The Chili – 30+GH/s BFL based Bitcoin Miner Assembly

[MrTeal]

Hi,

now I understand why one of my boards begins hashing but fails with a short on 12V after 30 minutes:

I will replace them. I hope I find one in one of my boxes. It will hard to replace them with the the backplate heatsink which is melted by thermal activated bonding foil... I will use my pre heater with low temperature (100 °C) and leaded solder.

Demounting was easy (twist left-right some times). After demontage the chilly works again and has not further an short circuit on the 12V rail any more. So I hope I get it back this evening ;-)

Cheers...

Wow, I can't believe that happened, or that it wouldn't have shorted out sooner.
The best way to solder the new one on is to use an iron, and either solder paste or flux. Clean the pad off, and apply the paste to the pads. Put the new cap down, and just heat the pad and lead until the solder paste flows.

[1714727571]

1714727571

[1714727571]

1714727571

[1714727571]

1714727571

[1714727571]

1714727571

[GrapeApe]

Wow, I can't believe that happened, or that it wouldn't have shorted out sooner.
The best way to solder the new one on is to use an iron, and either solder paste or flux. Clean the pad off, and apply the paste to the pads. Put the new cap down, and just heat the pad and lead until the solder paste flows.

Hey Mr.Teal I have several caps that broke off of my boards from Lucko from a rough delivery. I have little to no soldering experience but am willing to buy an iron and get some. Can you please give a little more instruction on how to solder them back on. Also I'm here in Austin so if you think it would be better to take them in somewhere do you have any recommendations?

Oh also in the image above his cap has the blue side on the inside I thought it was supposed to be on the outside.

Edit: You know after posting this I'm realizing I'm asking you to hold my hand again, so with that said any pointers would be appreciated but I'll start reading up on soldering. Thx

[asjfdlksfd]

Wow, I can't believe that happened, or that it wouldn't have shorted out sooner.
The best way to solder the new one on is to use an iron, and either solder paste or flux. Clean the pad off, and apply the paste to the pads. Put the new cap down, and just heat the pad and lead until the solder paste flows.

Yep, I did this with a little help from my air solder gun I found an 1500 µF 16 V type (the only which have had more then 12 V) capacity from an old defect pc main board. I used it as a replacement so the board is hashing now (hopefully longer than before.) The 160 mm x 100 mm x 10 mm backplate heat sink is a little bit expensive with costs of 6€ , but it helps to cool the bfl chips and the power dist chips

And all these only for a few month :/

Cheers...

[asjfdlksfd]

Hey Mr.Teal I have several caps that broke off of my boards from Lucko from a rough delivery. I have little to no soldering experience but am willing to buy an iron and get some. Can you please give a little more instruction on how to solder them back on. Also I'm here in Austin so if you think it would be better to take them in somewhere do you have any recommendations?

I cannot recommend this, if you have no experience. The boards are so good the heat of your iron will heat the most of the pcb before the solder will melt.

If you find someone in your near (austin looks like not so a small village you should do this by them. Best is to repair them on local shop which also repairs tv's and so on, if there any. It should not costs more than some $.

For your information, for the outer capacity + is in the inner side. That means the blue mark must shown to the outer side. On my post above you can have a look how it should not be mounted.

Cheers...

[GrapeApe]

Hey Mr.Teal I have several caps that broke off of my boards from Lucko from a rough delivery. I have little to no soldering experience but am willing to buy an iron and get some. Can you please give a little more instruction on how to solder them back on. Also I'm here in Austin so if you think it would be better to take them in somewhere do you have any recommendations?

I cannot recommend this, if you have no experience. The boards are so good the heat of your iron will heat the most of the pcb before the solder will melt.

If you find someone in your near (austin looks like not so a small village you should do this by them. Best is to repair them on local shop which also repairs tv's and so on, if there any. It should not costs more than some $.

For your information, for the outer capacity + is in the inner side. That means the blue mark must shown to the outer side. On my post above you can have a look how it should not be mounted.

Cheers...

Thanks for the advice I'll find someone on Monday to do it.

Edit: ChipGeek got in touch with me he is going to take care this for me.

[ZBC3]

Out of curiosity are there any plans to make boards that have more than 9 BFL chips, or do the BFL chips get too hot to effectively make a larger board?

[MrTeal]

Out of curiosity are there any plans to make boards that have more than 9 BFL chips, or do the BFL chips get too hot to effectively make a larger board?

Not really, no. There's no real point in doing another design, since BFL doesn't sell the chips anymore and they're getting increasingly hard to find. There'd be boards with nothing to put on them.

[ZBC3]

In that case, are you working on any designs using other chips ?

[GrapeApe]

Hey Mr.Teal I have several caps that broke off of my boards from Lucko from a rough delivery. I have little to no soldering experience but am willing to buy an iron and get some. Can you please give a little more instruction on how to solder them back on. Also I'm here in Austin so if you think it would be better to take them in somewhere do you have any recommendations?

I cannot recommend this, if you have no experience. The boards are so good the heat of your iron will heat the most of the pcb before the solder will melt.

If you find someone in your near (austin looks like not so a small village you should do this by them. Best is to repair them on local shop which also repairs tv's and so on, if there any. It should not costs more than some $.

For your information, for the outer capacity + is in the inner side. That means the blue mark must shown to the outer side. On my post above you can have a look how it should not be mounted.

Cheers...

Thanks for the advice I'll find someone on Monday to do it.

Edit: ChipGeek got in touch with me he is going to take care this for me.

ChipGeek got me fixed up both boards are up and hashing away.

[twib2]

I'm still trying to get my second Chili back up and reliable.  Right now, it stops talking to cgminer/bfgminer after 2-8 minutes and needs to be powered off/on and the miner restarted to resume mining.  Rinse and repeat.  I've posted the errors earlier (https://bitcointalk.org/index.php?topic=304250.msg4557778#msg4557778).  Anyone have any ideas/suggestions?

So far, I've tried: reflashing the the 1.1v-limited firmware, reflowing the FTDI chip, reseating a nearby small SMD capacitor that might have gotten dislodged, but so far still unstable.  Before I had the pin 5-6 shorting mishap (hit an adjacent pin at the same time during a reset, trashing my FTDI chip), it was pretty solid, though I did have to reset it about once a day.  Replaced the FTDI chip with a new one, programmed it (thanks Mr. Teal!), and was back hashing, but never stable or even usable.

Help?

[bx8389]

I'm still trying to get my second Chili back up and reliable.  Right now, it stops talking to cgminer/bfgminer after 2-8 minutes and needs to be powered off/on and the miner restarted to resume mining.  Rinse and repeat.  I've posted the errors earlier (https://bitcointalk.org/index.php?topic=304250.msg4557778#msg4557778).  Anyone have any ideas/suggestions?

So far, I've tried: reflashing the the 1.1v-limited firmware, reflowing the FTDI chip, reseating a nearby small SMD capacitor that might have gotten dislodged, but so far still unstable.  Before I had the pin 5-6 shorting mishap (hit an adjacent pin at the same time during a reset, trashing my FTDI chip), it was pretty solid, though I did have to reset it about once a day.  Replaced the FTDI chip with a new one, programmed it (thanks Mr. Teal!), and was back hashing, but never stable or even usable.

Help?

Lucko's cointamination thread have a lot of info about problems and some solutions for chilli boards. I think desserve some attention: start here but look before and after
https://bitcointalk.org/index.php?topic=267649.msg4417461#msg4417461

[Mudbankkeith]

I'm still trying to get my second Chili back up and reliable.  Right now, it stops talking to cgminer/bfgminer after 2-8 minutes and needs to be powered off/on and the miner restarted to resume mining.  Rinse and repeat.  I've posted the errors earlier (https://bitcointalk.org/index.php?topic=304250.msg4557778#msg4557778).  Anyone have any ideas/suggestions?

So far, I've tried: reflashing the the 1.1v-limited firmware, reflowing the FTDI chip, reseating a nearby small SMD capacitor that might have gotten dislodged, but so far still unstable.  Before I had the pin 5-6 shorting mishap (hit an adjacent pin at the same time during a reset, trashing my FTDI chip), it was pretty solid, though I did have to reset it about once a day.  Replaced the FTDI chip with a new one, programmed it (thanks Mr. Teal!), and was back hashing, but never stable or even usable.

Help?

Lucko's cointamination thread have a lot of info about problems and some solutions for chilli boards. I think desserve some attention: start here but look before and after
https://bitcointalk.org/index.php?topic=267649.msg4417461#msg4417461

There is a "slow rise version" of the software being installed on some of the newer boards, maybe MrTeal can send it to you?

[MrTeal]

In that case, are you working on any designs using other chips ?

I have a design here with the Coincraft A1 chips, but due to an error in their datasheet the pads are the incorrect size. It's already fixed, but I haven't sent the modified boards off to be made as I'm going to be extremely time limited this next month. To be honest, I'm not sure we'll pursue it at all. Their pricing needs to come down to make sense, $87.50 for a 30GH/s chip is going to be a tough sell with the delivery timelines they're talking about.

[MrTeal]

I'm still trying to get my second Chili back up and reliable.  Right now, it stops talking to cgminer/bfgminer after 2-8 minutes and needs to be powered off/on and the miner restarted to resume mining.  Rinse and repeat.  I've posted the errors earlier (https://bitcointalk.org/index.php?topic=304250.msg4557778#msg4557778).  Anyone have any ideas/suggestions?

So far, I've tried: reflashing the the 1.1v-limited firmware, reflowing the FTDI chip, reseating a nearby small SMD capacitor that might have gotten dislodged, but so far still unstable.  Before I had the pin 5-6 shorting mishap (hit an adjacent pin at the same time during a reset, trashing my FTDI chip), it was pretty solid, though I did have to reset it about once a day.  Replaced the FTDI chip with a new one, programmed it (thanks Mr. Teal!), and was back hashing, but never stable or even usable.

Help?

So, it doesn't reset or anything, it just stops talking?
Do you have a multimeter, and the next time it stops talking can you measure the voltage across C19 (one of the unpopulated ceramic capacitors between the PSU and ASICs).

[twib2]

I'm still trying to get my second Chili back up and reliable.  Right now, it stops talking to cgminer/bfgminer after 2-8 minutes and needs to be powered off/on and the miner restarted to resume mining.  Rinse and repeat.  I've posted the errors earlier (https://bitcointalk.org/index.php?topic=304250.msg4557778#msg4557778).  Anyone have any ideas/suggestions?

So far, I've tried: reflashing the the 1.1v-limited firmware, reflowing the FTDI chip, reseating a nearby small SMD capacitor that might have gotten dislodged, but so far still unstable.  Before I had the pin 5-6 shorting mishap (hit an adjacent pin at the same time during a reset, trashing my FTDI chip), it was pretty solid, though I did have to reset it about once a day.  Replaced the FTDI chip with a new one, programmed it (thanks Mr. Teal!), and was back hashing, but never stable or even usable.

Help?

So, it doesn't reset or anything, it just stops talking?
Do you have a multimeter, and the next time it stops talking can you measure the voltage across C19 (one of the unpopulated ceramic capacitors between the PSU and ASICs).

When that happens, the indicator lights go dark and the miner can't talk to it anymore.  It will actually reset smoe/most of the time, but of course the miner program doesn't notice that and has to restart before it'll catch again.  Once it's back up it repeats the crash with these three errors randomly streaming/repeating (mostly the timed out and unexpected queue result ones):

Code:

[2014-01-16 22:57:42] BFL 1: Failed to send queue
 [2014-01-16 22:57:43] BFL 1: Error: Get temp returned empty string/timed out
 [2014-01-16 22:57:43] BFL 1: Received unexpected queue result response:

I'll hook it back up again once I have a free PSU and look at the voltage across C19.  What should it be, and what are you hoping to see from that?

[twib2]

I'm still trying to get my second Chili back up and reliable.  Right now, it stops talking to cgminer/bfgminer after 2-8 minutes and needs to be powered off/on and the miner restarted to resume mining.  Rinse and repeat.  I've posted the errors earlier (https://bitcointalk.org/index.php?topic=304250.msg4557778#msg4557778).  Anyone have any ideas/suggestions?

So far, I've tried: reflashing the the 1.1v-limited firmware, reflowing the FTDI chip, reseating a nearby small SMD capacitor that might have gotten dislodged, but so far still unstable.  Before I had the pin 5-6 shorting mishap (hit an adjacent pin at the same time during a reset, trashing my FTDI chip), it was pretty solid, though I did have to reset it about once a day.  Replaced the FTDI chip with a new one, programmed it (thanks Mr. Teal!), and was back hashing, but never stable or even usable.

Help?

Lucko's cointamination thread have a lot of info about problems and some solutions for chilli boards. I think desserve some attention: start here but look before and after
https://bitcointalk.org/index.php?topic=267649.msg4417461#msg4417461

Thanks, but supposedly the non-Lucko boards don't have the hairdryer problem.  Still, I might try that for good luck.  Maybe the initial problem somehow duplicated the strangeness of the Lucko boards...

[MrTeal]

I'm still trying to get my second Chili back up and reliable.  Right now, it stops talking to cgminer/bfgminer after 2-8 minutes and needs to be powered off/on and the miner restarted to resume mining.  Rinse and repeat.  I've posted the errors earlier (https://bitcointalk.org/index.php?topic=304250.msg4557778#msg4557778).  Anyone have any ideas/suggestions?

So far, I've tried: reflashing the the 1.1v-limited firmware, reflowing the FTDI chip, reseating a nearby small SMD capacitor that might have gotten dislodged, but so far still unstable.  Before I had the pin 5-6 shorting mishap (hit an adjacent pin at the same time during a reset, trashing my FTDI chip), it was pretty solid, though I did have to reset it about once a day.  Replaced the FTDI chip with a new one, programmed it (thanks Mr. Teal!), and was back hashing, but never stable or even usable.

Help?

So, it doesn't reset or anything, it just stops talking?
Do you have a multimeter, and the next time it stops talking can you measure the voltage across C19 (one of the unpopulated ceramic capacitors between the PSU and ASICs).

When that happens, the indicator lights go dark and the miner can't talk to it anymore.  It will actually reset smoe/most of the time, but of course the miner program doesn't notice that and has to restart before it'll catch again.  Once it's back up it repeats the crash with these three errors randomly streaming/repeating (mostly the timed out and unexpected queue result ones):

Code:

[2014-01-16 22:57:42] BFL 1: Failed to send queue
 [2014-01-16 22:57:43] BFL 1: Error: Get temp returned empty string/timed out
 [2014-01-16 22:57:43] BFL 1: Received unexpected queue result response:

I'll hook it back up again once I have a free PSU and look at the voltage across C19.  What should it be, and what are you hoping to see from that?

Somewhere between 0.85V and 1.15V.
Most crashes are caused by the power supply turning off for some reason, but this sounds like it might be different. Do you know which other pin you shorted out? Some of the other ones around there are the communication pins between the VRM and the microcontroller. The micro will shut the power down and reset if it reads funky values from the power supply, perhaps that could be a source of the error.

Thanks, but supposedly the non-Lucko boards don't have the hairdryer problem.  Still, I might try that for good luck.  Maybe the initial problem somehow duplicated the strangeness of the Lucko boards...

Lucko and another user who has both types of boards reported they could get similar results on one board if they cool them down substantially (I believe Lucko did it at -10C). I tried a couple of mine at -15C and couldn't reproduce it, but I can't rule out that there is some kind of temperature dependent thing happening. I've only heard of the two units displaying those symptoms though, and none with the ambient above freezing.

[twib2]

Somewhere between 0.85V and 1.15V.
Most crashes are caused by the power supply turning off for some reason, but this sounds like it might be different. Do you know which other pin you shorted out? Some of the other ones around there are the communication pins between the VRM and the microcontroller. The micro will shut the power down and reset if it reads funky values from the power supply, perhaps that could be a source of the error.

Thanks, not sure of the numbering and I wasn't looking really well when I shorted the wrong pins (hence the mistake), but I *think* it would have been either one of the inner row of pins (farthest away from the edge of the PCB) right next to the pair I was trying to short.  So, if it looks like this:

-------------------edge of PCB
2  4  6  8  10
1  3  5  7  9

then I think it might have been pins 3 and 6 or even 3+5+6, but again that's a guess since the whole mis-reset happened because I was too quickly trying to reset the board when I really should have just turned off and on the PSU (was trying to keep the other Chili on the same PSU running during the reset).  All I know is that the USB comms didn't seem to work after that and so I replaced the FTDI chip and then all seemed pretty normal (once I had reprogrammed the FTDI).

[goggles1200]

Is there a site, web page or pdf out there with the schematic of the chili board? I ask for reference and to trouble shoot my fan.

Thanks.

[lin0sspice]

How can i fix Temperature problem ?
say 100-105C. ( 1 flashing light )
But i was running fine after that. I just restart pc then it doesnot work any more.
I disconnected all cable for couple hours but 2 light flashing and not off.
Then i turn PSU off and turn on again. Chili was flashing 1 light again with temperature 100-105 SICK again.
Please help.