BFL board project COINTAMINATION - EU facility - ORDERS OPENED for Chili - page 27.

Lucko

hero member

Activity: 826

Merit: 1000

Quote from: bx8389 on October 28, 2013, 04:25:34 AM

Quote from: GrapeApe on October 27, 2013, 02:43:01 PM

Well I haven't jumped ship. I have no where else to go so I (we) are counting on you.

Me too, It is difficult for me switch to another solution now
I would prefer to know board problem before take any decision.

Board is definitively not recoverable?
Or is recoverable but not on an reasonable time frame?

Regards

Hard to say. It could turned out to be a simple fix or error on a component or Sad

new board... So as long as I don't know that it is hard to say. But new board would take some time...

Lucko

hero member

Activity: 826

Merit: 1000

Quote from: Mudbankkeith on October 28, 2013, 06:06:21 AM

Batch 1 and Batch 3 chips are now at Ljubljana.
Is there any news on the batch 2 chips?

Yes but lack of documentation is making it difficult to get them out of customs...

I got no replay from BFL about batch 2....

jelin1984

legendary

Activity: 2408

Merit: 1004

Where is batch 2?

Mudbankkeith

hero member

Activity: 868

Merit: 1000

Batch 1 and Batch 3 chips are now at Ljubljana.
Is there any news on the batch 2 chips?

GrapeApe

sr. member

Activity: 476

Merit: 250

Quote from: Lucko on October 28, 2013, 01:12:55 AM

Updates...

Tracking is unfortunately down so I don't know where pacage is. I hope form gets it today...

I have a backup plan ready. MrTeal and I are working it out for me to licence his board... Unfortunately that second solution could cost more...

I have one of Mr.Teals Chilli boards. It's a good board (quiet) but it does cost more for sure.

Mudbankkeith

hero member

Activity: 868

Merit: 1000

Quote from: bx8389 on October 28, 2013, 04:25:34 AM

Quote from: GrapeApe on October 27, 2013, 02:43:01 PM

Well I haven't jumped ship. I have no where else to go so I (we) are counting on you.

Me too, It is difficult for me switch to another solution now
I would prefer to know board problem before take any decision.

Board is definitively not recoverable?
Or is recoverable but not on an reasonable time frame?

Regards

Still with you Lucko

bx8389

member

Activity: 80

Merit: 10

Quote from: GrapeApe on October 27, 2013, 02:43:01 PM

Well I haven't jumped ship. I have no where else to go so I (we) are counting on you.

Me too, It is difficult for me switch to another solution now
I would prefer to know board problem before take any decision.

Board is definitively not recoverable?
Or is recoverable but not on an reasonable time frame?

Regards

Lucko

hero member

Activity: 826

Merit: 1000

Updates...

Tracking is unfortunately down so I don't know where pacage is. I hope form gets it today...

I have a backup plan ready. MrTeal and I are working it out for me to licence his board... Unfortunately that second solution could cost more...

dwdoc

legendary

Activity: 966

Merit: 1000

- - -Caveat Aleo- - -

Likewise!

GrapeApe

sr. member

Activity: 476

Merit: 250

Well I haven't jumped ship. I have no where else to go so I (we) are counting on you.

Lucko

hero member

Activity: 826

Merit: 1000

Thanks all who are trying to help. For now test were done only on one board with real chips. There are 2 more and I just send all to form(my board guy is away on travelling). I don't have that much equipment as he. There might be also a possible that there is a mistake on a board or one of the components that prevents Atmel to work or error on Atmel itself or something on chips(that is a suspect at the moment) but this type of error we seen first time. I did get a replay from MrTeal and I would like to thank him for it and he give me idea but can't be tested(since boards are on the way to form). I will be in contact with him over the weekend and I hope we can figure it out... If not form will do his best next week.

Now what MrTeal told me is that this chips act differently then "old" one when you turn them on at low voltage. And we are turning them on at minimal voltage and then increase voltage to 1V. So that might be it... But it doesn't fit problem 100%...

Anyway I will do my best to have some backup plans ready ASAP...

Mudbankkeith

hero member

Activity: 868

Merit: 1000

https://bitcointalksearch.org/topic/m.3321352

refers to revision A to rev B chips and "Firmware changes"

ChipGeek

full member

Activity: 198

Merit: 100

Quote from: Mudbankkeith on October 25, 2013, 06:22:39 AM

Have BFL changed the pinout between batches?

No.

Mudbankkeith

hero member

Activity: 868

Merit: 1000

Quote from: Lucko on October 25, 2013, 02:02:19 AM

Quote from: GrapeApe on October 22, 2013, 10:19:41 AM

Yeah sorry to hear that dwdoc. Hey Lucko is everything okay? Is there anything we should know about?

Well now there is... I'm closing orders for now. Production chips(they have logo on it and they looks to be different inside to) will not work with this board. It worked OK with my 8 samples... Well as good as expected. One of removed chips from Coinado short circuit and I had to disable one 4 chips bank. And since dead chips do some troubles on SPI I can see from 0 to 4 chips. Depends on luck. Now we are not sure what will take to fix this problem. Me and form were up till 1am figuring it out... Board will got to him today so he can look what could be the problem... I also contacted MrTeal who had the same problem how he solved it...

Now for thus who send me chips please let me know what to do.

Have BFL changed the pinout between batches?

bx8389

member

Activity: 80

Merit: 10

Quote from: Lucko on October 25, 2013, 02:02:19 AM

Quote from: GrapeApe on October 22, 2013, 10:19:41 AM

Yeah sorry to hear that dwdoc. Hey Lucko is everything okay? Is there anything we should know about?

Well now there is... I'm closing orders for now. Production chips(they have logo on it and they looks to be different inside to) will not work with this board. It worked OK with my 8 samples... Well as good as expected. One of removed chips from Coinado short circuit and I had to disable one 4 chips bank. And since dead chips do some troubles on SPI I can see from 0 to 4 chips. Depends on luck. Now we are not sure what will take to fix this problem. Me and form were up till 1am figuring it out... Board will got to him today so he can look what could be the problem... I also contacted MrTeal who had the same problem how he solved it...

Now for thus who send me chips please let me know what to do.

From BFL Schematics documentation I see "Revision C.4-Bitforce SC chips Pinout corrected". Have you last BFL schematic ? ( i think is revision E )

Regards

Lucko

hero member

Activity: 826

Merit: 1000

Quote from: GrapeApe on October 22, 2013, 10:19:41 AM

Yeah sorry to hear that dwdoc. Hey Lucko is everything okay? Is there anything we should know about?

Well now there is... I'm closing orders for now. Production chips(they have logo on it and they looks to be different inside to) will not work with this board. It worked OK with my 8 samples... Well as good as expected. One of removed chips from Coinado short circuit and I had to disable one 4 chips bank. And since dead chips do some troubles on SPI I can see from 0 to 4 chips. Depends on luck. Now we are not sure what will take to fix this problem. Me and form were up till 1am figuring it out... Board will got to him today so he can look what could be the problem... I also contacted MrTeal who had the same problem how he solved it...

Now for thus who send me chips please let me know what to do.

Lucko

hero member

Activity: 826

Merit: 1000

Quote from: asjfdlksfd on October 23, 2013, 04:31:03 PM

Hi,

the heatsink looks like not to transports enough power from the chips. What are the stats? And what is the advantage against CPU cooler which is replaced by a simple all-aliminium heat sink?

Indeed I found good air flow cooling solutions here: http://www.kuehlkoerper.de/kammprofile-40-mm-190-mm/
My favorites are solutions with < 0,1 K/W at 10 cm length. This gives a max. of 25 K at 250 W :-)

Your heatsink looks like more a 0,5 - 1 K/W cooling solution, isn't it? So every newer cpu cooler could be better than yours?

Cheers...
Pierre

First plan was make it out of block of aluminium. But it didn't work out so this is a backup option... It is still batter then what we know it works so I guess it will be ok. Will see...

CPU cooler is not big enough. And it is made for cooling one point not 16... So you need a coper plate...

asjfdlksfd

full member

Activity: 128

Merit: 100

Hi,

the heatsink looks like not to transports enough power from the chips. What are the stats? And what is the advantage against CPU cooler which is replaced by a simple all-aliminium heat sink?

Indeed I found good air flow cooling solutions here: http://www.kuehlkoerper.de/kammprofile-40-mm-190-mm/
My favorites are solutions with < 0,1 K/W at 10 cm length. This gives a max. of 25 K at 250 W :-)

Your heatsink looks like more a 0,5 - 1 K/W cooling solution, isn't it? So every newer cpu cooler could be better than yours?

Cheers...
Pierre

Lucko

hero member

Activity: 826

Merit: 1000

We have backplate too... But it was not pack in this box... Some hiccups we have with suppliers... From what I understand it was packing error and they are on the way. It is a 5 year since my uncle died and we always get together on this date in Austria(where he lived) so I didn't have time to go for it myself. But I have 2 possible configuration ready for backplate. One with rubber where chips are and one with rubber in spaces between chips banks... Will see what will work best... And pads are planed so there is some room for hight error...

bx8389

member

Activity: 80

Merit: 10

Quote from: Lucko on October 22, 2013, 05:23:26 PM

Quote from: xselam1988 on October 22, 2013, 04:35:15 PM

Quote from: Lucko on October 22, 2013, 02:21:25 PM

Well... hard to say... We did ask him not to trash it and we will check it... But he never got around to it... So hard to say... If he didn't load firmware on it might even not be toasted... It could even happen that it has good chips on it... Luke-Jr did have a lot of them but cooler was not put on properly so you can guess what happened... I did even post that on forum that he didn't do it right but he didn't fix it...

One thing I dont sure is that if this thing uses more than 300W power Why are you putting a simple CPU cooler on it? Isn't this gonna destroy the chips?

It uses less then 300 and probably less then 250W. And we are no longer using that... Just got first box of custom made ones... It was not made 100% to my specifications(some breking tools problems) so I'll need to test it but I guess it will work...

That black cooler works if it is on properly... It was problem that all chips didn't touch heatsink...

Hello lucko
A massive heatsink like showed on photo due its weight will put a lot of mechanical stress on PCB and possibly bend it if appropriate measures aren´t taken. If PCB isn't flat thermal transfer between chips and heatsink is not even and optimal.
Check w/ form about but is wise to have some mechanical frame in opposite side of pcb to avoid this problem
regards

Topic: BFL board project COINTAMINATION - EU facility - ORDERS OPENED for Chili - page 27. (Read 73149 times)