Oh Damn.....one of those....do you have another fan with the same connector that you can use to compare?  Otherwise try to look up the pinout of a motherboard fan header and that should tell you.  I'd do it but I'm on my phone and it's kind of a pain in the ass.

make sure the correct pins line up on your fan.  Last time I checked the fan plug the pins were 'reversed' as opposed to a standard 3-pin connection.  Match the red and black wires, yellow will have no connection.

Has anyone gotten CGMiner 4.5 to run with the Gridseed A1 units?  I see A1 files in it, I've compiled it with the A1 support....but it never detects the A1 hardware when I run it. 

command line:  sudo ./cgminer -o pooladdress -u username -p password --bitmine-a1-options 16000:800000:2000

The unit has a built in rpi and the A1 controller board is connected to the GPIO pins.  I've tried running on the image that came with the hardware, as well as compiling on a generic raspbian image....same result....it doesn't find the A1 hardware. 

What am I missing?

The hardware I'm speaking of is this one in particular:

Here is the image.  i forgot to take out my information so if you really feel like using it to mine feel free.  otherwise just go to the settings page and set things up the way you want.

some notes:
no failover pool support yet
i added fields for clock speed (max is 1000mhz)
added a field for spi speed
the image waits 60 seconds after bootup to start mining (this is the way the default crap came)
it's all in english now
i added the current clock speed to the status page.


as it is now with the settings in it im getting 1.1ths poolside with it.

https://www.dropbox.com/s/0njs18kdgx2guda/Stock_English_Fixed.7z

Obviously every miner is going to be a little different, but I would say 1.35v - 1.42v would do the trick.  I believe mine are set to 1.41v.  Less than that and they gave more errors, more than that gave quickly diminishing returns so I backed it down to 1.4v-1.41v.

I use an Rpi...not sure if you can use a windows box to mine with it, but rpi is best for the job I think.

It's compressing and will be uploaded immediately after that.  I'll follow up  with a link when it's done.

You enter that info into your browser...Not ssh.

If it's using the standard chinese shitware that came on mine, you'll need to SSH into the pi and edit the files in /www/miner/conf with the information you want.
Login is Pi
Password is pi3.1415

I tried using the dragonminer image but couldn't get it to detect but 2 blades for whatever reason.  The Gridseed A1 unit I got has 5 blades with room to add a 6th.

I'm waiting on a delivery today,  and after that comes I can pull my card out, read the image, and post it for you to download.  I fixed most of their screw-ups in the software, and added a couple input fields for customization.  It doesn't have failover pool support yet, but at least it's in English and you can access the settings page to change things.

Just be warned, with their stock software and cgminer, I have run into instances where I make changes, apply them, and then for whatever reason it doesn't detect the A1 boards again.  I've had to pull the power for about a minute and let it boot back up to get things working again. 

The test points should be marked on the board.  I think I have a picture of one in the 1st post where you can see test points marked 1.2v or 1.2vcc - those are the points you can test the chip voltage. 

Also, most of the time when you measure a component "in-circuit" you will get a different reading than out of the circuit. 

You can simply remove the existing resistor (unless you were planning to parallel it) and just run some small wires to where you have the larger resistor.  I have trim-pots on my Fury's which are connected to the pads via small wire, works fine.  It may not look sexy but who cares.

Has anyone tried getting the Gridseed A1 hardware working under Minera?  Right now I'm using the UGLY AS HELL software that shipped with it....that was either written by a retard, or intentionally "broken" to permanently mine on their pool.  For one, it's pretty much all in chinese.  Second....when you click the settings button to change the pool info and stuff....you get a blank screen.  It immediately starts mining for someone else.  I promptly pulled the SD out and finally found the username/password (pi/pi3.1415) and was able to ssh in and change the settings manually.  So right now it's mining at a steady 1.2Th/s...but using that god aweful interface and no ability to switch pools easily.

I tried following these instructions from https://bitcointalk.org/index.php?topic=495357.msg6672849#msg6672849 and I can get the version of cgminer (3.9.0) from the chinese shitware to launch but it doesn't detect the blades.  The blades and controller board are all hooked up to the rpi GPIO pins. 

I tried the "official" cgminer but I think it's looking for a USB connection and not through GPIO.

Anyone have any idea on what I might be missing to get this rolling?

I have img files of the original software, as well as the binary and config files from the chinese shitware that I can zip and up post if needed.  Michelem - I can open up an ssh tunnel to the unit if you'd like direct access for testing sometime. 

For anyone else that may experience this....I solved it.  Swapped out the 4 port USB hub with a new 7-port hub (with individual on/off switches for each port).  Now it works perfect.  The only thing I found is that the falcon doesn't like to play nice with the hub either.  It's using the FTDI USB chip though.  What happens is that even if the hub port is switched off, it still seems to light up the indicator somewhat.  When that happens, cgminer/rpi won't recognize it.  To get it to work, I have to switch the port on first, then plug the falcon in...defeating the purpose of the on/off switch really.  Guess that's what you get for $8 and free shipping.  I believe part of the problem is the falcon feeds voltage through the usb cable and the hub likely doesn't have diodes to block that incoming device voltage.  Dunno...haven't taken the hub apart...and for $8 that resolves the walls of hw errors...I don't really care because I don't intend on making a habit of unplugging things very often.

I requested 649. 

Have you tried going to 3M's website to see if they offer any 'samples' of it?  I have gotten some samples of their Vikuiti LCD backlight films before (made a small 7" standard LCD sunlight readable with them).  They (Like Texas Instruments) offer samples on a lot of their products.  Being samples, it's not much...but it might be enough for a blizzard/fury if they offer it.

:Edit:  http://solutions.3m.com/wps/portal/3M/en_US/3MNovec/Home/Support/RequestSample/   Novec Engineering Fluid is in the list

Just for the hell of it, I submitted for a sample order.

+1.  I don't have any of those blowers but they would be good for blowing the air through the chambers.  
I actually have a couple blowers from old GPU's....hmmmm.....great....new idea...thanks

:Edit:  Total tard moment...complete brain fart...I misread Novec and thought Nidec who make some decent blower fans

Try running a Falcon/Thunderx3.  I run my miners in my basement, ambient temp is around 69f....so it's nice and cool for operating miners.  The Falcon has 3 "turbo fans" on it and it sounds like a dang vacuum cleaner.  Compared to the Falcon the fans on my fury's (inlcuding the 50mm fan mounted on the end) are inaudible.  Hell...I can't hear a damned thing when I'm near the falcon in the basement.  I may end up moving it to the garage down there. 

If you're really feeling adventurous you could relocate the power connector to the other end of the board so that all the cables are connected to one end.  Reducing clutter, and allowing for a somewhat larger fan to be mounted on the one end to blow air through both units.  Check out post 596 on page 30.  Sorry the pics are so big, took them with my phone and it creates huge pics.  One of these days I'll resize them.

If you want the MOST, definitely use thermal adhesive.  The sticky adhesive that comes on the heatsinks is ok, but nowhere near as good as something like arctic ceramique   or another dedicated thermal adhesive.  I used thermal adhesive on my first two Fury's and was able to hit the 350/354mhz mark with a 4%-5% error rate.  This becomes even more important if you voltmod it as it generates even more heat.  Airflow through the circuitboard chamber is also important...the more the better but don't go deaf due to diminishing returns.  The second two Fury's I first tried the sticky adhesive that came on the heatsinks and I had much higher error rates.  The sizes of the heatsinks is also important.  Try to get the biggest heatsinks that will fit in the chamber.  Tiny mosfet heatsinks will not do much.

There is a list of rpi images in the first post.  Minera was just added and it supports just about everything and has a very active and responsive developer. 

I'd love to build our own miners.  I've been working with computer's for roughly 30 years myself.  Originally went to school for electronics, got sidetracked by the pretty pixels, but have always had my hands in the hardware side changing and modifying things.  I look at the hardware that Gridseed and Zeus have put out there and am baffled by some of the design decisions they chose.  The week power components that were used on the blades, and Zeus with wires coming off the boards on two ends....I suppose if you like a messy setup with cables everywhere it would be alright.