Topic: Re: [ANN] Technobit HEX16A2 - 24 Ghs - 50 EUR 48 h shippment - page 47. (Read 104895 times)

Done

technobit.eu/0_1_3.rar

0.1.3 Milestone release - Nanfury support is added with native libusb api. No hid api's are required!
* cgminer ./autegen.sh --enable-hexmineru to add Nanos' manufactured by TechnoBIT known as HEXu
* cgminer --hexmineru-frequency command line to set chip frequency - range 1-62
* cgminer Added to default /etc/config/cgminer  Factory default or web Save+Apply is required for changes to be applied!
* cgminer updated to 3.8.5 rev afe7710858e4ce28bb60f6ae6e167a18d687634f
* cgminer patch to cgminer 3.8.5 rev afe7710858e4ce28bb60f6ae6e167a18d687634f.patch - various code cleanup and optimization needs to be done
* hotplugd added hotplug support for HEXu in /udev/rules.d/01-cgminer.rules. Factory default is required for changes to be applied!
* openwrt updated to r39151

That parameter already right. And i hope you not forget to power on the PSU 

Okay I will try, Thank you very much

Do I have to start cgminer with any parameters?

Ok, let try to compile cgminer 3.8.5 since there is bug with hexminerc driver on 3.8.1 patch.

If that still not work, then try to use powered usb hub.

Edit: Just curious, what is the PSU brand?

ohh sorry,
it is the HEX16A2 connected to raspberry pi with debian wheezy, PSU has 16A at 12V, so it should be enough power available.
I tried with cgminer 3.8.1 with patch, board gets recognized but no hashes

i used this command to start cgminer:
sudo ./cgminer --hexminerc-voltage 900 --hexminerc-options 16:540

Great, Thanks marto 

Yes.
The 5 Volts are delivered by 7805 IC or from USB. Both are possible , depending on 0 Ohm resistor placement.
But from molex We get only 12 Volts, so you can direct solder wires on the connector terminals on the back.

Could you please tell us you configuration & which miner board(HEX16A, HEX16B, HEX16C, Nanofury)  

Hi guys,

I followed the instructions on page 11 to get the board running with cgminer, but no success so far. Also on windows it was not possible for me the get it running.

The board LED's are blinking sometimes , sometimes not. Has anyone any Idea or had the same problems?
At least it got recognized on my raspberry, but thats all so far.

Hi marto,

I had plan to not use the molex connector but will just solder the power cable to molex connector pin.
I checked the schematic & see that +5 VDC from molex isolated by unmounted R42.
So i want to make sure that i only need +12 VDC only right ?

  Thanks for the explanation marto 

So generally as far the heatsink is machined there with 0.5 mm step and this edge of the board( around USB connector) is in the air, and can eventually bend
This cardboard is to prevent from shortening a few 3.3 and 5 V + points from shortening to the heatsink .
This was last minute patch on hex16A and it worked great -  600 units delivered.
So it was left like this as far we are always in a hurry with the next 2 products.
3 revisions of the heatsink and We forget to put a boss there in a suitable place.
Do not worry about fire. This is in the cool part of the board

Martin

Yup, i really love to hear marto explanation since i don't like to open the heatsink to see it 

If mine wasn't just a one-off I wonder if it was a rough-and-ready way of insulating some of the solder points on the bottom from shorting out against the alu heatsink, but from its positioning, I wasn't sure. Maybe Martin can offer some insight?

My two HEX16A board also had that kind of cardboard but i just leave it there since i though maybe there is a reason for that.
Already run for 1 month without problem 

Nice  Compiling cgminer on pi right now.

marto, i hope you not forget to upload TPLINK image 

Hi Marto, glad to see you checking in and keeping an eye out for feedback. If there's no functional difference between the production and test ones (what are the changes you made between test and final, just out of curiosity), I'm fine with the board to be honest now I've tweaked those bits, Just glad to have one here. Who knows, maybe it I can call it a collectors edition further down the line

To re-iterate my main point above, it's a good value product, and for those in doubt, they really are being shipped, just be patient.

Sorry I see that you got one of first 10 samples made on eurocircuts fast(2 day) boards.
Actually these are not supposed to be sent to customers.
The production one have excess termo paste too, we believe better more than less.
Also the hexagon risers are not standard .
In terms of speed it is the same you just got a test board.
We can exchange it for you if you wish.
Best Martin

 

Received my package, ordered on 10th Nov '13, delivered on 17 Dec '13.



In brief, happy with the performance overall. Well packaged, great value product. That said, I have some concerns over obviously rushed assembly, lacking basic quality control.



See image above, there's one screw, nearest the molex connector which has a screw which is blatantly different from all the rest, it does not fit properly, and actually makes multiple points of contact with the conductive elements of components on the board.



This grey bit looked strangely out of place so I thought I better have a quick inspection, considering I've got the replace the terrible screw above. This actually turned out to be a piece of cardboard. WTF.



This is the bit of cardboard, it was flipped over and slightly differently positioned when the heatsink is on. I've no fucking idea why someone thought it was a sensible idea to put a piece of cardboard between the heat sink and the board. Remembering these boards run VERY hot. I've no idea if this would have ever caught fire, but I HIGHLY recommend that everyone else check their units. You can probably also tell there was masses and masses of thermal paste strung across the face side of the board too, and thermal paste finger prints on all the steel hex mount legs, but these didn't really concern me.

I was replacing the fan with an Arctic F9 too, and I noticed that the stock fan that was used also had dust build-up on the leading edge of the blades and had obviously been run, for some time, too. Charitably, this could have been a burn-in test.

I cleaned all the excess thermal paste from the board face. Replaced the stock fan. Removed the cardboard from between the PCB/HSF unit. Fixed the appallingly bad screw. The non-HSF side of the chips measured 62deg C on the settings I'm running, so I added 16 small copper heat sinks, one to each chip. I will possibly add a second fan to the undersides too if I'm concerned about the heat. It's running very stable at 1500/1.050v @ 23.14GHs

All the above considered, it's a great product at the price. It wouldn't be if it cost much more though.

Yes Dude 2.5Gh + per nano ....
Just refresh it frequently...
This is in return for your nice post card. I liked it