NiteShdw
Member
Offline
Activity: 69
Merit: 10
|
|
August 20, 2013, 05:21:23 PM |
|
Is there any significant (>1%) performance difference between 20130703 and 20130819 (or 20)? I'm trying to keep the reflashing to a minimum. Are there changelogs somewhere for ckolivas' firmwares?
I just upgraded last night and my "--avalon-auto" clocks are up at least 5mhz. My pool is reporting my workers running at 260GH/s versus 242GH/s before the upgrade.
|
|
|
|
ProfMac
Legendary
Offline
Activity: 1246
Merit: 1001
|
|
August 20, 2013, 05:40:21 PM |
|
I used the re-flashed retail unit with no hardware modification for an hour or so. The mini-USB supplied power. The Avalon USB plugged into the side. I removed one of the screws from the top of the Avalon and fed the USB cable through the slot, and mounted the TP-Link outside the case.
I later modified this unit, I removed a 0R resistor to disconnect the internal antenna, and added a 0R resistor to backfeed power from the Avalon USB connector, and added the antenna, and replaced the dead internal unit, and closed it all up.
Does the blue light stay on with your TP-Link? If you push the reset button while it is starting up, does the blue LED flash about twice/second? Are you trying to connect via the LAN jack? I think the WiFi has to be configured.
Can't the original TP-link be send to TP-link for reflashing? Maybe for a small fee or so? There is a serial connection on the board of the TP-Link to do a serial boot. The OpenWRT community is very familiar with this process. The gotcha with the TP-Link is that those traces are very fragile, and if someone tacks wires to it and does not provide strain relief, they will rip the traces off the board. I might be willing to sell a spare 703n flashed with the ckolvias latest release firmware, and to buy a distressed unit. Shipping would cost almost as much as a buying a new unit, then there is my time, so I am an expensive, last resort option.
|
I try to be respectful and informed.
|
|
|
candoo
|
|
August 20, 2013, 09:18:11 PM |
|
Hey,
did anyone manage to install the avalon openwrt firmware on a TP-Link TL-MR3020 ?
|
Einer trage des andern Last, so werdet ihr das Gesetz Christi erfüllen.
|
|
|
jermwerty
|
|
August 21, 2013, 12:48:29 AM |
|
Actually turns out the 850watt didn't burn out. After 2 weeks, I think it overheated and shutoff when my ac was turned off. ... Still though, I ordered a corsair hx 1050 gold to serve as primary psu.
Well that is good to hear! Just FYI the Corsair HX 1050 is about 1cm too long to fit inside. That's why I was hoping the V1000 would fit, but I am not sure if all the connectors will clear. It may work with using 2 out of the 3 PCI-E cables, not sure. Any problem with leaving just PCI-E 3 disconnected? Not sure how the power board distributes power.
|
|
|
|
thejestre
Member
Offline
Activity: 76
Merit: 10
|
|
August 21, 2013, 01:26:04 AM |
|
So, I am running late Batch 2 with 20130703. Will flashing to the latest version increase hash rate? Thanks!
The 20130818 firmware increased my speed by at least 1 GH/s consistently, so I'd say choose that one or later to flash.
|
|
|
|
Bogart
Legendary
Offline
Activity: 966
Merit: 1000
|
|
August 21, 2013, 01:43:17 AM |
|
Actually turns out the 850watt didn't burn out. After 2 weeks, I think it overheated and shutoff when my ac was turned off. ... Still though, I ordered a corsair hx 1050 gold to serve as primary psu.
Well that is good to hear! Just FYI the Corsair HX 1050 is about 1cm too long to fit inside. That's why I was hoping the V1000 would fit, but I am not sure if all the connectors will clear. It may work with using 2 out of the 3 PCI-E cables, not sure. Any problem with leaving just PCI-E 3 disconnected? Not sure how the power board distributes power. I put a HX1050 in my Batch 1 unit. I removed the PDU to install the power supply. Then I reinstalled the PDU after all the cables were attached. I removed the side panel which made things easier. Do not leave any inputs disconnected from the PDU. You may well melt the connected wires as they are unable to carry the amperage the unit is drawing, and overheat. Mine takes the 24-pin EATX connector, the 8-pin CPU Power connector, and 2 8-pin PCIe connectors. Later batches may be different, I don't know.
|
"All safe deposit boxes in banks or financial institutions have been sealed... and may only be opened in the presence of an agent of the I.R.S." - President F.D. Roosevelt, 1933
|
|
|
Ghost of USD
Newbie
Offline
Activity: 31
Merit: 0
|
|
August 21, 2013, 03:48:46 AM |
|
I have a batch 3 unit and replaced the PSU with one slightly too long to fit.
I'm running it now with the rear panel removed, and it actually runs about 5 degrees cooler than before.
|
|
|
|
DeathAndTaxes
Donator
Legendary
Offline
Activity: 1218
Merit: 1079
Gerald Davis
|
|
August 21, 2013, 03:59:17 AM |
|
Personally I am about to order one of these, as I am in your same boat (950W at the wall on a 910W PC Power & Cooling - hoping it doesn't fry with calculated 840W actual load, but 12V rail is still 12.2 so I think I'm good...) on my 4-module: PSU are rated on DC load. So a 910W PC P&C means it has a peak load of 910W DC. If 90% efficient at 100% load that would mean 910W DC = 1,011W AC (910/0.9). 950W AC = 855W DC (950W*0.9) I assumed 90% but you should look up efficiency curve for your PSU. If you cant find one and it is an "80Plus" rated unit you can check the 80Plus website. They have test which show efficiency at 100% load.
|
|
|
|
bitdude
|
|
August 21, 2013, 10:16:48 AM |
|
So, I am running late Batch 2 with 20130703. Will flashing to the latest version increase hash rate? Thanks!
The 20130818 firmware increased my speed by at least 1 GH/s consistently, so I'd say choose that one or later to flash. Thanks. I went for 20130820 and I can confirm your measurements! It is +1GH on 3 module Batch 2 machine. 24 hours uptime with avg. 82.876 GH and it was 81.8 avg. before the update.
|
|
|
|
|
tarui
|
|
August 21, 2013, 01:03:59 PM |
|
anyone tried removing the rear panel for batch 1 and lowers temp?
|
|
|
|
lastbit
|
|
August 21, 2013, 03:49:19 PM |
|
anyone tried removing the rear panel for batch 1 and lowers temp?
I have removed for batch#3 and lowered temperature by 2C. The presure felt when keeping panel with the hand was pretty substantial. I'm sure that if the slots were larger and beveled on the inside too it would have been better. You might give it a try, it's not even necessary to power down the machine if you're careful and don't mind back panel to lye captive by the power cord.
|
|
|
|
zoro
|
|
August 21, 2013, 05:11:00 PM |
|
i also removed it and i see good results! frequency is up since i have --avalon-auto
|
|
|
|
|
ProfMac
Legendary
Offline
Activity: 1246
Merit: 1001
|
|
August 21, 2013, 05:16:20 PM |
|
anyone tried removing the rear panel for batch 1 and lowers temp?
I have removed for batch#3 and lowered temperature by 2C. The presure felt when keeping panel with the hand was pretty substantial. I'm sure that if the slots were larger and beveled on the inside too it would have been better. You might give it a try, it's not even necessary to power down the machine if you're careful and don't mind back panel to lye captive by the power cord. I cringe at the thought of one of those metal nuts bouncing onto the pdu board.
|
I try to be respectful and informed.
|
|
|
silverston
|
|
August 21, 2013, 06:06:00 PM |
|
thanks, will try
|
|
|
|
PaJo2000
Newbie
Offline
Activity: 13
Merit: 0
|
|
August 21, 2013, 10:13:33 PM |
|
20130821 works fine, will see in next 12h
|
|
|
|
Bogart
Legendary
Offline
Activity: 966
Merit: 1000
|
|
August 22, 2013, 12:48:59 AM |
|
anyone tried removing the rear panel for batch 1 and lowers temp?
I have removed for batch#3 and lowered temperature by 2C. The presure felt when keeping panel with the hand was pretty substantial. I'm sure that if the slots were larger and beveled on the inside too it would have been better. You might give it a try, it's not even necessary to power down the machine if you're careful and don't mind back panel to lye captive by the power cord. I cringe at the thought of one of those metal nuts bouncing onto the pdu board. You can't grab 'em with a magnet either. I guess they're zinc or something.
|
"All safe deposit boxes in banks or financial institutions have been sealed... and may only be opened in the presence of an agent of the I.R.S." - President F.D. Roosevelt, 1933
|
|
|
johnny5
Member
Offline
Activity: 90
Merit: 10
|
|
August 22, 2013, 04:31:23 AM Last edit: August 22, 2013, 06:15:05 PM by johnny5 |
|
I'm running the 20130821 firmware on a 4-module batch 3 box, and my yellow LED (next to the green one) now flashes, on for a half-a-second, and off for a half-a-second over and over - still mining OK - but this yellow flashing is new - anyone else's also blinking with the 130821 firmware?
|
|
|
|
E
|
|
August 22, 2013, 05:05:55 AM |
|
Human readable changelog:
- HW error % will show in the cgminer API log tab on the avalon web interface. <boggle>
|
|
|
|
|