[DIY] - Reward $100 | Antminer S1/S3 Blade on Raspberry Pi

[Mineit]

I flashed to the last one, now i flashed back to the one that supports S3+ which i first used but still a no go.

i checked on loose connectors but it still looks good as far i could see.
its kinda late now im not to sharp.

I saw this in my kernel:

Tue Aug 18 05:57:18 2015 daemon.info sysinit: pic update
Tue Aug 18 05:57:18 2015 daemon.info sysinit: open /dev/bitmain0 to restart pic
Tue Aug 18 05:57:18 2015 daemon.info sysinit: open /dev/bitmainbl0 ok fd = 0x4
Tue Aug 18 05:57:18 2015 daemon.info sysinit: read data
Tue Aug 18 05:57:18 2015 daemon.info sysinit:

• {0x1}, [1]{0x10}, [2]{0x1}, [3]{0x10}, [4]{0x1}, [5]{0}, [6]{0x10}, [7]{0x1}, [8]{0x10}, [9]{0x4}, [10]{0x4},

Tue Aug 18 05:57:18 2015 daemon.info sysinit: Version 1.0
Tue Aug 18 05:57:18 2015 daemon.info sysinit: open pic update file fail

Think there is the problem. but i dont know how to solve that

EDIT

This is btw on the first supported for S3+ and on the last version same error

[Mineit]

nvm i found out thats normal...

[pinhead666]

Just a hunch:Maybe regulator(s) are too hot and 1V8 voltage isn't stable.That affects levelshifting and data isn't coming through?Do you have multimeter?

[pinhead666]

Now I'm pretty sure there is problem with 3V3 regulator. My prototype board worked well over 12 hours but my second version reported 16 blades with 1 chip each when I woke up this morning! Prototype board has better heat-sink for 3V3 regulator. L1117 should handle that drop from 3V3 to 1V8 but my version is D-PAK2 so it would be better to use TO-220 version which would be easier to cool. Right now I'm gonna feed 3V3 from lab-power supply and leave it on for next night to confirm my conclusions.

Edit: It doesn't work with external 3V3. Voltages are good but no communication between controller and blades. I sadly don't have logic analyser so I'm only guessing here.Too much interference from external psu? I replaced heatsink with bigger-one and it works right away.Now I leave it on and do other things.I'll check in the evening if it's still working.

[Mineit]

My 3.3V reports 3.34 with my multimeter and my 1.8v is between 1.79 and 1.8v

I have the same problem as u, no connection between controller and the blade.

[Mineit]

3.3v from the blade is also 3.34/3.35 so i think thats good

[pinhead666]

There is problem with transistor driver.Disconnect pin 3 from controller(s) and blades should be recognised. When I connect pin 3 on the fly,blade starts hashing but quite quickly dies. You can measure pin 14,which should be 1V8 when blade is recognised.It drops little over 1V when it's not working.And controller pin 3 is little over 2V when it should be 3V3. I have this problem too with my second adapter. First one worked so long with no problems.I'll have to connect it again and see if it still works.I used little different value resistors connected from t1 to ground.And different transistors.But I have to sleep sometimes,so I'll be back later.

[pinhead666]

Now I wonder if I have killed that blade.I removed link to my schematic.I will try with original S3 controller and see if the blade still works.But 'cause I don't have cables it will be painfull solder job again.
Edit: Good news is that board works still with original controller. I might take a break,do other things and then re-think this thing.I have ordered bi-directional level shifters from ebay,but I think it'll be maybe two weeks or so when they arrive.

[Mineit]

Your new schematic sended me in the right direction!

After i saw that u changed the 33k resistor location. I found out that there where some loose connection and that is why it worked only for a little while when i just builded it. it worked when the 33k resistor where loose but when it makes contact it didnt work. So after changing that it where all running fine, even on my BM1117.

Then i ran into some other problem. I thought it where running on S3+ blades but they where just normal S3 Blades. When i connect the prototype to the S3+ blades they wont do anything. Even a original controlboard ( S3 versions) wont work. The blade only gives values on the TX, the RX ones dont do anything   Are my S3+ blades dead  

Other that that the prototype works! i just connected the fan wires and they all work fine.

Ill let it hash over the night and see how it goes.

I only have one question though, maybe someone knows it... Does a S3 Controller work on S3+ Blades? Else its 100% sure that the blades died with the controller when it where fried.

[pinhead666]

I also came to conclusion that those resistors are the reason.I used other values in first prototype than my second one.I was gonna try 100k trimmer there but I will remove them and watch what will happen.Thank you mate! I don't know about s3+ blades 'cause don't have any.I try to buy one to play with. But I remember that when I had S3+ and S3's, I mixed controllers and blades and they did work.I had one with s3 controller driving 1 s3+ blade and 1 s3 blade.

[Mineit]

Thank you too mate!

Well then i guess the S3+ blades died sadly, but i still wont give up lol. There's got to be something there, ill check it later.

EDIT

I wanted to solder my second connector, but to think of it. Do i need to add another mosfet curcuit or can i just add it to the same group?

[pinhead666]

Do you mean to use same regulators for two controller? I think they should handle that. Could you please clarify how is your T1 connected now? And about dead blades there's been talk about dead oscillators. I think "somebody said somewhere" that he had replaced one and got the blade running again. Or it might have been S1 blade but they are so similar that I would try replacing it anyway.It's pretty easy to solder if you have hot air solder station.

[Mineit]

The board is the same as your 'corrected scheme' that u removed.

Yeah i saw that about the dead blades but mine doesn't have oscillators on it. So i cant try that

Edit

Found the oscillators on the blades, i need to order the replacements first before i can try it

[Mineit]

Ill try to solder it rq, i prepared the cables already. ill post my results

[Mineit]

right, like i thought they only work separately. That why u got 2 more mosfets on your new version right? To control 2 blades?

[pinhead666]

If by "mosfet" you mean transistors t1 and t2 so yes you need separate lines for each blade.Same regs but 4 trannys and 2 resistor voltage dividers for two blades.Every tx and rx line must be separated,otherwise you get signals mixed and that's not so good.
About oscillators: If you have old S1 blades you can salvage it from there.

[Mineit]

Yes sorry meant transistors. Alright ill try to get it work with 2 blades thanks!
If i get it right im gonna start to make a fine pcb with connectors. Im getting crazy of all this mess!  

About my try to repair my S3+ blades:

After i took the whole thing apart again i found the oscillators on the S3+ blades, which i first thought there weren't any on it.

I had an old, still working S1 miner. First i checked how the oscillators responded with an oscilloscope and they where working fine as expected.
Then i checked my S3+ blades and they where dead indeed. No signals, no voltage, nothing.
The donated oscillator works after replacing it on the S3+ blades, again i checked it with an oscilloscope. It gets it voltage and signals on all sides again.

But STILL no signal on the RX. The TX still around 1.8v

So there's got to be something else that is fried and i cant find out what it is. Im about to give up on it now.

[adib]

Yes sorry meant transistors. Alright ill try to get it work with 2 blades thanks!
If i get it right im gonna start to make a fine pcb with connectors. Im getting crazy of all this mess!  

About my try to repair my S3+ blades:

After i took the whole thing apart again i found the oscillators on the S3+ blades, which i first thought there weren't any on it.

I had an old, still working S1 miner. First i checked how the oscillators responded with an oscilloscope and they where working fine as expected.
Then i checked my S3+ blades and they where dead indeed. No signals, no voltage, nothing.
The donated oscillator works after replacing it on the S3+ blades, again i checked it with an oscilloscope. It gets it voltage and signals on all sides again.

But STILL no signal on the RX. The TX still around 1.8v

So there's got to be something else that is fried and i cant find out what it is. Im about to give up on it now.

If you look closely you will see that there are two places where you can put those oscillators, i wanted to try to move the osc from the side to tha middle of the blades(i have seen blades with ocs in the middle).
I dont have a controller or the components to build one anymore.
You should try that.

You also have to desolder some very small resitors and condensors surrounding the original osc and move them toghter with the new placed osc as those components are also missing and i belive they are needed for the new placed osc to function.

Let me know how it goes.

[bmoscato]

What happened to the guys that were building boards? Did we settle on using the CP2102 adapters?

[HerbPean]

What happened to the guys that were building boards? Did we settle on using the CP2102 adapters?

he's busy with other project... he had one prototype running.