Pages:
Author

Topic: Hacking a BFL Jalapeno to 32GH and beyond....(???) - page 21. (Read 54329 times)

legendary
Activity: 3164
Merit: 2258
I fix broken miners. And make holes in teeth :-)
Yeah, we have a new heatsink that goes on the bottom of the units that are shipping with newer units that replaces the flat reinforcement backplate.
That is making a nice difference. One of my outstanding questions was why two of the new chips were hashing at 3gh or so while the third was at 4.3gh. With the new sink covering all the chips and wicking away the heat, I am seeing 4.0+ on the two (obviously corner) chips and 4.3 on the inside one.

Very impressive. Moral: If you're going to double your chips, put them on the cross positions so they are all covered by the heat sink. Better yet, get a better heat sink with pipes to remove the heat more efficiently.

This weekend I might try putting mine back in the case and see if it will keep itself cool. Then if I get my chips I'll go for six.

Thanks Inaba!
legendary
Activity: 3164
Merit: 2258
I fix broken miners. And make holes in teeth :-)
A note to everyone watching:

I have been in contact with the gentleman/lady/person who is selling chips here on the forum. Apparently he has a limited supply of them, so if you want chips from this source I highly recommend buying them... soon.

https://bitcointalksearch.org/topic/closed-bfl-4-ghs-chips-06-btc-per-chip-batch-29-236103

Kudos to drug5bitz for finding this person. I bought one (in addition to 3 more coming in from another source) so I should be good to go for now.

Edit: AND all the chips are sold again. What the hell is it with me? :-)

C
full member
Activity: 123
Merit: 100
So I've been looking at reballing stations as well what ya think of this

http://www.ebay.com/itm/like/131046801460?lpid=82
sr. member
Activity: 333
Merit: 250
Wicking in from the sides and tilting the board is how I did Xbox 360 GPUs and it worked well.  They were a much bigger BGA so it should work just fine on these chips.

I may try reflowing the chip I think has cold joints tonight.   I'll report back with what I find.  Will also add a heatsink on the bottom per your suggestion.

legendary
Activity: 3164
Merit: 2258
I fix broken miners. And make holes in teeth :-)
Yeah 120C is what I had it set for.... on top IR heat it just pulses on and off... stays on longer if you set it "higher".

Further info... I float the chips twice now...
1-apply flux
2-set chip
3-IR heating
4-Repeat 123 for the other chip
5-wait 5min
6-more liquid flux on chip 1
7-reheat chip 1 another 2min at 120C
8-more liquid flux on chip 2
9-reheat chip 2 another 2min at 120C
10- turn off heat wait 10 min

I'm picking up my rpi heatsinks (for the 'fets), some heat pads(chip heights), some recycled big ol' aluminum heat sinks for the bottom... gonna sit all 3 on a laptop cooler:)

I'll keep you posted...
Well, looks like I am going to Ebay and will be spending another chunk of change for one of these. That's a lot better than searing chips with the flamethrower.

How did you get more flux on the chips? Dump it around the chip and let it wick in?

C
full member
Activity: 139
Merit: 100
Yeah 120C is what I had it set for.... on top IR heat it just pulses on and off... stays on longer if you set it "higher".

Further info... I float the chips twice now...
1-apply flux
2-set chip
3-IR heating
4-Repeat 123 for the other chip
5-wait 5min
6-more liquid flux on chip 1
7-reheat chip 1 another 2min at 120C
8-more liquid flux on chip 2
9-reheat chip 2 another 2min at 120C
10- turn off heat wait 10 min

I'm picking up my rpi heatsinks (for the 'fets), some heat pads(chip heights), some recycled big ol' aluminum heat sinks for the bottom... gonna sit all 3 on a laptop cooler:)

I'll keep you posted...
full member
Activity: 123
Merit: 100
That was another idea I had, how about changing out critical components such as the mosfets as well?  I really wouldn't know but they seem to be the weak link for now. That and getting a real power supply for these things. 

I lost faith that my 4 jalapenos (just in the mail now). Thinking I ended up wasting my money (which inevitably i did). If I have anything to say about it I'm going to have 4 little singles. Water cooled possibly if it seems beneficial. Kinda tend to be a junkie with water and PC's, I just don't know if  the added dissipation is going to help more than just cool the fets a litttle. The chips clock up in the start up, and after they're running only down clock if to much heat is present right?
legendary
Activity: 1260
Merit: 1000
It depends on which MOSFETs you have.  We have had a lot of trouble sourcing large quantities of MOSFETs, so we've had different ones on there. Some run much, much cooler than others.
legendary
Activity: 3164
Merit: 2258
I fix broken miners. And make holes in teeth :-)
Cheapo China special T862+ was marketed for bga replacement (yeah right with the tiny bottom heater it has)... however it is perfect for mobile phone repairs... and hacking Jalapeno's!!

Interesting. Off to Ebay to buy one; they look like garbage, but why not. :-)

It really only needs 120 degrees C of heat on top to float the chips? Or is that 320C

Quote
Flux is Kester 951 no clean flux from dickiesgaragesale.com  ... I bought it from his ebay store...

edit just finished my second jally... clocks in at just under 16gh... a bit too late to do the last one but I need to pick up some supplies so I can just put 5 on all 3... I'll keep you posted...

I am so tickled that this works....

Quote
just an fyi... 1 jally at 15.9gh@46C, 1 jally at 13.2gh@41C.... the last was at 14.7 and 10% HW errors (forgot to turn back on a single error check no reason to waste the energy and stress it for no extra hashes)
That's about right, now grab a cheap-o IR thermometer and check the temps on those FETs. With 4 it's not a problem, going to 5 starts to push into odd places. I assume this is running with open case, fan down, no sink on the bottom?

Good work!
C
full member
Activity: 139
Merit: 100
Just an update...

I used my own modded 1.2.9 firmware first... worked at 6.4gh... crappy chips...

Bottom heat at 100C for about 10min

Applied no clean liquid flux  (edit)

IR lamp set to 120C (pulses on and off)
about 2.5min (x2) for chip to drop (each one)

Let stand for about 10min

Assemble

Enjoying 14.7gh at about 48C (nice cold basement! )

Yum!
120 degrees C? Man what are you using, and where can I get something like that? Also what flux did you use?

Cheapo China special T862+ was marketed for bga replacement (yeah right with the tiny bottom heater it has)... however it is perfect for mobile phone repairs... and hacking Jalapeno's!!

Flux is Kester 951 no clean flux from dickiesgaragesale.com  ... I bought it from his ebay store...

edit just finished my second jally... clocks in at just under 16gh... a bit too late to do the last one but I need to pick up some supplies so I can just put 5 on all 3... I'll keep you posted...

just an fyi... 1 jally at 15.9gh@46C, 1 jally at 13.2gh@41C.... the last was at 14.7 and 10% HW errors (forgot to turn back on a single error check no reason to waste the energy and stress it for no extra hashes)
legendary
Activity: 3164
Merit: 2258
I fix broken miners. And make holes in teeth :-)
Wow yeah it does all kind of make sense. All the pics from early have 1 chip.  Also from what I've seen the mini rigs are water cooled. I think it's 8 (prolly 6) chips on a board sandwiched between two water blocks. If I can do that, that would be a thread as well. Did you mount the copper sink right to the back plate and add thermal paste? Could I get a better pic of your mounting setup? Also whats the surface size the chips cover like 50x50mm for all of the pads?
Close. The copper sink I had came from an IBM x366 system, big ole Pentium XEON chips. I wish I had saved another sink, but you know how it goes.

The surface plate is too big to fit on the back of the board directly because there are 8 little vias to the left of the chips, and those are copper on the bottom. Short out potential. So I initially drilled two dimples in the heat sink, and stuck the jally with the AL plate on it with some cheap-o Radio shack heat sink compound. Worked well, the aluminum heat sink gets *HOT* and it was wicking heat into the plate.

Then I said "Hm", went out to the shed and drilled two holes into it. Removed two fins from the centerline, and ran two screws through it, AL plate, board, crummy AL sink on top. Works *well*.

Then I said "hm hm" and put heat sink compound on both sides of the AL plate (board and sink). IT WICKS A LOT OF HEAT AWAY and solves the heat problem completely. If I got a router and put a channel into the copper plate so it would clear those vias I would just mount it to the whole back of the unit and call it a day.

I don't quite recall the surface size, it might be in the BFL specifications PDF for the little single. Take yours apart and check with a set of calipers, actually you can see the chip outlines on the back so measure it there without removing the heat sink. Go as big as you can, and if possible get it under the FETs on the side over to the CPU board. That's the heat spot.

I'll shoot some better pics this weekend, this little guy is almost a work of art at this point.

C
full member
Activity: 123
Merit: 100
Wow yeah it does all kind of make sense. All the pics from early have 1 chip.  Also from what I've seen the mini rigs are water cooled. I think it's 8 (prolly 6) chips on a board sandwiched between two water blocks. If I can do that, that would be a thread as well. Did you mount the copper sink right to the back plate and add thermal paste? Could I get a better pic of your mounting setup? Also whats the surface size the chips cover like 50x50mm for all of the pads?
legendary
Activity: 3164
Merit: 2258
I fix broken miners. And make holes in teeth :-)
Oh as far as water cooling? Forget the top, water cool the bottom of the board. The chips can sink an amazing amount of heat through the board, it's flat and with the exception of a few copper vias pretty perfect for a nice big plate that can cool the chips, the FETs, half of everything. Then just put a normal sink and fan on top and go to town.

C
legendary
Activity: 3164
Merit: 2258
I fix broken miners. And make holes in teeth :-)
Just an update...

I used my own modded 1.2.9 firmware first... worked at 6.4gh... crappy chips...

Bottom heat at 100C for about 10min

Applied no clean liquid flux  (edit)

IR lamp set to 120C (pulses on and off)
about 2.5min (x2) for chip to drop (each one)

Let stand for about 10min

Assemble

Enjoying 14.7gh at about 48C (nice cold basement! )

Yum!
120 degrees C? Man what are you using, and where can I get something like that? Also what flux did you use?
legendary
Activity: 3164
Merit: 2258
I fix broken miners. And make holes in teeth :-)
Check this out, looking around to help.

http://imgur.com/a/CMazz#0

Every BFL PCB with chip layout. Cheesy

P.S. I'm looking at water cooling options could you find me dimensions on the chip area (would a 50x50mm block work?) and the hole centers for the heat sink mount.
Oh those... pictures... are interesting. They look like BFL pictures, what were they thinking?

Let's see:

The jally board has one chip. Way more importantly, it has *no* MOSFETs on the 1 volt supply line. Those guys were going to power it off the USB line like some big Erupter. Well, we know that didn't work. But very interesting, no capacitors, no inductors, no nothing. That board would not have been able to do anything interesting.

The little single board is also interesting. 4 chips. 30 GH was the rated speed, I think they were either smoking something or the chips did not do quite as much as they thought they would. One would need 7+gh in order to pull 30gh with 4 chips, and it took eight to make 30, and even that is stressing the board's power supplies.

THAT is what happened. They totally mis-read the power consumption on the chips. They thought you could run a chip at super low power and 4.5gh for the jally, but it turned out they needed full power plus. So they had to use fully powered SC boards plus two chips because one wouldn't quite hash it (hah, I kill myself). Which sank the possibility of passive cooling, so into the dustbin went the coffee cup warmer cases.

The single/30 needed 8 chips, and you can't even run 4 chips in the box there and expect it not to overheat. Believe me, I tried, fail. Even with all eight, the power FETs were overheating so they had to add bottom cooling. Can't do that in a little single box, so they had to build the bigger box with space under it and a side fan. Made one box for 30 and 60 (dual 30's) at which point they just said "screw it, we'll use the little single boards and mid-quality chips for the jalapenos, offer a "bump up" to try to recoup costs by putting two grade A chips, and delay the singles and 60's.

And that is exactly what happened. Although running at full 30gh is at the edge of what the single board can handle without blowing FETs so they wrote code to limit the hash speed to 30+.04%. And they slowly shipped 30's and 60's, which explains why people there were grumpy.

But imagine being in the BFL factory: You have to put down twice as many chips, had to scrap a whole pile of jally cases, and worst you had to fill jalapeno orders with much more expensive quality cases, fans, power supplies, and fuck knows what else including boards with full power support for a product that cost several times as much.

And THEN you have to build new cases, double the heatsink counts, chip counts, and run the thing at the edge to meet the 30gh "promise".

Now your jallies hit the streets, and people start modding them. Word gets out, BFL leaves a lot of performance there, boost it up dude! The jally people can get away with it, and then the single/30 and single/60 people start doing it....

If the single/30 had split the chips across both sides of the board the users might have gotten away with hot-dogging it to 35+gh. But BFL was already way under the money rock with this; they needed to keep costs down so they sold the other side of the board with no components. Not needed. And they reduced the new speed sold to 25gh, which is much safer and will probably last forever.

So the Single/30 and Single/60 people load up code revs with no restrictions. Power draw goes way up, FETs explode, and people start bitching on the forums. Because they thought they could make their singles faster, expected it.

Ooops. Poor BFL. They have really tried to do the right thing here, even when many other companies said fuck it and quit. I do have to give them a lot of credit here, and I think the bashing is missing the point. They weren't malicious, indeed their solution was pretty damn impressive.

Maybe this is what happened. But the chips and pics and stories are all coming together with what I am seeing here totally independently. I wouldn't go over 6 chips....

Fascinating.
full member
Activity: 139
Merit: 100
Just an update...

I used my own modded 1.2.9 firmware first... worked at 6.4gh... crappy chips...

Bottom heat at 100C for about 10min

Applied no clean liquid flux  (edit)

IR lamp set to 120C (pulses on and off)
about 2.5min (x2) for chip to drop (each one)

Let stand for about 10min

Assemble

Enjoying 14.7gh at about 48C (nice cold basement! )

Yum!
legendary
Activity: 3164
Merit: 2258
I fix broken miners. And make holes in teeth :-)
speed may be incorrect, there were no work units.  mosfets were uncooled, but it was not on long enough to matter.  the unit currently powers on, but does not detect.  and will not allow Jtag to get into program mode.  I had someone perform the work for me, so I am passing along what I was told.

sadly, no pics.  

just figured I would get your take, since you have adding for a while.  
Well hm....

So let's think about this: The jally was a (stock code or not-stock code?) unit, two chips were added, it came up, and now does nothing. Let's try some diagnostics:

1) Does the front light flash on power-up and then go solid?
2) Do any lights on the back come on? There should be the power LED by the power plug, then at least two of the eight LEDs along the back between the power plug and the USB plug.

It should light the rear power LED, then flash the front light, then stop flashing the front light and light the appropriate number of rear LEDs from that set of eight. That indicates the system is up and running.

Nothing we did here should affect the JTAG port; have you ever had it programmed before now? In all my cases, I could always erase memories, then reprogram. Yep, checked my notes, even when the thing was dead as a doornail.

Can you get pics of the board without the heat sink on it? Something's odd.

Before I did my first chip, I knew I had to go to 1.2.9 firmware. I tried Tarkin's code stock and it failed my Jally before I even put another chip on. I had to realize that one has to define LITTLE_SINGLE and not SINGLE and compile to get things to work.

If you have the 8 pin programmer on backwards you will get the 0 volt thing. Always a good idea to do stuff one step at a time, never change two things (new programming and new chips) at once if you can help it.

I'd say have the person check the LEDs and report back, and make absolutely sure pin 1 on the dragon JTAG is on the red wire side of the 10 pin connector, and the jally's pin 1 is on the red wire side of the 10 pin connector. On mine the dragon has to be upside down in order for the cable not to have a 180 twist in it. If this person is going straight he (or she) is doing it wrong.

If you're still sunk, drop me a line and maybe I can take a look at your unit. The JTAG thing makes no sense, even shorting the +1 supply didn't stop that from coming up (and I spent hours doing that one fun night)

C
hero member
Activity: 699
Merit: 504
speed may be incorrect, there were no work units.  mosfets were uncooled, but it was not on long enough to matter.  the unit currently powers on, but does not detect.  and will not allow Jtag to get into program mode.  I had someone perform the work for me, so I am passing along what I was told.

sadly, no pics. 

just figured I would get your take, since you have adding for a while. 
legendary
Activity: 3164
Merit: 2258
I fix broken miners. And make holes in teeth :-)
Hello LightFoot, 

I have a question for you.  I recently tried to follow your footsteps, and mod a Jally.  I dropped in 2 chips.  When I powered on, it looked like all was going smoothly....Jally was up to 25ghs...and all the sudden it konk'd out.  it still powers on, but just gives errors.  When I try to Jtag it, it will not allow me to.  It indicates volts are 0.0.   I am hoping you may be able to shed some light, if I still have hope or if I am super F'd...LoL


thanks

Hmmmm. Can you post pics what were the temps how.was it cooled and how did you get 25gh on 4 chips total?

Check the fets and does it smell odd.

C
hero member
Activity: 699
Merit: 504
Hello LightFoot, 

I have a question for you.  I recently tried to follow your footsteps, and mod a Jally.  I dropped in 2 chips.  When I powered on, it looked like all was going smoothly....Jally was up to 25ghs...and all the sudden it konk'd out.  it still powers on, but just gives errors.  When I try to Jtag it, it will not allow me to.  It indicates volts are 0.0.   I am hoping you may be able to shed some light, if I still have hope or if I am super F'd...LoL


thanks
Pages:
Jump to: