Pages:
Author

Topic: The Chili – 30+GH/s BFL based Bitcoin Miner Assembly - page 32. (Read 138054 times)

newbie
Activity: 29
Merit: 0
Please somebody upload the original firmware. After updating firmware, my chilis run slow.
How much slower?  How long did you let them run before deciding they are slow? 

When first powering on, they intentionally run slow then speed up based on temperature.

Sorry it was due to cooling problem, not firmware. Now it runs 35GH/s and 38GH/s. Thanks!!
sr. member
Activity: 267
Merit: 250
Regarding the firmware, do you think it can be tweaked to increase performance any?  Just wondering since it'd be a huge benefit to those of us running these devices already!
full member
Activity: 198
Merit: 100
Please somebody upload the original firmware. After updating firmware, my chilis run slow.
How much slower?  How long did you let them run before deciding they are slow? 

When first powering on, they intentionally run slow then speed up based on temperature.
newbie
Activity: 29
Merit: 0
Please somebody upload the original firmware. After updating firmware, my chilis run slow.
sr. member
Activity: 267
Merit: 250
Earlier this week, I recieved 4x more Chilis from another forum member. The hashrates reported to me from the other member before taking delivery were 29-31ghs.  I asked for them without the heatsinks, since I already had 4x extra Antec Shelfs.  I used the above method again which I posted about before and I am now getting between 32 and 35 ghs.  Seems like it's quite an improvement, if I may say so myself!  

Also, one of my original 4 Chilis is acting very different than the rest.  It always shows 58-59C and it is my highest performing one at around 36.5 ghs.  Very interesting.

My original 4x are still getting between 34 and 36 ghs (and the one which runs at 58C has been running above 36.5 ghs for some reason)

Have you tried adding airflow on the undersides as well?

Yes, I have 3pin case fans blowing onto them...
legendary
Activity: 1973
Merit: 1007
Really happy w/ my chilis atm - bought a sunzi (6 chilis)

In my cold room ... they are clockin between 36-39GHs/avg ... freakin awesome!
My overall error rate at this speed is 7%
Power consumption @ wall - 1280W

Total avg hashrate: 225GH


Wow, that is awesome! You must be keeping those chilis chilly.
hero member
Activity: 681
Merit: 500
Earlier this week, I recieved 4x more Chilis from another forum member. The hashrates reported to me from the other member before taking delivery were 29-31ghs.  I asked for them without the heatsinks, since I already had 4x extra Antec Shelfs.  I used the above method again which I posted about before and I am now getting between 32 and 35 ghs.  Seems like it's quite an improvement, if I may say so myself!  

Also, one of my original 4 Chilis is acting very different than the rest.  It always shows 58-59C and it is my highest performing one at around 36.5 ghs.  Very interesting.

My original 4x are still getting between 34 and 36 ghs (and the one which runs at 58C has been running above 36.5 ghs for some reason)

Have you tried adding airflow on the undersides as well?
sr. member
Activity: 267
Merit: 250
Sorry ahead of time for the low quality images.  Also, I figure everyone is already past this step at this point in the game... but I'd post it anyway.

Originally, I was going to use the ARCTIC Accelero Mono Plus VGA Coolers and grab some PCI brackets, but that setup was considerably more expensive and proved to show lower hashing results as posted by others.

Then, I was going to do the double Antec Shelf idea like MrTeal. I ordered 8 coolers for 4 boards, but I tried it with a single one and I am fine with the results... so now I am going to buy 4x more boards to use up my 4 extra Shelfs.

So to mount the single shelf to the Chili, I mounted it directly to the board with the provided 1156 back plates.  I mounted the top mount to the heatsink as it was directed in the instructions (to the underside of the heatsink).  I then used misc computer screws which threaded into the back plates... I have a ton of screws laying around from buying internal computer devices (CDROM, floppy, etc drives) and they provided a very tight connection.  The board doesn't look like it's bowing at all.  I used nothing besides a rice sized amount of the Antec included thermal grease on each chip (maybe too much but it seemed to work).  I didn't really think about standoffs, since I have them mostly sitting in their sides like the first picture showing the back plates.  If you wanted to sit them upright, they sit on the back plates.  It's probably not 100% safe (something could short the power solder points underneath), but oh well. It works for me.

I tried one Chili with a thermal pad, but that one got 27 Ghs.  I redid it the way I did the others and it raised up the Ghs to around 34 or so.

All in all, I'm getting between 34 and 36 ghs.











Earlier this week, I recieved 4x more Chilis from another forum member. The hashrates reported to me from the other member before taking delivery were 29-31ghs.  I asked for them without the heatsinks, since I already had 4x extra Antec Shelfs.  I used the above method again which I posted about before and I am now getting between 32 and 35 ghs.  Seems like it's quite an improvement, if I may say so myself!  

Also, one of my original 4 Chilis is acting very different than the rest.  It always shows 58-59C and it is my highest performing one at around 36.5 ghs.  Very interesting.

My original 4x are still getting between 34 and 36 ghs (and the one which runs at 58C has been running above 36.5 ghs for some reason)
hero member
Activity: 711
Merit: 532
Has anyone found a new *cheap* source for the Antec Kuhler Shelf? The $10 ones are out of stock:
http://store.antec.com/Product/cooling-cpu_cooler/kuhler-shelf/0-761345-77081-1.aspx
legendary
Activity: 2450
Merit: 1002
Really happy w/ my chilis atm - bought a sunzi (6 chilis)

In my cold room ... they are clockin between 36-39GHs/avg ... freakin awesome!
My overall error rate at this speed is 7%
Power consumption @ wall - 1280W

Total avg hashrate: 225GH
member
Activity: 67
Merit: 10
Any news about batch 3? Are you still planning on auctioning off some populated boards?
Yeah, I need to get on that. I should be posting it up tonight or tomorrow.

I got to take a look at that little board last night.... I would be interested in leftovers from batch 3 :-).

If I'm not mistaken, batch 3 has not been released yet.

To clarify I saw a board from one of the earlier batches (that I wasn't a part of)...

I wasn't trying to imply that Batch 3 had any released boards.. sorry about that :-)
sr. member
Activity: 262
Merit: 250
Any news about batch 3? Are you still planning on auctioning off some populated boards?
Yeah, I need to get on that. I should be posting it up tonight or tomorrow.

I got to take a look at that little board last night.... I would be interested in leftovers from batch 3 :-).

If I'm not mistaken, batch 3 has not been released yet.
legendary
Activity: 2450
Merit: 1002
Is it possible to flash new firmware to these using linux? (a rpi w/ minepeon specifically)
If not, can someone point me to the windows flashing procedure & files? Thanks
member
Activity: 67
Merit: 10
Any news about batch 3? Are you still planning on auctioning off some populated boards?
Yeah, I need to get on that. I should be posting it up tonight or tomorrow.

I got to take a look at that little board last night.... I would be interested in leftovers from batch 3 :-).
sr. member
Activity: 262
Merit: 250
newbie
Activity: 9
Merit: 0
We might take a look at something like that. The minions are interesting chips in a lot of ways. A lot of the basic design is done, so while it wouldn't be a dead simple port like Avalon Gen1 to Gen2, it would be a similar interface and setup to the BFL chips.


EPIC!

I am in! Should we start a tally somewhere to get an idea of how much support there is for this?

Bob, what do you think is a fair price per chip since you are THE first person in line?

I cant really sell the chips at a discount, but I can offer them to be delivered ASAP. I guess I should make the thread now..

Count me in as well.   Cool

I do a simple and rough math.
The unit cost per GH/s of Chili-Minion is about 1/10 of Chili-BFL.
Suppose the difficulty adjust is about 20% rising exponential. (1.2)^x=10, x~=6.
It means the shipment of Chili-Minion board should be within 60 days, or it's hard to break-even assume there's no another BTC/USD surge in the future.

Even so, I still want in Smiley Please reserve 8 Minion chips for me, if MrTeal and ChipGeek would do it.
legendary
Activity: 1274
Merit: 1004
Any news about batch 3? Are you still planning on auctioning off some populated boards?
Yeah, I need to get on that. I should be posting it up tonight or tomorrow.
legendary
Activity: 1274
Merit: 1004
Quote
LED decoder ring

The LEDs are numbered 1 to 8 with #1 closest to the USB connector and #8 closest to the power and fan connectors.  During first power up, LEDs 5, 6, 7, and 8 will light up and LEDs 1, 2, 3,  and 4 will indicate a failure code if anything is wrong with the hardware.  This happens VERY quickly - in less than a second.

If the initial hardware checks are good, #7 will blink indicating it is waiting for the board to cool down.  This will always blink for a few seconds even if the board is cold. 

Then #7 and #8 will blink indicating ASIC self test. 

After that, 1, 2, 3, 4 will indicate how many jobs from cgminer are waiting to run in the input queue.  LED 5 is just a debug output for me but it roughly indicates a job has completed - but sometimes blinks too fast for the human eye to see.  Just ignore LED 5.

I am working on some firmware updates that will fix most (or all?) of the problems the boards are having.  For example, one failure I think I have figured out is occasionally a board will jump to more than 100 GH/s but have 100% hardware errors.  I have one board that does this once per day or so and needs to be rebooted.  After chasing this for a while, I believe I have finally figured out what is going on and will have a fix in the next release.

Hi,

what does it mean if a board leds 1-4 will be stay off and 5-8 will be stay on? No usb connection is registred if I plug in on my linux box? TP5V and TP3V3 are looking fine with measured 4,99V and 3,38V.

I registred the problem on one board before mounting the bgas and on one board after mounting the bgas, so maybe the problem allready exists before I mounted the chips.

Hi, I also checked now the other unflowed boards without mounted chips. They are all starting as well with the four leds for x/10 seconds also short time as now expected and  then different leds are flickering for seconds before they will all be go off. This looks like great for the rest of my boards. So first I will continue to assemble and test them.

I have not a plug with this dimension to the jtag connector so I cannot check at the moment the connection and status of the arm which maybe is not working well. I will order an cable so I can connect my rpi to the jtag.
Because of missing documentation do you can give me instructions what I can check to analyze and fix the boards by myself? Maybe the test points F0-F7 are helpfull? I didn't checked them.

Thank you,
The connector is a 10 pin 0.05" ARM Cortex JTAG connector, using the standard pinout.
http://www.digikey.com/product-search/en?x=0&y=0&lang=en&site=us&KeyWords=609-4055-nd
http://infocenter.arm.com/help/topic/com.arm.doc.faqs/attached/13634/cortex_debug_connectors.pdf

Test points F0 to F7 are the frequency outputs of the ASICs, you can see the actual speed of each chip at those points.

Send me an email with as much detail as you can on what's happening with the board that you mounted the ASICs on, and I'll see what I can do to help. It sounds like the other ones are starting up correctly, so it could be a reflow issue. LEDs 5-8 coming on and staying on usually indicates that the 1V supply is either not initializing or is not asserting the power good signal.
full member
Activity: 128
Merit: 100
Quote
LED decoder ring

The LEDs are numbered 1 to 8 with #1 closest to the USB connector and #8 closest to the power and fan connectors.  During first power up, LEDs 5, 6, 7, and 8 will light up and LEDs 1, 2, 3,  and 4 will indicate a failure code if anything is wrong with the hardware.  This happens VERY quickly - in less than a second.

If the initial hardware checks are good, #7 will blink indicating it is waiting for the board to cool down.  This will always blink for a few seconds even if the board is cold. 

Then #7 and #8 will blink indicating ASIC self test. 

After that, 1, 2, 3, 4 will indicate how many jobs from cgminer are waiting to run in the input queue.  LED 5 is just a debug output for me but it roughly indicates a job has completed - but sometimes blinks too fast for the human eye to see.  Just ignore LED 5.

I am working on some firmware updates that will fix most (or all?) of the problems the boards are having.  For example, one failure I think I have figured out is occasionally a board will jump to more than 100 GH/s but have 100% hardware errors.  I have one board that does this once per day or so and needs to be rebooted.  After chasing this for a while, I believe I have finally figured out what is going on and will have a fix in the next release.

Hi,

what does it mean if a board leds 1-4 will be stay off and 5-8 will be stay on? No usb connection is registred if I plug in on my linux box? TP5V and TP3V3 are looking fine with measured 4,99V and 3,38V.

I registred the problem on one board before mounting the bgas and on one board after mounting the bgas, so maybe the problem allready exists before I mounted the chips.

Hi, I also checked now the other unflowed boards without mounted chips. They are all starting as well with the four leds for x/10 seconds also short time as now expected and  then different leds are flickering for seconds before they will all be go off. This looks like great for the rest of my boards. So first I will continue to assemble and test them.

I have not a plug with this dimension to the jtag connector so I cannot check at the moment the connection and status of the arm which maybe is not working well. I will order an cable so I can connect my rpi to the jtag.
Because of missing documentation do you can give me instructions what I can check to analyze and fix the boards by myself? Maybe the test points F0-F7 are helpfull? I didn't checked them.

Thank you,
Pages:
Jump to: