[Update: Dec-27 2-13] Hey guys, i have redone the UDOO Power consumption testing. Tests were completed for both the UDOO Dual and UDOO Quad using the Ubuntu 12.04 image, results in the link. An interesting thing was found regarding power, in that upon startup the X Server / desktop runs and churns on CPU cycles and in turn power. If left for about 10 minutes with no desktop activity, this stops and there is a notable drop in power consumption. In most of the testing in the updated tables i waited for that drop to occur when the desktop went to sleep. There is a test at the end with it active as well, just for reference. I have updated the format a bit and will be adding more rows for things like running a SATA SSD, and running the UDOO 7" and 15" LCD Panels to help give further guidance on power usage. [Original Post] Hey guys, since the UDOO has a battery connection, and its likely people are going to be interested in portable UDDO's and more specifically understanding the power usage, i have started to chart some of that data here http://goo.gl/82EAmo This was just something i started last night and tossed the results into a spreadsheet. i would welcome further feedback or other types of loads or stress tests. So far this is Just the UDOO Linux usage, i will be doing more testing with the Arduino side active soon as well. Anyways, just figured i would start sharing the results. Updates * added new testing parameters - - No WIFI, No USB devices, No HDMI - in this test the WiFi adapter was physically removed nothing connected to the UDOO except the serial cable - - No WIFI, No USB devices, with HDMI - in this test still no wifi adapter, serial cable attached, but added HDMI with active monitor - - New Identified drop in power when X Server goes to sleep, took into account in new data - - New Updated results using the Ubuntu 12.04 Image - - New Added tab to the spreadsheet for the UDOO Dual core board
So I gather from your testing that it is possible to enable the number of processors required in order to save power? How is that done? That would be handy when on battery power.
@pjc123. Im not enabling or disabling cores. im just running a single threaded command that will only consume 1 core. i then run it twice to consume 2 cores and so on. you could artificially lower the clock speed to conserve power, something i haven't tested yet but technically feasible Additionally I have updated the table to include power requirements under a few new scenarios. I completed the tests for no WiFi at all levels and also added a test with no wifi and no HDMI See table for new data
It would be interesting if you could do a long-term test to see the statistical data on power usage. (I.e. run it for an hour and take 1 second samples) It would be much more scientifically acceptable if you took more samples, averaged them and took the standard deviation of them, as I can use this to approximate the true mean of power usage for the device. Another interesting scenario I would like to see is with a little processing on the SAM3X side. (Particularly the monitoring of a couple PWM GPIO's and some digital GPIO's.) If you don't have the time for it I understand, it's just something I would like to see happen. (With the plan I have for it, I'm assuming CPU usage will be around 20% and I'll be monitoring 20+ GPIO's, so it would be nice to have those power figures so I can determine what I need to do to make it use less power, if necessary. (Though initially I won't be doing this, simply providing a voltage to the RTC and causing the Udoo to boot when I need it.) Thanks, EBrown
No I don't but a good question. The methodology to determine that would be identify the power regulator and look up that parts datasheet to determine its spec. (devices can normally be trusted to draw a sensible amount of power, but if your thinking of putting a fuse inline I would choose a 3A).
does it actually have the right/matching barrel jack on it? if yes, it probably will tbh, you can get a proper power cord on Amazon for like $5-10 so my 2 cents is to just get one you know will work.. with the right polarity, and size barrel..
Hey guys, i have redone the UDOO Power consumption testing. Tests were completed for both the UDOO Dual and UDOO Quad using the Ubuntu 12.04 image, results in the link. An interesting thing was found regarding power, in that upon startup the X Server / desktop runs and churns on CPU cycles and in turn power. If left for about 10 minutes with no desktop activity, this stops and there is a notable drop in power consumption. In most of the testing in the updated tables i waited for that drop to occur when the desktop went to sleep. There is a test at the end with it active as well, just for reference. I have updated the format a bit and will be adding more rows for things like running a SATA SSD, and running the UDOO 7" and 15" LCD Panels to help give further guidance on power usage. Link to results (same as in OP) http://goo.gl/82EAmo Here are the new Updates - - Identified drop in power when X Server goes to sleep, took into account in new data - - Updated results using the Ubuntu 12.04 Image - - Added tab to the spreadsheet for the UDOO Dual core board I now have the ability to do this and i will try and put together a couple logs. I dont want to do a 60 minute log with 1 sec polling interval for each of my line items, but i think i may do a standard setup, WiFI, HDMI, USB RF KB, and 50% CPU load for boththe dual and Quad boards. I can then put those into .csv files and post them for review. The power usage does fluctuate, but not a great deal.. its usualy +/- 0.005 mA form what i have seen in the brief testing i have done so far I now have a sketch that i can run that will give the SAM3X8 a workout, and will be running this to measure both thermal output and also power consumption. I may be able to get to this soon.. but it is on the table of things for me to do.
Thanks for sharing this data Draco. If you ever get have the chance, it would be interesting to see how the power consumption changes for lower supply voltages (say 6V and 9V). This would be useful when estimating what size of battery supply would be suitable for a particular project.
Somewhere around here are several repetitions of a statement that says that the Udoo can become unstable at lower voltages like that. All you really have to do to find that is multiply amps by 2 or 1.33 respectively. (12/n) Thanks, EBrown
The UDOO can be safely run at 7v or higher. So 9v is ok but 6v is not. Now in reality it will work, but it will likelt not be stable and you may have issues as EBrown noted. Regarding doing the testing, i did just to valudate that you can extrapolate using common formulas and they stay correct.. And you can, i thought i had this noted already in this thread but maybe not. The current always stays the same regardless off the voltage (this is what my initial testing validated) so if you have a lower known voltage you can still calculate the used watts. You will know the voltage, and from my table, the current. From this you can always calculate the watts. This is why i removed my 9v test results and did not test under 9v in the additional tests i did.
I made some own, yet simple tests today with the Android 1.3 image and a Udoo Quad. I used a 12v 2a power connector with the correct barrel jack. My goal was to find out how much power is drawn if Udoo is used mostly as a (portable) gaming station (not a reason with the best reputation, I know, but still a reason good enough for making some tests) but also staying flexible, like browsing the web etc. Just running Android "naked" with nothing connected but the HDMI monitor it was stable at 6.0W. There was no load on the Android system otherwise. Then I connected the LAN cable, which costs around 0.5W-1.2W, depending on the activity. (=> average usage 6.5W) The I added a USB host with 4 ports, which was another 0.1W. (=> 6.6W) Then I added a USB wifi dongle to the USB host for the logitech K400 keyboard which has also a touchpad integrated, which costed 0.8W - 1.6W, depending on if the touchpad was touched or not. (=> 7.5W) Then I added a USB wifi dongle to the USB host for a logitech gamepad, which costed 0.2W. (=> 7.7W) Finally I added a USB stick to the USB host, which costed about 1.0W-1.5W. (=> 9.0W) Then I copied 1.5 GB of data from the USB stick to the SD card and watched the overall Watt usage, which was at 9.2W - 10.2W. Then I removed the USB stick again (for saving power) and started a gaming emulator (for playstation (PSX)) in order to cause some proper load for the system. While running everything the usage was between 9.2W and 9.5W. Then I stopped the emulator (back to 7.7W) and disconnected the LAN cable (=> 7.1W). Then I activated WiFi OnBoard, which costs 0.6W (=> 7.7W). Then I opened a view websites. Depending on what was shown there, like moving live tickers etc, the usage went up to 9.3W. On heavy loaded and moving websites it stayed at about 8.8W to 9.3W, on more "quite" web pages it went down again to 7.7W. So for myself I take it that 9W-10W usage, if the Udoo on Android is really "used", is like realistic. According to those numbers a proper battery can be chosen for being portable.
[quote="Lifeboat_Jim"The methodology to determine that would be identify the power regulator and look up that parts datasheet to determine its spec. (devices can normally be trusted to draw a sensible amount of power, but if your thinking of putting a fuse inline I would choose a 3A).[/quote] From the schematic on page one, there are at least two regulator chips (3.3V, 5.0V) part number: RT8287AZQW. There's a data sheet here: http://www.farnell.com/datasheets/1756300.pdf. They're rated at 4.0 Amps, each and there's thermal shutdown protection.
Good discussion on power uses of UDOO board. Really appreciate the analysis done by @Darco for Ubuntu 12.04 I am working with UDOO advanced plus board with windows 10 installed and a visual studio application running. The application is required to take input from 3 USB COM ports apart from wifi, two displays connected (one display having own power supply while other is taking from USB port). Initially I was working with 12V/3A power supply but some serial ports didn't provided data. I changed the adapter with 4A without any success. What should I do? Does windows requires more power due to more processing requirement compared to Ubuntu?
Use an USB hub with external power supply? The board can only supply limited power to it's peripherals. So if you have extensive requirements use an external power supply to power these peripherals.
By the way, you are reacting on a Udoo dual/quad board topic but you use a Udoo x86. So it is better to start a new topic in that forum if you still have power issues with your Udoo X86 Advanced board.