Topic: [DIY] - Reward $100 | Antminer S1/S3 Blade on Raspberry Pi - page 33. (Read 82132 times)
Awesome! Are you going to offer these as a kit? Would love to assemble one.
It's time for some soldering job.
I have 12 S1 boards available @$185 shipped for all 12 in continental US.
Send me a PM.
Thanks,
Alex.
Send me a PM.
Thanks,
Alex.
All the chips are connected clock in/clock out which leads to the tx/rx header (pin 3&5). The other pins are for fan and temp signals other than pin 4 for ground. The power header is used to power the control board (2 ground and 2 3.3v). Ill see if I can run the board at what is to be thought full voltage/full clock and report back. I will also try setting clock speeds and see what I end up with when I get out of work.
im home tuesday and i should be getting a bunch of large heatsinks in, so as soon as i get the controller board from j4bber i will start looking into the code.
All the chips are connected clock in/clock out which leads to the tx/rx header (pin 3&5). The other pins are for fan and temp signals other than pin 4 for ground. The power header is used to power the control board (2 ground and 2 3.3v). Ill see if I can run the board at what is to be thought full voltage/full clock and report back. I will also try setting clock speeds and see what I end up with when I get out of work.
im home tuesday and i should be getting a bunch of large heatsinks in, so as soon as i get the controller board from j4bber i will start looking into the code.
All the chips are connected clock in/clock out which leads to the tx/rx header (pin 3&5). The other pins are for fan and temp signals other than pin 4 for ground. The power header is used to power the control board (2 ground and 2 3.3v). Ill see if I can run the board at what is to be thought full voltage/full clock and report back. I will also try setting clock speeds and see what I end up with when I get out of work.
I was getting 27 gh average but def seen 30-35 for a while. What I did was ground pin 4 on to the uart converter ground. I was getting those speeds at .8 volts. I would try one at full voltage but I don't have a proper heatsink on it.
That would mean that all the chips are performing as normal, but the results aren't read by cgminer, or something similar.
This could be helpful to fix the software.
I suspected it from the begining since all my chips were quite hot to touch as if they were all hashing full speed.
I was getting 27 gh average but def seen 30-35 for a while. What I did was ground pin 4 on to the uart converter ground. I was getting those speeds at .8 volts. I would try one at full voltage but I don't have a proper heatsink on it.
this is just running cgminer with no chip freq or options. undervolted 2.55 @ the resistor. gave it a little bit but i dont wanna run it too long, all i have are little heat sinks on each chip.
![]()
what specific version of cgminer are you using? (link)
Looks like it's the same as I was using.
The one for the U2 on bitmaintech site.
3.8.5 from here
https://bitmaintech.com/support.htm?pid=007201404130633105772QFG6R2u0635
ok, im home on tuesday and will start to poke around the source then, and wait for the control board so i can do some testing then....
this is just running cgminer with no chip freq or options. undervolted 2.55 @ the resistor. gave it a little bit but i dont wanna run it too long, all i have are little heat sinks on each chip.
![]()
what specific version of cgminer are you using? (link)
Looks like it's the same as I was using.
The one for the U2 on bitmaintech site.
3.8.5 from here
https://bitmaintech.com/support.htm?pid=007201404130633105772QFG6R2u0635
this is just running cgminer with no chip freq or options. undervolted 2.55 @ the resistor. gave it a little bit but i dont wanna run it too long, all i have are little heat sinks on each chip.
![]()
what specific version of cgminer are you using? (link)
offtop special registed for this thread.
J4bberwock , i think better way use fubarino sd(http://fubarino.org/sd/index.html) + plug 6!!! boards
or apply Fubarno Mini or Pengino Kit Pinguino 32MX250
http://www.pinguino.cc/shop/index.php?route=product/product&path=64&product_id=52
I thik wll more easy to configure software for these bards based as original on PIC32
special registed for this thread.J4bberwock , i think better way use fubarino sd(http://fubarino.org/sd/index.html) + plug 6!!! boards
or apply Fubarno Mini or Pengino Kit Pinguino 32MX250
http://www.pinguino.cc/shop/index.php?route=product/product&path=64&product_id=52
I thik wll more easy to configure software for these bards based as original on PIC32
J4bberwock - this was posted here for cgminer: https://bitcointalk.org/index.php?topic=28402.15600
Maybe compile for your RasPi with the changes and see if you get a reasonable hash speed?!
Maybe compile for your RasPi with the changes and see if you get a reasonable hash speed?!
Was able to get it to work on an S1 making these changes to driver-bitmain.c:
Code:
477c477
< work_id = DATAW(witem)->work->id;
---
> work_id = DATAW(witem)->wid;
1292,1293c1292,1293
< min = DATAW(info->work_list->tail)->work->id;
< max = DATAW(info->work_list->head)->work->id;
---
> min = DATAW(info->work_list->tail)->wid;
> max = DATAW(info->work_list->head)->wid;
1509c1509
< if (DATAW(witem)->work->id == rxnoncedata.nonces[j].work_id)
---
> if (DATAW(witem)->wid == rxnoncedata.nonces[j].work_id)
2437,2438c2437,2438
< //DATAW(witem)->work->id = ++info->last_wid;
< //info->queued++;
---
> DATAW(witem)->wid = ++info->last_wid;
> info->queued++;
This is a hack to get S1 to work on ckpool
LoL, I disagree since I was the one who posted it...
Don't worry, I won't share any more findings, I get the hint! ;-)
8 )
Hi, anybody who can help is welcome to.
Can you explain what is this mod supposed to do in the bitmain driver?
I assume it's for the original antminer control board.
I'm still looking for people able to "hack" the driver to allow the use of an open board using CP2102 (and CP2108) as the UART/USB interface. If you are interested in one of the prototypes to work on the code, let me know, I should have 1 or 2 spare.
First proto will be 2102, "big" boards might use the 2108.
The good thing about 2108 is that it seems no actual miner is using it. Plus it is a Quad UART>USB bridge, it saves space and components on board, but I need to be sure each board will be detected and work assigned correctly.
J4bberwock - this was posted here for cgminer: https://bitcointalk.org/index.php?topic=28402.15600
Maybe compile for your RasPi with the changes and see if you get a reasonable hash speed?!
Maybe compile for your RasPi with the changes and see if you get a reasonable hash speed?!
Was able to get it to work on an S1 making these changes to driver-bitmain.c:
Code:
477c477
< work_id = DATAW(witem)->work->id;
---
> work_id = DATAW(witem)->wid;
1292,1293c1292,1293
< min = DATAW(info->work_list->tail)->work->id;
< max = DATAW(info->work_list->head)->work->id;
---
> min = DATAW(info->work_list->tail)->wid;
> max = DATAW(info->work_list->head)->wid;
1509c1509
< if (DATAW(witem)->work->id == rxnoncedata.nonces[j].work_id)
---
> if (DATAW(witem)->wid == rxnoncedata.nonces[j].work_id)
2437,2438c2437,2438
< //DATAW(witem)->work->id = ++info->last_wid;
< //info->queued++;
---
> DATAW(witem)->wid = ++info->last_wid;
> info->queued++;
This is a hack to get S1 to work on ckpool
LoL, I disagree since I was the one who posted it...
Don't worry, I won't share any more findings, I get the hint! ;-)
8 )
there wasnt any "hint" in it, i just skimmed the thread and saw what looked like a discussion on getting S1/2/3/4 on ckpool...
so, in other words, your more than welcome to post any findings you may find :-)
J4bberwock - this was posted here for cgminer: https://bitcointalk.org/index.php?topic=28402.15600
Maybe compile for your RasPi with the changes and see if you get a reasonable hash speed?!
Maybe compile for your RasPi with the changes and see if you get a reasonable hash speed?!
Was able to get it to work on an S1 making these changes to driver-bitmain.c:
Code:
477c477
< work_id = DATAW(witem)->work->id;
---
> work_id = DATAW(witem)->wid;
1292,1293c1292,1293
< min = DATAW(info->work_list->tail)->work->id;
< max = DATAW(info->work_list->head)->work->id;
---
> min = DATAW(info->work_list->tail)->wid;
> max = DATAW(info->work_list->head)->wid;
1509c1509
< if (DATAW(witem)->work->id == rxnoncedata.nonces[j].work_id)
---
> if (DATAW(witem)->wid == rxnoncedata.nonces[j].work_id)
2437,2438c2437,2438
< //DATAW(witem)->work->id = ++info->last_wid;
< //info->queued++;
---
> DATAW(witem)->wid = ++info->last_wid;
> info->queued++;
This is a hack to get S1 to work on ckpool
LoL, I disagree since I was the one who posted it...
Don't worry, I won't share any more findings, I get the hint! ;-)
8 )
this is just running cgminer with no chip freq or options. undervolted 2.55 @ the resistor. gave it a little bit but i dont wanna run it too long, all i have are little heat sinks on each chip.
15 GH sounds like 8 or so chips are working. 32 chips at 15 GH would mean were only getting .5 GH per chip which is a 3rd of a u1 at close the same voltage. Do some of the other pins need to be connected/shorted/grounded together to chain all the chips? If my board sits plugged in not hashing they all get warm at some point.
That's what needs to be sorted out software side.
Either 8 chips are working, or cgminer only receive/reads data from 8 chips.
Bridging 2 pins on each chip can be done and should result in single chain mode, but it's too hard to do for the Lambda user.
I've done half the chips on one board so far, and I might try to finish it on this weekend, just as a proof of (don't?) work.
Got my 4 s1 blades in last night and a few usb to uart converters. Undervolted to .8 @ 2.6-2.7. With cgminer run it was getting 45-60 ghs, it worked without any options set, also just the basic freq setting worked. Turned it all off and now it won't set freq, still have .8 volts.... so i tried another board without undervolting real quick to see what the issue was and it started to hash. So maybe I have a dead board now? I have a few more boards on the way as well.
I'd check the poolside hashrate, because 60 GH undervolted is way over what I'm getting.
I assume you are using the old U2 cgminer build from bitmain?
I'm believe I'm using the cgminer build from the u1, I'm at work atm but ill check when I get home. It wasn't 60+, it was more like 57-60. Pool side it was at 15 Gh and climbing until I tinkered and restarted it. I havent been able to get the board to hash after that. I'll undervolt another board later tonight and give it a go and snap some pics.
Ok, that makes sense, you should be able to get 15-16 "real" GH/s when undervolted.
I was unable to go over 27 GH/s pool side with non modded board.
Isn't the u1 rated for 1.6-1.8 on ~.8v. So at 32 chips we should be able to get 50-57 GH per 1 board or close to it. 15 GH sounds like 8 or so chips are working. 32 chips at 15 GH would mean were only getting .5 GH per chip which is a 3rd of a u1 at close the same voltage. Do some of the other pins need to be connected/shorted/grounded together to chain all the chips? If my board sits plugged in not hashing they all get warm at some point.
Got my 4 s1 blades in last night and a few usb to uart converters. Undervolted to .8 @ 2.6-2.7. With cgminer run it was getting 45-60 ghs, it worked without any options set, also just the basic freq setting worked. Turned it all off and now it won't set freq, still have .8 volts.... so i tried another board without undervolting real quick to see what the issue was and it started to hash. So maybe I have a dead board now? I have a few more boards on the way as well.
I'd check the poolside hashrate, because 60 GH undervolted is way over what I'm getting.
I assume you are using the old U2 cgminer build from bitmain?
I'm believe I'm using the cgminer build from the u1, I'm at work atm but ill check when I get home. It wasn't 60+, it was more like 57-60. Pool side it was at 15 Gh and climbing until I tinkered and restarted it. I havent been able to get the board to hash after that. I'll undervolt another board later tonight and give it a go and snap some pics.
Ok, that makes sense, you should be able to get 15-16 "real" GH/s when undervolted.
I was unable to go over 27 GH/s pool side with non modded board.
Got my 4 s1 blades in last night and a few usb to uart converters. Undervolted to .8 @ 2.6-2.7. With cgminer run it was getting 45-60 ghs, it worked without any options set, also just the basic freq setting worked. Turned it all off and now it won't set freq, still have .8 volts.... so i tried another board without undervolting real quick to see what the issue was and it started to hash. So maybe I have a dead board now? I have a few more boards on the way as well.
I'd check the poolside hashrate, because 60 GH undervolted is way over what I'm getting.
I assume you are using the old U2 cgminer build from bitmain?
We can say that the HW part is solved or almost done.
Taugeran branch will use the original chips used by Bitmain, so it should handle 2 hashing units without any trouble, and software should work as-is.
I decided to follow a different route to allow more hashing boards to be attached, and control units to be chained.
This route will need some software hack to allow multiplechain with the driver from the U2, or to allow proper detection of the hacked boards with the S1 driver.
My protoype PCB finally left china today, so I'll have them delivered worst case on monday by DHL.
Then, soldering, and tests.
If I did a good job, they will work, and I'll send some to people who offered some help on the software.
Jump to: