Got my JALAPENO today mining at 5.9 Gh/s with some tweaks

[julz]

Purchased by rjk who sold his preorder to sadpandatech who then sold to me.

SO this was real product that was shipped not a sale given priority for a PR reason??  IOW - a trickle of jalepenos is actually leaving bfl still - which of course is great news..

Yeah - this week I received 2 units here in Australia - and I was suprised to find a nice pair of Alpaca socks included! 
I ordered on the 2nd day.

[PuertoLibre]

3. Flip fan around so that it is blowing down

Wait, so the fan was originally pulling air through the heatsink? BFL engineering excellence right there.

I also found that strange.

By the way, did he pay for 7 Gh/s?

If not, then people may very well get more than 5 for free....why pay more for the upgrade?

[vapourminer]

interesting, Ill have to try that.

my jally runs 5.56 ghs (2 day ave) @ 38 C and 3.2 v with a 140 mm fan blowing at its intake.

without the fan it was around 44 C and 3.35 v. cant remember the exact hash rate though.

seems the warmer it gets the more volts it gets.

really wish bfl would pay attention to cooling over looks. its so noisy no one would put it on their desk anyway, my single and jally are in the basement.

[Enigma81]

Unfortunately no. You want to increase the air that moves over the entire surface of the heat sink. If you suck, the air will take the shortest path possible through the heat sink blades leaving some surface area with sub-optimal surface flow. If you blow air, it forces air into every nook and cranny of the heat sink increasing airflow over the entire surface. Proper heat sink blade design will prevent back-pressure and eddies.

{Rolls Eyes}

[vapourminer]

I noticed there are 5 HW errors in that photo. Might want to look into that.

looked at percentage wise its not significant.

my jally in 2 days has had 4 hw errors.

0.000060% if Im doing the math anywhere near right.

[erk]

I noticed there are 5 HW errors in that photo. Might want to look into that.

looked at percentage wise its not significant.

my jally in 2 days has had 4 hw errors.

0.000060% if Im doing the math anywhere near right.

Yeah that's insignificant. I wouldn't want to see more than 1 per hour on a device like an ASIC or I would be asking the software dev for advice.

[Cluster2k]

Skeptical post is skeptical.

A bunch of people have received Jalapenos now. 

35 on this list: http://bfl.ptz.ro/

First two days of Jalapeno orders filled.  Almost a year to go.

[PuertoLibre]

Skeptical post is skeptical.

A bunch of people have received Jalapenos now. 

35 on this list: http://bfl.ptz.ro/

First two days of Jalapeno orders filled.  Almost a year to go.

How long did it take to fill out 2 days of Jalepenos?

Should I start counting from April 1st or May 1st?

60 days to fulfill 2 days worth of partial orders....yikes.

[vapourminer]

interesting, Ill have to try that.

my jally runs 5.56 ghs (2 day ave) @ 38 C and 3.2 v with a 140 mm fan blowing at its intake.

without the fan it was around 44 C and 3.35 v. cant remember the exact hash rate though.

seems the warmer it gets the more volts it gets.

OK tried it. open case, fan blows down. EDIT: same 140 mm fan blowing at it. ambient 20 C

one other thing I did was remove the heatsink and dab some MX2 around where the now broken up thrmal tape was. originally I was going to use straight MX2 without thermal tape but the chips are bare and I was worried about cracking them. so I left the pieces on and filled it in with MX2. I also applied MX2 to the PCB opposite the 2 hashing chips so the load plate would maybe pull some heat off the back of the PCB.

end result: 33 C temps, 3.1 v. same hash rate though.

[erk]

OK tried it. open case, fan blows down.

one other thing I did was remove the heatsink and dab some MX2 around where the now broken up thrmal tape was. originally I was going to use straight MX2 without thermal tape but the chips are bare and I was worried about cracking them. so I left the pieces on and filled it in with MX2. I also applied MX2 to the PCB opposite the 2 hashing chips so the load plate would maybe pull some heat off the back of the PCB.

end result: 33 C temps, 3.1 v. same hash rate though.

33C would feel cold to touch, hardly worth having a fan!

[vapourminer]

33C would feel cold to touch, hardly worth having a fan!

heh.

20 C ambient in my basement. also forgot to mention the 140 mm fan is still blowing at it (edited my above post to reflect this).

[k9quaint]

Unfortunately no. You want to increase the air that moves over the entire surface of the heat sink. If you suck, the air will take the shortest path possible through the heat sink blades leaving some surface area with sub-optimal surface flow. If you blow air, it forces air into every nook and cranny of the heat sink increasing airflow over the entire surface. Proper heat sink blade design will prevent back-pressure and eddies.

{Rolls Eyes}

Did you have a specific criticism? Or were you just disagreeing with who posted the information?

[vapourminer]

in case anyone wondered what the guts look like, heres a (crappy) pic:

[GenTarkin]

there any way to overclock them?

[jhansen858]

Skeptical post is skeptical.

A bunch of people have received Jalapenos now.  

35 on this list: http://bfl.ptz.ro/

Please dont accept this information as fact...

1) the information submitted is 100% honor system
2) I have been marking them as delivered based purely on the order dates that have been submitted by matching up what BFL_Jody says compared to what some random person on the internet claims they ordered and when.  
3) people frequently put fake orders in showing them ordering 50 mini rigs.  I just delete those, however, if somone was just putting what looked like a reasonable order in, I would have no way to verify it and would just add it.  It is entirely possible that some of these orders are not real.

If anyone see's any errors / problems with this data, please PM me and I'll correct it.  

[wrenchmonkey]

Skeptical post is skeptical.

A bunch of people have received Jalapenos now.  

35 on this list: http://bfl.ptz.ro/

Please dont accept this information as fact...

1) the information submitted is 100% honor system
2) I have been marking them as delivered based purely on the order dates that have been submitted by matching up what BFL_Jody says compared to what some random person on the internet claims they ordered and when.  
3) people frequently put fake orders in showing them ordering 50 mini rigs.  I just delete those, however, if somone was just putting what looked like a reasonable order in, I would have no way to verify it and would just add it.  It is entirely possible that some of these orders are not real.

If anyone see's any errors / problems with this data, please PM me and I'll correct it.  

With as much BFL haterade gets drank around here, I wouldn't be surprised if half of those "orders" were faked. Not to mention that there's no way for people to update their information indicating their orders have been received.

[ct1aic]

That's great news. But at this velocity, I'll get mine in 2015...

2827   2013/04/30   100049721   No   1   0   0   0      ct1aic   Undelivered

[Xian01]

The Jalapeno case doesn't allow for a lot of good air flow.
When running out of the box I was getting 5.8 - 5.9 when it first started mining and then dropped to 5.6-5.7 Gh/s after it warmed up. Soto increase air flow I did the following:
<snip>
Getting 5.9 - 6.0 Gh/s now

 Nice work. Looks much happier "out in the open" like that with plenty of room to breathe.

[KS]

The Jalapeno case doesn't allow for a lot of good air flow.
When running out of the box I was getting 5.8 - 5.9 when it first started mining and then dropped to 5.6-5.7 Gh/s after it warmed up. Soto increase air flow I did the following:
<snip>
Getting 5.9 - 6.0 Gh/s now

 Nice work. Looks much happier "out in the open" like that with plenty of room to breathe.

No wonder singles are still MIA.

They probably need to machine out a whole side before the chips will be able to "breathe" enough and that's only with *2* chips in.

[wrenchmonkey]

That's great news. But at this velocity, I'll get mine in 2015...

2827   2013/04/30   100049721   No   1   0   0   0      ct1aic   Undelivered

I'm right there with you, man.  

Hopefully things start gaining momentum. It does seem to be going that way; lots of these things starting to show up in the wild.