitely the power button as references in the device tree for the board,  I've determined (and have seen others likewise verify) that the pins closest to the edge are the power button input. The other is not documented at all, and I've not wanted to tempt fate by shorting it out.

 

Software/Support Considerations:

 

          The Documentation for this board is terrible.  Incomplete, non-existent, etc.  The Official ASUS image is a series of workarounds and, until release 1.6, was not properly available to the community.  Even then, development does not appear to be occurring publicly (if it is that means development has stopped).  Rockchip representatives (seemingly not the ones working on the Tinker Board) have at least come forward to provide some helpful hints concerning issues, but ASUS has been entirely silent. 

 

My opinion after use/development:

 

          This is a very powerful board.  Unfortunately I had to build an adapter to power it over GPIO so it would run properly with any moderately demanding USB peripherals, I added a larger heat sink to stabilize the thermal situation, and am currently trying to find a way to get the board to reset properly without using what the Tinker Board source code itself labels a "HACK".  I can not recommend this board to a new buyer.  It's a shame, really, this board had every opportunity to be a really good solution. 

 

  • If the prospective buyer wants nothing more than a 4K media player, there are other options that will serve that niche better, including a small mountain of inexpensive TV boxes. 
  • This board is not ideal for a NAS due to the USB Hub (unless you want to test the limits of the SD card interface). 
  • CPU intensive operations will throttle the device to under 1 GHz with the factory cooler, so without modification you are limited there.
  • Powering peripherals through the board is simply not possible out of the box due to the Micro-USB power solution.  Powering through GPIO is the only sane option.
  • Raspberry Pi compatibility is not absolute.  The GPIO libraries (WiringPi, etc) are not exact, some of the pins serve multiple purposes on the header, etc.
  • This board may be adequate as a small kiosk linux desktop, it is fast enough to provide a snappy interface, and will fit in many of the available cases for the RPi.  I would still recommend GPIO power and probably improved cooling in case a lot of video/etc are needed.

 

[update]  I've been running the Tinker Board as a daily driver for over a week, powering it via micro USB with my normal peripherals (mouse/keybd, wireless active, touchscreen attached)  My findings are what would be expected:

 

  • Power supplied to micro USB port:  5.25 volts 800 - 950 mA "normal" use
    • Playing a Youtube Video (software render) this hits 1.7 Amps
  • Voltage present at Tinker Board USB Host port:  4.7 Volts under "normal" use
    • Playing a Youtube Video this drops to 4.2 Volts, meaning a > 1 Volt drop.

 

Now, you might be saying "I run my Tinker on micro USB all the time and don't have any issues"  You're right, and you're wrong all at once.

 

The processor/RAM use much lower voltages provided by the RK808 PMIC, so the system doesn't fold up and crash when the input voltage gets too low.  HOWEVER, here is a snippet from my dmesg:

 

Spoiler

[224179.241058] usb 1-1.1: new low-speed USB device number 16 using dwc2
[224179.323680] usb 1-1.1: New USB device found, idVendor=0461, idProduct=4d2b
[224179.323693] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[224179.323697] usb 1-1.1: Product: HP Wireless Laser Mini Mouse
[224179.323701] usb 1-1.1: Manufacturer: Primax
[224179.329307] input: Primax HP Wireless Laser Mini Mouse as /devices/platform/ff540000.usb/usb1/1-1/1-1.1/1-1.1:1.0/0003:0461:4D2B.02DE/input/input734
[224179.330436] hid-generic 0003:0461:4D2B.02DE: input,hidraw0: USB HID v1.10 Mouse [Primax HP Wireless Laser Mini Mouse] on usb-ff540000.usb-1.1/input0
[224179.816494] usb 1-1.1: USB disconnect, device number 16
[224180.000057] usb 1-1.1: new low-speed USB device number 17 using dwc2
[224180.477056] usb 1-1.1: device not accepting address 17, error -71
[224180.745058] usb 1-1.1: new low-speed USB device number 18 using dwc2
[224180.827684] usb 1-1.1: New USB device found, idVendor=0461, idProduct=4d2b
[224180.827696] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[224180.827701] usb 1-1.1: Product: HP Wireless Laser Mini Mouse
[224180.827705] usb 1-1.1: Manufacturer: Primax
[224180.832362] input: Primax HP Wireless Laser Mini Mouse as /devices/platform/ff540000.usb/usb1/1-1/1-1.1/1-1.1:1.0/0003:0461:4D2B.02DF/input/input735
[224180.832638] hid-generic 0003:0461:4D2B.02DF: input,hidraw0: USB HID v1.10 Mouse [Primax HP Wireless Laser Mini Mouse] on usb-ff540000.usb-1.1/input0
[224181.352236] usb 1-1.1: USB disconnect, device number 18
[224181.532063] usb 1-1.1: new low-speed USB device number 19 using dwc2
[224181.809039] usb 1-1.1: device descriptor read/64, error -71
[224181.999684] usb 1-1.1: New USB device found, idVendor=0461, idProduct=4d2b
[224181.999696] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[224181.999701] usb 1-1.1: Product: HP Wireless Laser Mini Mouse
[224181.999705] usb 1-1.1: Manufacturer: Primax
[224182.004183] input: Primax HP Wireless Laser Mini Mouse as /devices/platform/ff540000.usb/usb1/1-1/1-1.1/1-1.1:1.0/0003:0461:4D2B.02E0/input/input736
[224182.004405] hid-generic 0003:0461:4D2B.02E0: input,hidraw0: USB HID v1.10 Mouse [Primax HP Wireless Laser Mini Mouse] on usb-ff540000.usb-1.1/input0
[224182.632104] usb 1-1.1: USB disconnect, device number 19
[224182.816050] usb 1-1.1: new low-speed USB device number 20 using dwc2
[224183.089049] usb 1-1.1: device descriptor read/64, error -71
[224183.279450] usb 1-1.1: New USB device found, idVendor=0461, idProduct=4d2b
[224183.279458] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[224183.279462] usb 1-1.1: Product: HP Wireless Laser Mini Mouse
[224183.279466] usb 1-1.1: Manufacturer: Primax
[224183.283977] input: Primax HP Wireless Laser Mini Mouse as /devices/platform/ff540000.usb/usb1/1-1/1-1.1/1-1.1:1.0/0003:0461:4D2B.02E1/input/input737
[224183.284613] hid-generic 0003:0461:4D2B.02E1: input,hidraw0: USB HID v1.10 Mouse [Primax HP Wireless Laser Mini Mouse] on usb-ff540000.usb-1.1/input0
[224183.656348] usb 1-1.1: USB disconnect, device number 20
[224183.866054] usb 1-1.1: new low-speed USB device number 21 using dwc2
[224184.349041] usb 1-1.1: device not accepting address 21, error -71
[224184.617033] usb 1-1.1: new low-speed USB device number 22 using dwc2
[224184.700159] usb 1-1.1: New USB device found, idVendor=0461, idProduct=4d2b
[224184.700168] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[224184.700172] usb 1-1.1: Product: HP Wireless Laser Mini Mouse
[224184.700176] usb 1-1.1: Manufacturer: Primax
[224184.704646] input: Primax HP Wireless Laser Mini Mouse as /devices/platform/ff540000.usb/usb1/1-1/1-1.1/1-1.1:1.0/0003:0461:4D2B.02E2/input/input738
[224184.705304] hid-generic 0003:0461:4D2B.02E2: input,hidraw0: USB HID v1.10 Mouse [Primax HP Wireless Laser Mini Mouse] on usb-ff540000.usb-1.1/input0
[225358.440414] usb 1-1.1: USB disconnect, device number 22
[225358.626474] usb 1-1.1: new low-speed USB device number 23 using dwc2
[225358.711827] usb 1-1.1: New USB device found, idVendor=0461, idProduct=4d2b
[225358.711835] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[225358.711840] usb 1-1.1: Product: HP Wireless Laser Mini Mouse
[225358.711843] usb 1-1.1: Manufacturer: Primax
[225358.716409] input: Primax HP Wireless Laser Mini Mouse as /devices/platform/ff540000.usb/usb1/1-1/1-1.1/1-1.1:1.0/0003:0461:4D2B.02E3/input/input739
[225358.716694] hid-generic 0003:0461:4D2B.02E3: input,hidraw0: USB HID v1.10 Mouse [Primax HP Wireless Laser Mini Mouse] on usb-ff540000.usb-1.1/input0
[229843.816078] usb 1-1.1: USB disconnect, device number 23
[260964.116257] usb 1-1.1: new low-speed USB device number 24 using dwc2
[260964.198547] usb 1-1.1: New USB device found, idVendor=0461, idProduct=4d2b
[260964.198558] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[260964.198562] usb 1-1.1: Product: HP Wireless Laser Mini Mouse
[260964.198566] usb 1-1.1: Manufacturer: Primax
[260964.204243] input: Primax HP Wireless Laser Mini Mouse as /devices/platform/ff540000.usb/usb1/1-1/1-1.1/1-1.1:1.0/0003:0461:4D2B.02E4/input/input740
[260964.205837] hid-generic 0003:0461:4D2B.02E4: input,hidraw0: USB HID v1.10 Mouse [Primax HP Wireless Laser Mini Mouse] on usb-ff540000.usb-1.1/input0

 

What you're seeing here is my little wireless mouse receiver giving up the ghost because of voltage starvation.  More or less, when I get these voltage dips, anything that needs 5 volts (like USB peripherals, say that external HDD, webcam, card reader, mouse) shut down and/or could be damaged/corrupted.  I have not had a single system failure, however were I to be reading/writing external media (or running this off of a flash drive for some reason) I'd have experienced some real problems.

Share this post


Link to post
Share on other sites
On 6/30/2017 at 0:10 AM, TonyMac32 said:

This device should be powered via the GPIO header using a filtered supply if you wish to have any semblance of stability.

Thanks for the summary:

 

Ok, so is there any over-the-counter GPIO Power Supplies that one can purchase ?, -instead of me trying to make my own plug-n-pray one.

By "filtered" power supply does that mean a switching power supply.?

 Not to mention the plethora of other glaring issues with this ASUS Tinker Board, <- can u believe it "Asus" -you'd think they knew how to make boards by now eh. 

 

Or, (being a new user to SBC), should I wait for the Tinker 2 Board, when hopefully everything should be fixed?.

 

Thanks for any advice folks.

 

 

Share this post


Link to post
Share on other sites

I've been toying with the idea of designing a simple board and publishing the artwork/BOM for it, but if you look around, one for the raspberry pi should be available, look for DIN rail pi accessories, some of them have regulator hats.

 

As far as filtered (I forgot to hit that point), I am referring to a capacitor/zener diode combo to protect against switching noise and over voltage. 

Share this post


Link to post
Share on other sites

Updated OP with "real world" use findings.  System stability with micro USB has not been a problem as long as SD is used for file system thanks to the RK808 buck converters and LDO's, however anything attached to USB can fail/detach-reattach depending on system load.  I would expect longer term testing could show some HDMI issues as well due to 5V requirements.

Share this post


Link to post
Share on other sites

I just want to throw in my two cents about the alleged Tinkerboard power issues. In my experience, I first used it with an average 2.5A power source, and a very bad cable, and I got frequent random reboots. But after I changed the microusb cable for a good one, while keeping the same average 2.5A source, I have never had any problem. I have streamed video for hours, used intensively Wifi and Bluetooth speakers, and not even a single crash. Notice I am also powering a small 200mA fan. Of course, it might be different if you hook up some hungry USB peripherals (such as external hard disks).

 

About performance: with the stock heatsink, and the board enclosed in a case, with no fan, stressing it with cpuburn-A9, it will throttle down to an average of around 1 Ghz, occasionally dropping to around 800 Mhz. But simply using one of those Raspberry Pi cases with a small fan, it will keep the maximum 1.8Ghz for long periods of time, hardly throttling down to 1.6 Ghz, and very seldom dropping to 1.4 Ghz. Notice these are under the heaviest stress conditions, which you will not reach under normal use.

 

All these tests done with the latest ASUS 4.4.71+ kernel from TinkerOS 2.0.4. I don't use Armbian in the Tinker, even though I think the Armbian OS is way superior, because the kernel for Thinkerboard still has some issues with audio, and no graphic accel.

Share this post


Link to post
Share on other sites

 

2 hours ago, JMCC said:

I have never had any problem

Congratulations!

 

That said, the laws of physics still hold, and you missed providing the relevant information concerning your supply of choice:  The voltage of the supply.  Also, see:

 

That post contains empirical data and observations while running the Tinker Board.   If you wish to continue this discussion, we can do it in the review thread as it is off-topic for this thread.  I would actually recommend that @chwe or another moderator make that move of these last two (or more as deemed prudent) posts to that thread if possible, since we're far off topic of RK3328 vs RK3288 and the power discussion may be enlightening to others. 

 

From that post:

 

On 30.6.2017 at 6:10 AM, TonyMac32 said:

Now, you might be saying "I run my Tinker on micro USB all the time and don't have any issues"  You're right, and you're wrong all at once.

Remember that Armbian gives users the flexibility to run off of USB (after booting SD of course), and that use case would be a complete failure without spending some time seeing warnings like those above.  I also am referencing "as packaged" when I discuss the "factory" heat sink, since obviously with fans/etc it will perform better.  I can operate at a continuous 1.5 GHz with 30 second or so "sprints" at 1.6 when it cools slightly at full load, simply by attaching a superior low-cost heat sink, something I would highly recommend to anyone who did not want to suffer immediate and severe throttling (45 to 55% throttling is severe).  Add a fan and it may never throttle under normal use.

Edited by chwe
it is a particular honour for me to do it for you... ​​​​​​​:D

Share this post


Link to post
Share on other sites
1 hour ago, TonyMac32 said:

 

 you missed providing the relevant information concerning your supply of choice:  The voltage of the supply.

 

You're right, I was taking for granted that the power source voltage is 5V as recommended by the manufacturer, but I should have specified it.

 

I already knew about those posts you mention, and have read them several times. I am sure your findings are solid and well-founded. I'm just an amateur, and I won't dare to question what you pro guys state after many hours of research and tests.

 

But, at the same time, I want to contribute with a real-use experience that is not as bad as those numbers may suggest. I have seen a growing panic in several posts in different forums, as if the TB was useless without GPIO powering. In other words: when I ordered my Tinker, and started googling for info about it while waiting for it, all these posts here made me think I had wasted my money in a useless piece of scrap. But my experience is that it can be ideal for certain projects, where you need to combine a powerful SoC with good graphics, together with the flexibility of a SBC (as is my case: 24/7 playing video and PPT presentations, with programmed schedule, and communicating through GPIO to RS232 HDMI switch for a live camera feed at certain times).

 

So, totally agreed that the TB is not the best choice for any project that requires USB HDD's (I got an Opi+2e precisely for that). But in my experience, it can be a very good choice for other kind of projects, as long as you choose a good microusb cable and put a small fan on the case. As i said, not even a single problem in many hours of testing.

 

P.S.: Is it possible that ASUS has improved the quality of some components in recent batches (I got my TB in December), so now it performs better than the one you used for your tests? Or maybe they have introduced some tweaks in recent kernels, to cope with the power issues? I don't know, just asking.

Share this post


Link to post
Share on other sites
1 hour ago, JMCC said:

In other words: when I ordered my Tinker, and started googling for info about it while waiting for it, all these posts here made me think I had wasted my money in a useless piece of scrap

 

Well, some would agree with that assessment, as far as I'm concerned it is disappointing, but not useless.  Were it a completely useless board I would not have put the time into supporting it.  The danger here is that people use Armbian for servers and data-integrity-sensitive applications, and the Tinker simply can't fill that role without a lot of care.  

 

As far as my background, I have a degree in electrical engineering and design automotive electronics (safety related devices).  When a board has 4 USB 2 ports on it, I expect them to meet the specification.  If it chooses a power plug, it must be one capable of supplying the board at full load (so in this case board + 2 Amps of USB load).  No board that I know of meets that requirement, so the Tinker is not alone.  So expectations were extremely high, and the board didn't quite deliver.

 

1 hour ago, JMCC said:

You're right, I was taking for granted that the power source voltage is 5V as recommended by the manufacturer, but I should have specified it.

 

Well, at 5.0 Volts my stability testing repeatedly failed.  That's common for RPi, 'x'Pi, etc etc etc.  The voltage drop due to the micro-USB assures that.  So I use 5.25 volts and sit *just* above the minimum allowed supply voltage for the USB peripherals.  

 

1 hour ago, JMCC said:

P.S.: Is it possible that ASUS has improved the quality of some components in recent batches

 

I would assume not, given the issue is a specification problem, not just a component source issue.  I believe the parts are of sufficient quality, this is an ASUS product after all, it's just that the RPi foundation set a rather unfortunate standard, and their boards have a lot of workarounds because of it, which is replicated in the newer ASUS board. 

 

As you see in December I updated with more recent info, using the device for 2-3 weeks as my desktop, noting that system stability was not an issue, as long as the filesystem stayed on SD.  

Share this post


Link to post
Share on other sites

@Meta I'd be careful with the term "as advertised".  ASUS sells it as a Pi killer, and it is, even with the drawbacks.  It is not a workstation or a server, it's a multimedia desktop / thingy. 

 

   My biggest grief with the design concerns the decisions that could have gone differently to make it better at the things a typical Armbian user wants, like the soundcard having a dedicated USB while everything else has to share, no dedicated power option besides the microUSB, etc.  I'm waiting for even one company to provide a board with a header for powering the board without using the microUSB, so far nothing.

 

Now, if you want faster board there aren't a lot of options unless you're doing something that ARMv8 has extra support for that will overcome the slowness of an a53 vs the a17.  The Odroid XU4 has 4 A15 cores and I believe performs similarly, and has a bonus of 4 small cores as well.  In more practical terms, it also has USB 3.0.  Downsides are GPIO are 1.8 volt, cases are expensive or cheap feeling, and the processor goes from 40 to 80 C in about 1 second, being very hard to cool (I liquid cooled mine just for kicks).  It is what runs my Plex server.  

 

Otherwise I think even the 3399 is only questionably faster in certain use cases, only having 2 high-powered cores and 4 "slow" ones.  

 

With the new "S" model there is voltage detection to see if the supply is dropping.  I need to look closer at it, it may be possible to avoid writing to external media in that situation/etc, helping to safegaurd data.  Otherwise, like with all other boards, power the drive separately from the tinker.  If you want to do industrial grade things, get a chassis supply with at least 10 amps available and power all the components accordingly.  

I want to design a power supply hat for pi-factor boards, focusing on cheap + effective, if there's interest I can give it some priority.  It would ideally expose barrel jack, and a terminal block, and possibly put a USB A to power externals.  Like the Pi touchscreen.

Share this post


Link to post
Share on other sites

@TonyMac32 Fair enough, I guess my grievance is more with it lacking the polish in many of the areas you're talking about.

 

On the note of the power options, I've been looking for ways to power various Orange Pi devices via GPIO instead of their barrel (with the goal of being able to power it from a USB hub instead of a wall wart), and it made me curious about the TinkerBoard too.  I think I've come up with an okay method for the OPi, using a little circuit to filter out noise from USB power sources, and it would appear the same circuit would work on the TinkerBoard as well. Would that give it any more stability, powering the USB host through the rails instead of drawing it all from the micro-USB input?

 

The power HAT idea you mentioned sounds like it would be really useful for a lot of situations. Are you thinking of just posting schematics, or actually manufacturing it yourself?

Share this post


Link to post
Share on other sites
5 hours ago, Meta said:

The power HAT idea you mentioned sounds like it would be really useful for a lot of situations. Are you thinking of just posting schematics, or actually manufacturing it yourself?

 

That is a very good question.  I am going to make a few test runs, meanwhile I'm trying to explore options about building.  The problem(s) I have are fairly predictable: 

  • Small market
  • No logistics capabilities.  Shipping is $$$$ no one wants to spend.
  • Time to invest in new thingies. 

If anyone knows of a place where you can make designs available for sale that would be awesome, I'm looking at Sparkfun but seeing an issue with Eagle being their EDA of choice, and Eagle now being a subscription thanks to Autodesk.  And they'd probably not touch a GPIO power hat for obvious reasons.  I'll probably have "do not use" written into the silkscreen, and the bottom copper layer.  :P  I am reviewing open source hardware licensing as well, my intention is to "make cool stuff", not to "hoard cool stuff"

Share this post


Link to post
Share on other sites

I put in an order for prototype boards today (3), and I have components coming as well, I think Fab is scheduled for the 7th (could have been sooner but I needed at least 2 oz copper).  In any case, if it works nicely I'll upload the whole mess to GitHub with an open source license on it and maybe put out a hat for appreciative people to drop change into so I can buy some beer.  ;-). There are some tweaks I want to make, but that's for later or else nothing will ever be built...

 

PCB's can be shipped in the small bubble envelopes, so there could be some options there.  

Share this post


Link to post
Share on other sites

Hi all.

 

Sharing my somewhat peculiar use case here.

 

I am a control freak. Yes. I don't like Carrier-provided modems controlling my network. Or a off-the-shelf (read: TP-Link, Netgear, Linksys, etc.) Router/Wifi too.

 

And I also keep tab of the power consumption, because, the lower the power consumption, the longest the endurance of my UPS ;-)

 

I elected the Tinker Board as my solution after some research. It's been a month since it went 'production' and I have to say that I am a hundred percent happy with the stability and performance.

 

A little more about my scenario.

 

The Tinker Board is mainly my Router / NAT device. I only use the on-board Gigabit interface: I don't use any other external interfaces (USB Ethernet Dongles) or the WiFi / Bluetooth devices: I relinquish my Wireless ops to a Ubiquiti UAP-AC-Pro.

 

The 'magic' of the NIC partitioning is achieved by using VLAN tag. It creates several other different and independent interfaces while using the same underlying physical device. Do not mistake with Interface Aliases - which are neither isolated nor independent.

 

VLAN tag requires a capable switch, so it can tag and forward the traffic accordingly to its respective ports.

 

After running synthetic tests, I was able to pull around 930 Mbps sustained NATed throughput.

 

The Tinker Board persistence resides in two external hard drives; one is a small factor (slim) 2½" laptop 500 GB HDD, powered by its caddy and a standard 3½" 6 TB hard drive as warm storage. Both caddies are the same, USB 3 capable and attached to a powered USB hub, so no power is drawn from the Tinker Board during its run time. As a side note, the 500 GB HDD spins all the time, while the 6 TB is spun only on demand, and spins down after 10 minutes if no activity.

 

The boot is provided by a 256 MB (grins) SD card, which refers the rest of the boot process to the 500 GB HDD.

 

The USB hub also has a Huawei GSM Dongle, in order to provide a cellphone gateway to my Asterisk VoIP infrastructure.

 

Besides VoIP and NAT, the Tinker Board also runs Samba services and a DLNA media server, VPN, DHCP, DNS, HTTP, git, rsync (in daemon mode, because SSH is... costly!) and UPS monitoring.

 

The memory figures usually stays behind 400 MB of RAM, as well its average load average below 0.2. VoIP is smooth, no cracks / lag, and the Samba service throughput is limited by the USB 2.0 bus bandwidth.

 

It currently runs a stripped out to bare kernel, tailored to my needs and it is 100% headless.

 

It is powered by a 1A power supply, using the loathed micro USB port and runs with the shipped heat sink. Temperature is monitored and it is always behind 50 Celsius. Looks like graph processing is the villain here when speaking about power consumption.

 

Since I tally the power consumption (thanks to the UPS), the Tinker Board is using the same amount of power as the Raspberry Pi used to: Thrice the computing power with the same power consumption.

 

I am positive that this scenario might not tailor the general use of the board. But it stated here, in case anyone has a similar use case ;-)

 

If interested or have any question don't hesitate to get in touch.

 

- Rodrigo.

Share this post


Link to post
Share on other sites

 

Spoiler
On 6/30/2017 at 6:10 AM, TonyMac32 said:

[update 12/2017 at bottom]

Possibly late, but I would like to put everything we know in one place for anyone who might think of buying this board.

 

DSC_0640-1038x576.jpg

 

Overview:

 

   This is a form factor and (mostly) I/O clone of the Raspberry Pi 3 with a much more powerful quad-core Cortex-A17 Rockchip rk3288.  It supports HDMI 2.0, has 2 GB RAM, Gigabit Ethernet, Wifi and BT on board, etc:  https://www.asus.com/us/Single-Board-Computer/Tinker-Board/

 

   As numerous other sites have covered all the typical performance metrics and extolled the power and so forth of this board, I'm going to go ahead and give you the less exciting information and the tradeoffs/problems.

 

Mainline:

 

Getting the mainline kernel to boot on this machine was pretty straightforward, mainline support for the hardware, including WiFi, makes for less patching and allows a lot of functionality from the mainline kernel without excessive patching.  That said, so far Bluetooth and squashing a reboot bug have not been successful (I'm under the impression the rk3288 was never truly intended to boot solely from external sdmmc devices)

 

Important Hardware Considerations:

 

          Power Solution:  This board is equipped with a micro-USB connector as it's power input.  Micro-USB is only rated for 1.8 Amps, no matter how big the numbers on your power supply are.  It is entirely possible, even likely, that you will hang this board by plugging in peripherals to the USB 2 slots.  Micro-USB is a terrible method of providing power to a single board computer, and is the most serious problem with this device.  This device should be powered via the GPIO header using a filtered supply if you wish to have any semblance of stability.

 

          Heat:  The rk3288 is not a low-power chip, and the heat sink supplied (pictured above), is not adequate for any CPU-intensive activity, quickly throttling performance when it gets too hot. 

 

          USB throughput:  I have not empirically tested this, mostly because it is unnecessary.  For some reason the 4 USB 2.0 ports on the board are all routed through a single USB Hub as on the Raspberry Pi.  Not incredibly useful, other than not having to buy an external hub to make the one exposed USB port into 4.  (unless of course those devices use power, then you need a powered hub anyway)  In case you are wondering, there are 2 USB2 ports available on the SoC, however the dev team for this board decided to dedicate one to an "HD Audio codec" instead of using the dedicated I2S/PCM output to do that job.

 

          Undocumented pins:  The 4 pin header  next to the micro-USB power serve no documented purpose.  One pair is definitely the power button as references in the device tree for the board,  I've determined (and have seen others likewise verify) that the pins closest to the edge are the power button input. The other is not documented at all, and I've not wanted to tempt fate by shorting it out.

 

Software/Support Considerations:

 

          The Documentation for this board is terrible.  Incomplete, non-existent, etc.  The Official ASUS image is a series of workarounds and, until release 1.6, was not properly available to the community.  Even then, development does not appear to be occurring publicly (if it is that means development has stopped).  Rockchip representatives (seemingly not the ones working on the Tinker Board) have at least come forward to provide some helpful hints concerning issues, but ASUS has been entirely silent. 

 

My opinion after use/development:

 

          This is a very powerful board.  Unfortunately I had to build an adapter to power it over GPIO so it would run properly with any moderately demanding USB peripherals, I added a larger heat sink to stabilize the thermal situation, and am currently trying to find a way to get the board to reset properly without using what the Tinker Board source code itself labels a "HACK".  I can not recommend this board to a new buyer.  It's a shame, really, this board had every opportunity to be a really good solution. 

 

  • If the prospective buyer wants nothing more than a 4K media player, there are other options that will serve that niche better, including a small mountain of inexpensive TV boxes. 
  • This board is not ideal for a NAS due to the USB Hub (unless you want to test the limits of the SD card interface). 
  • CPU intensive operations will throttle the device to under 1 GHz with the factory cooler, so without modification you are limited there.
  • Powering peripherals through the board is simply not possible out of the box due to the Micro-USB power solution.  Powering through GPIO is the only sane option.
  • Raspberry Pi compatibility is not absolute.  The GPIO libraries (WiringPi, etc) are not exact, some of the pins serve multiple purposes on the header, etc.
  • This board may be adequate as a small kiosk linux desktop, it is fast enough to provide a snappy interface, and will fit in many of the available cases for the RPi.  I would still recommend GPIO power and probably improved cooling in case a lot of video/etc are needed.

 

[update]  I've been running the Tinker Board as a daily driver for over a week, powering it via micro USB with my normal peripherals (mouse/keybd, wireless active, touchscreen attached)  My findings are what would be expected:

 

  • Power supplied to micro USB port:  5.25 volts 800 - 950 mA "normal" use
    • Playing a Youtube Video (software render) this hits 1.7 Amps
  • Voltage present at Tinker Board USB Host port:  4.7 Volts under "normal" use
    • Playing a Youtube Video this drops to 4.2 Volts, meaning a > 1 Volt drop.

 

Now, you might be saying "I run my Tinker on micro USB all the time and don't have any issues"  You're right, and you're wrong all at once.

 

The processor/RAM use much lower voltages provided by the RK808 PMIC, so the system doesn't fold up and crash when the input voltage gets too low.  HOWEVER, here is a snippet from my dmesg:

 

  Reveal hidden contents



[224179.241058] usb 1-1.1: new low-speed USB device number 16 using dwc2
[224179.323680] usb 1-1.1: New USB device found, idVendor=0461, idProduct=4d2b
[224179.323693] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[224179.323697] usb 1-1.1: Product: HP Wireless Laser Mini Mouse
[224179.323701] usb 1-1.1: Manufacturer: Primax
[224179.329307] input: Primax HP Wireless Laser Mini Mouse as /devices/platform/ff540000.usb/usb1/1-1/1-1.1/1-1.1:1.0/0003:0461:4D2B.02DE/input/input734
[224179.330436] hid-generic 0003:0461:4D2B.02DE: input,hidraw0: USB HID v1.10 Mouse [Primax HP Wireless Laser Mini Mouse] on usb-ff540000.usb-1.1/input0
[224179.816494] usb 1-1.1: USB disconnect, device number 16
[224180.000057] usb 1-1.1: new low-speed USB device number 17 using dwc2
[224180.477056] usb 1-1.1: device not accepting address 17, error -71
[224180.745058] usb 1-1.1: new low-speed USB device number 18 using dwc2
[224180.827684] usb 1-1.1: New USB device found, idVendor=0461, idProduct=4d2b
[224180.827696] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[224180.827701] usb 1-1.1: Product: HP Wireless Laser Mini Mouse
[224180.827705] usb 1-1.1: Manufacturer: Primax
[224180.832362] input: Primax HP Wireless Laser Mini Mouse as /devices/platform/ff540000.usb/usb1/1-1/1-1.1/1-1.1:1.0/0003:0461:4D2B.02DF/input/input735
[224180.832638] hid-generic 0003:0461:4D2B.02DF: input,hidraw0: USB HID v1.10 Mouse [Primax HP Wireless Laser Mini Mouse] on usb-ff540000.usb-1.1/input0
[224181.352236] usb 1-1.1: USB disconnect, device number 18
[224181.532063] usb 1-1.1: new low-speed USB device number 19 using dwc2
[224181.809039] usb 1-1.1: device descriptor read/64, error -71
[224181.999684] usb 1-1.1: New USB device found, idVendor=0461, idProduct=4d2b
[224181.999696] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[224181.999701] usb 1-1.1: Product: HP Wireless Laser Mini Mouse
[224181.999705] usb 1-1.1: Manufacturer: Primax
[224182.004183] input: Primax HP Wireless Laser Mini Mouse as /devices/platform/ff540000.usb/usb1/1-1/1-1.1/1-1.1:1.0/0003:0461:4D2B.02E0/input/input736
[224182.004405] hid-generic 0003:0461:4D2B.02E0: input,hidraw0: USB HID v1.10 Mouse [Primax HP Wireless Laser Mini Mouse] on usb-ff540000.usb-1.1/input0
[224182.632104] usb 1-1.1: USB disconnect, device number 19
[224182.816050] usb 1-1.1: new low-speed USB device number 20 using dwc2
[224183.089049] usb 1-1.1: device descriptor read/64, error -71
[224183.279450] usb 1-1.1: New USB device found, idVendor=0461, idProduct=4d2b
[224183.279458] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[224183.279462] usb 1-1.1: Product: HP Wireless Laser Mini Mouse
[224183.279466] usb 1-1.1: Manufacturer: Primax
[224183.283977] input: Primax HP Wireless Laser Mini Mouse as /devices/platform/ff540000.usb/usb1/1-1/1-1.1/1-1.1:1.0/0003:0461:4D2B.02E1/input/input737
[224183.284613] hid-generic 0003:0461:4D2B.02E1: input,hidraw0: USB HID v1.10 Mouse [Primax HP Wireless Laser Mini Mouse] on usb-ff540000.usb-1.1/input0
[224183.656348] usb 1-1.1: USB disconnect, device number 20
[224183.866054] usb 1-1.1: new low-speed USB device number 21 using dwc2
[224184.349041] usb 1-1.1: device not accepting address 21, error -71
[224184.617033] usb 1-1.1: new low-speed USB device number 22 using dwc2
[224184.700159] usb 1-1.1: New USB device found, idVendor=0461, idProduct=4d2b
[224184.700168] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[224184.700172] usb 1-1.1: Product: HP Wireless Laser Mini Mouse
[224184.700176] usb 1-1.1: Manufacturer: Primax
[224184.704646] input: Primax HP Wireless Laser Mini Mouse as /devices/platform/ff540000.usb/usb1/1-1/1-1.1/1-1.1:1.0/0003:0461:4D2B.02E2/input/input738
[224184.705304] hid-generic 0003:0461:4D2B.02E2: input,hidraw0: USB HID v1.10 Mouse [Primax HP Wireless Laser Mini Mouse] on usb-ff540000.usb-1.1/input0
[225358.440414] usb 1-1.1: USB disconnect, device number 22
[225358.626474] usb 1-1.1: new low-speed USB device number 23 using dwc2
[225358.711827] usb 1-1.1: New USB device found, idVendor=0461, idProduct=4d2b
[225358.711835] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[225358.711840] usb 1-1.1: Product: HP Wireless Laser Mini Mouse
[225358.711843] usb 1-1.1: Manufacturer: Primax
[225358.716409] input: Primax HP Wireless Laser Mini Mouse as /devices/platform/ff540000.usb/usb1/1-1/1-1.1/1-1.1:1.0/0003:0461:4D2B.02E3/input/input739
[225358.716694] hid-generic 0003:0461:4D2B.02E3: input,hidraw0: USB HID v1.10 Mouse [Primax HP Wireless Laser Mini Mouse] on usb-ff540000.usb-1.1/input0
[229843.816078] usb 1-1.1: USB disconnect, device number 23
[260964.116257] usb 1-1.1: new low-speed USB device number 24 using dwc2
[260964.198547] usb 1-1.1: New USB device found, idVendor=0461, idProduct=4d2b
[260964.198558] usb 1-1.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[260964.198562] usb 1-1.1: Product: HP Wireless Laser Mini Mouse
[260964.198566] usb 1-1.1: Manufacturer: Primax
[260964.204243] input: Primax HP Wireless Laser Mini Mouse as /devices/platform/ff540000.usb/usb1/1-1/1-1.1/1-1.1:1.0/0003:0461:4D2B.02E4/input/input740
[260964.205837] hid-generic 0003:0461:4D2B.02E4: input,hidraw0: USB HID v1.10 Mouse [Primax HP Wireless Laser Mini Mouse] on usb-ff540000.usb-1.1/input0

 

What you're seeing here is my little wireless mouse receiver giving up the ghost because of voltage starvation.  More or less, when I get these voltage dips, anything that needs 5 volts (like USB peripherals, say that external HDD, webcam, card reader, mouse) shut down and/or could be damaged/corrupted.  I have not had a single system failure, however were I to be reading/writing external media (or running this off of a flash drive for some reason) I'd have experienced some real problems.

 

 

This forum just blew my mind wide open 

 

I'm sorry peeps I'm a SBC virgin I thought I could just pick any SBC and throw a OS like the android Firetv on it and it would be all plug and play 

 

Could someone please help me picking the best equipment for this kinda job 

 

Yes yes yes I know what you going say why don't you just buy you another fire stick, but what's the fun in that 

 

 

Share this post


Link to post
Share on other sites

Well, as far as that goes, the Tinker is one of your better bets at present, thanks to the work of @JMCC.  Using his "how-to's", you'll have multimedia capabilities on Linux in short order.  Now, if you want Android, I believe that is also supported, but only by ASUS, not by us. 

Share this post


Link to post
Share on other sites
2 hours ago, TonyMac32 said:

Well, as far as that goes, the Tinker is one of your better bets at present, thanks to the work of @JMCC.  Using his "how-to's", you'll have multimedia capabilities on Linux in short order.  Now, if you want Android, I believe that is also supported, but only by ASUS, not by us. 

Hi there 

 

Can you advise where I could purchase a Tinker board S please...? 

Share this post


Link to post
Share on other sites
2 minutes ago, Gaaht said:

Can you advise where I could purchase a Tinker board S please...? 

That I do not know, I do not believe they have been officially released yet.  If you must have the eMMC, then wait, otherwise the standard one is fine.  Now, make sure you read everything you possibly can about things like Amazon, netflix,etc.  a lot of devices have strange behaviors in this arena, and I am no expert (I  have only had 3 TV boxes, one that was a complete failure (s905x "Trongle"), and the other 2 have... quirks... (Z28 Pro and Ugoos UT3s)

Share this post


Link to post
Share on other sites
5 hours ago, Gaaht said:

Can you advise where I could purchase a Tinker board S please...? 

Netflix will work OK in the browser (with some tweaks I am about to release), Amazon won't. I know there are Kodi add-ons for both, but I haven't tested them.

 

About the best place to buy the TB, if you can buy in amazon.co.uk they put it on sale for a very good price from time to time. I got mine there for £46 before last Christmas, and now it is for £56, with free shipping within the UK. In the USA, ameridroid.com has the best price ($59). And also, they have one of the funniest testimonials I have ever read:

Spoiler

"I really appreciate your fast shipping. I got the order even before I could come up with a story for my wife about why I needed another SBC!! That's impressive." -Bill

LOL :D

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
4 4