Pages:
Author

Topic: The Habanero - 650GH/s - OOS - page 19. (Read 96051 times)

newbie
Activity: 27
Merit: 0
June 24, 2014, 12:29:06 PM
Not sure what I am missing, but what screw are you using to attach the Nepton 280L cooling head to the board, nothing seems to fit.
I had to go to home depot and get 8x 3/4" 6-32(think thats what they are, well normal coarse thread PC) screws

Same here. I actually prefer this method to the thumbscrews, have more control and easier to make small adjustments.
http://i.imgur.com/R8Sytcj.jpg


Thanks.. Got it sorted Smiley


Anyone know what is this error?

 [2014-06-25 00:22:11] Hotplug: Hashfast added HFA 50
 [2014-06-25 00:22:12] HFA 50 HFUSBInit usb write err:(-7) LIBUSB_ERROR_TIMEOUT
 [2014-06-25 00:22:12] HFA 50 attempted reset got err:(0) LIBUSB_SUCCESS
 [2014-06-25 00:22:12] HFA : send_packet: OP_USB_INIT USB Send error, ret 0 amount 0 vs. length 8
 [2014-06-25 00:22:17] HFA 0: No response to name query - failed init or firmware upgrade required.
 [2014-06-25 00:22:17] Hotplug: Hashfast added HFA 51
 [2014-06-25 00:22:18] HFA 51 HFUSBInit usb write err:(-7) LIBUSB_ERROR_TIMEOUT
 [2014-06-25 00:22:18] HFA 51 attempted reset got err:(0) LIBUSB_SUCCESS
 [2014-06-25 00:22:18] HFA : send_packet: OP_USB_INIT USB Send error, ret 0 amount 0 vs. length 8
 [2014-06-25 00:22:23] HFA 0: No response to name query - failed init or firmware upgrade required.
 [2014-06-25 00:22:23] Hotplug: Hashfast added HFA 52
 [2014-06-25 00:22:24] HFA 52 HFUSBInit usb write err:(-7) LIBUSB_ERROR_TIMEOUT
 [2014-06-25 00:22:24] HFA 52 attempted reset got err:(0) LIBUSB_SUCCESS
 [2014-06-25 00:22:24] HFA : send_packet: OP_USB_INIT USB Send error, ret 0 amount 0 vs. length 8
sr. member
Activity: 358
Merit: 255
June 24, 2014, 06:53:50 AM
Not sure what I am missing, but what screw are you using to attach the Nepton 280L cooling head to the board, nothing seems to fit.
I had to go to home depot and get 8x 3/4" 6-32(think thats what they are, well normal coarse thread PC) screws

Same here. I actually prefer this method to the thumbscrews, have more control and easier to make small adjustments.
http://i.imgur.com/R8Sytcj.jpg
legendary
Activity: 2450
Merit: 1002
June 24, 2014, 01:33:43 AM
Not sure what I am missing, but what screw are you using to attach the Nepton 280L cooling head to the board, nothing seems to fit.
I had to go to home depot and get 8x 3/4" 6-32(think thats what they are, well normal coarse thread PC) screws
newbie
Activity: 27
Merit: 0
June 24, 2014, 01:23:03 AM
Not sure what I am missing, but what screw are you using to attach the Nepton 280L cooling head to the board, nothing seems to fit.
legendary
Activity: 2450
Merit: 1002
June 24, 2014, 12:58:01 AM
So for anyone struggling to keep temps under control... I just applied some Liquid Pro(100% liquid metal, till it cures then its more solid) tonight on my 2 habs....
Unfuckingbelievable!!!!
Temps dropped from 100-105C down to 83-88C (on hottest dies)

Running a nepton 280L, stock fans, well cept for one, on just one of the habs.

BTW, upon remounting I noticed that I used too much pressure on these chips, on one of my habs, one die had a tiny tiny chip in the outermost corner and the die opposite from it had a larger chip. There must be a bit of safety padding around the actual logic on these. Like good ol fashion AMD socket A processors =), could chip entire corners out of their die and theyd run fine.

Mounting them the second time around w/ liquid pro, used much less pressure and eyed the evenness of the mounting points. Using the pepper app .. was able to easily dial in acceptable temps for all the dies, I think the liquid pro is LARGELY responsible for how it easy it was this time.
SVK
sr. member
Activity: 378
Merit: 250
June 24, 2014, 12:17:15 AM


Hello...anybody home... Huh

Any progress...production...sales...new batches...any other exciting news...?

ZiG

With 26% difficulty increase it's going to be hard. If I remember correctly they were waiting for bigger order confirmation nearly 1 1/2 week ago. No word might mean that it fell through.
ZiG
sr. member
Activity: 406
Merit: 250
June 23, 2014, 06:54:57 PM


Hello...anybody home... Huh

Any progress...production...sales...new batches...any other exciting news...?

ZiG
member
Activity: 84
Merit: 10
Kinex - The New Frontier
June 23, 2014, 06:49:17 PM
hero member
Activity: 617
Merit: 543
http://idontALT.com
June 22, 2014, 04:29:01 AM
yay!! i am done with testing my boards out. the dual Habaneros are now steady at:
Code:
 cgminer version 4.3.5 - Started: [2014-06-21 17:43:15]
----------------------------------------------------------------------------------------------------
 (5s):1.510T (1m):1.431T (5m):1.377T (15m):1.368T (avg):1.368Th/s
 A:29061566  R:26018  HW:84868  WU:19116.1/m | ST: 64  SS: 88  NB: 216  LW: 31241145  GF: 2  RF: 0
 Connected to us.ozco.in diff 579 with stratum as user quiveringgibbage.habanero
 Block: 2dac8ebc...  Diff:13.5G  Started: [19:00:35]  Best share: 77.4M
----------------------------------------------------------------------------------------------------
 [U]SB management [P]ool management [S]ettings [D]isplay options [Q]uit
 0: HFB FATALII : 937MHz  95C 100% 0.87V  | 632.6G / 700.2Gh/s WU:9782.2/m A:14800033 R:15027 HW:24303
 1: HFB FRUTESCE: 937MHz  93C 100% 0.88V  | 680.6G / 668.1Gh/s WU:9334.0/m A:14261533 R:10991 HW:60565
----------------------------------------------------------------------------------------------------
 [2014-06-22 19:10:42] Accepted 433b1d5d Diff 975/579 HFB 1 pool 0
 [2014-06-22 19:10:46] Accepted 2c4bfdee Diff 1.48K/579 HFB 0 pool 0
 

24hrs no reboots. if i give it any more they hash less. Pulling 1750W at the wall (240V).
Pics can be found here: http://imgur.com/a/SMBJP#25




I still have 1 more Habanero humming away in my room. going to try to Rpi as control box.
Code:
cgminer version 4.2.3 - Started: [2014-06-22 01:00:22]
----------------------------------------------------------------------------------------------------
 (5s):554.5G (1m):640.3G (5m):673.3G (15m):658.6G (avg):651.5Gh/s
 A:9984421  R:10536  HW:5553  WU:9101.4/m | ST: 5  SS: 131  NB: 163  LW: 10853007  GF: 0  RF: 0
 Connected to us.ozco.in diff 579 with stratum as user quiveringgibbage.habanero
 Block: 1a934abf...  Diff:13.5G  Started: [19:23:10]  Best share: 7.74M
----------------------------------------------------------------------------------------------------
 [U]SB management [P]ool management [S]ettings [D]isplay options [Q]uit
 2: HFB JOLOKIA : 912MHz  86C 100% 0.87V  | 492.8G / 637.1Gh/s WU: 8904.2/m A: 637064 R:   0 HW: 335
----------------------------------------------------------------------------------------------------
 [2014-06-22 19:23:56] Accepted 23f41caa Diff 1.82K/579 HFB 0 pool 0

Pic of the rest of my h/w here: http://imgur.com/a/mRbMZ

Cheers,
QG
hero member
Activity: 857
Merit: 1000
Anger is a gift.
June 21, 2014, 11:24:59 AM
It may not be pretty, but moving the radiators off my desk and into my shelving dropped temps 3C. Hottest cores are 96C instead of 99C. Have a box fan blowing on the radiators and my 6 Module Jupiter, which is on the 2 racks above these.

member
Activity: 86
Merit: 10
June 19, 2014, 10:31:33 PM
Not sure yet, I haven't figured it out. They're just standard FCI PowerBlade connectors for AC input and DC output though.
Figured it out now, I'll post up something later. Other than being hugely long, they're actually pretty cool little supplies and are amazingly quiet at low power levels.

Edit:
This is the connector for AC Input.
http://www.digikey.com/product-search/en?vendor=0&keywords=51940-059LF
Pinout is L-G-L

This is the DC output connector on the PSU.
http://portal.fciconnect.com/Comergent/fci/drawing/51939-186.pdf
I haven't found the mate for that one yet, but to turn it on connect G2 (PSKILL) to U2 (5Vsb), and U3 (PS_ON) to T1 (GND).

Edit 2:
This is the DC side connector
http://portal.fciconnect.com/Comergent/fci/drawing/51940-160.pdf

THANK YOU! you are the best! I want in on your next batch.
legendary
Activity: 1274
Merit: 1004
June 19, 2014, 04:22:05 PM
Not sure yet, I haven't figured it out. They're just standard FCI PowerBlade connectors for AC input and DC output though.
Figured it out now, I'll post up something later. Other than being hugely long, they're actually pretty cool little supplies and are amazingly quiet at low power levels.

Edit:
This is the connector for AC Input.
http://www.digikey.com/product-search/en?vendor=0&keywords=51940-059LF
Pinout is L-G-L

This is the DC output connector on the PSU.
http://portal.fciconnect.com/Comergent/fci/drawing/51939-186.pdf
I haven't found the mate for that one yet, but to turn it on connect G2 (PSKILL) to U2 (5Vsb), and U3 (PS_ON) to T1 (GND).

Edit 2:
This is the DC side connector
http://portal.fciconnect.com/Comergent/fci/drawing/51940-160.pdf
hero member
Activity: 784
Merit: 501
June 19, 2014, 02:13:48 PM
Any plans for another batch of these?
hero member
Activity: 539
Merit: 500
June 19, 2014, 02:11:49 PM
These things are built like tanks, aren't they? Cool

Agreed.  Both of mine should be very broken based on what I've done to them.  Lucky for me and my wallet, they're hashing nicely 24/7 @ 700+ gh/s.
member
Activity: 69
Merit: 10
June 19, 2014, 12:08:02 PM
I'm running cgminer on an ARMv7 linux (ubuntu)  platform.
hero member
Activity: 857
Merit: 1000
Anger is a gift.
June 19, 2014, 12:04:25 PM
Help!
One of my habbies died this morning. They were both happily hashing away for days @825 when I noticed that one of the habbies stopped hashing and went zombie. Tried resetting everything but same result. Tried turning hash clock down to 500 and same result. The board powers up fine, but when it's asked to hash it appears to go through a reset, over and over, each time cgminer detects it and tries hashing.
Anyone have any ideas for me to try?


What OS are you using? If it is windows, run Zadig again and it should fix it.
member
Activity: 69
Merit: 10
June 19, 2014, 11:59:28 AM
Help!
One of my habbies died this morning. They were both happily hashing away for days @825 when I noticed that one of the habbies stopped hashing and went zombie. Tried resetting everything but same result. Tried turning hash clock down to 500 and same result. The board powers up fine, but when it's asked to hash it appears to go through a reset, over and over, each time cgminer detects it and tries hashing.
Anyone have any ideas for me to try?

*edit*
I'm getting this error from cgminer:
[2014-06-19 17:03:44] ERR: Asked to memcpy 0 bytes from usbutils.c _usb_read():3065
[2014-06-19 17:03:44] HFA : OP_USB_INIT failed! Operation status 20 (Regulator programming error)

*edit 2*
Oops, never mind. Embarrassed
Just had a loose PCI-e connection.

These things are built like tanks, aren't they? Cool
 
legendary
Activity: 1274
Merit: 1004
June 19, 2014, 01:20:26 AM
Not sure yet, I haven't figured it out. They're just standard FCI PowerBlade connectors for AC input and DC output though.
legendary
Activity: 1274
Merit: 1004
June 19, 2014, 12:24:21 AM
So, quick update again in my journey of discoving how this core shedding works:
noshed enabled: 8 cores(per hab) get disabled on start of mining.
noshed disabled: ALL cores enabled on start of mining, over 24hr period, one of my habs incremented "shed count" to 2

Theoretically, if ur cores arent complete shit, you will benefit from running w/o noshed!

It seems to be very common(as in Ive had several people report this to me) that using noshed...causes 8 cores to get disabled right from the start, per GN!!! why?! no fucking clue, but they work fine!
That's interesting, and that it appears to be the same four cores on dies 0 and 1. I would say it's likely a FW issue that we'll have to track down.
legendary
Activity: 2450
Merit: 1002
June 18, 2014, 10:54:00 PM
So, quick update again in my journey of discoving how this core shedding works:
noshed enabled: 8 cores(per hab) get disabled on start of mining.
noshed disabled: ALL cores enabled on start of mining, over 24hr period, one of my habs incremented "shed count" to 2

Theoretically, if ur cores arent complete shit, you will benefit from running w/o noshed!

It seems to be very common(as in Ive had several people report this to me) that using noshed...causes 8 cores to get disabled right from the start, per GN!!! why?! no fucking clue, but they work fine!
Pages:
Jump to: