Quote
Regardless, 3 out of 4 of my batch 3 units end up with <10Gh/s after 3-12hrs of mining, some more frequently than others.
Check the System Log. Do you see any errors with the error code -71? 
Topic: Avalon ASIC users thread - page 104. (Read 438596 times)
Quote
any idea how to fix the NMW
Please search before asking; it saves everyone time.NMW is just another kind of hardware error. Best to just ignore it.
Sorry if this has already been discussed:
Should I be connecting the second EPS12V CPU connection on the PDU board? It seems to be working just fine with one, but I'm not interested in toasting anything!
And if I should connect it, can I just use an 8-pin PCI-e connection? I've got only one "CPU" connection on my PSU.
Thanks all
Should I be connecting the second EPS12V CPU connection on the PDU board? It seems to be working just fine with one, but I'm not interested in toasting anything!
And if I should connect it, can I just use an 8-pin PCI-e connection? I've got only one "CPU" connection on my PSU.
Thanks all
Touch the wires. Slightly warmer than ambient is OK. Anything more is dangerous.
You absolutely CANNOT swap EPS12V for PCIE. Be careful because depending on the manufacturing tolerances, sometimes those molex plugs can fit the wrong socket. But I recommend you don't even try, it will short out the PSU.
That said, if someone knows what they are doing, they could theoretically change around the positives and negatives to make it work.
any idea how to fix the NMW
[STATS] => 0
[ID] => AVA0
[Elapsed] => 1439
[Calls] => 0
[Wait] => 0.000000
[Max] => 0.000000
[Min] => 99999999.000000
[baud] => 115200
[miner_count] => 24
[asic_count] => 10
[timeout] => 35
[frequency] => 354
[fan1] => 1560
[fan2] => 2400
[fan3] => 2400
[temp1] => 26
[temp2] => 50
[temp3] => 50
[temp_max] => 50
[no_matching_work] => 81
[match_work_count1] => 1139
[match_work_count2] => 1077
[match_work_count3] => 1146
[match_work_count4] => 1112
[match_work_count5] => 1129
[match_work_count6] => 1151
[match_work_count7] => 1144
[match_work_count8] => 1147
[match_work_count9] => 1083
[match_work_count10] => 1157
[match_work_count11] => 1147
[match_work_count12] => 1099
[match_work_count13] => 1163
[match_work_count14] => 1175
[match_work_count15] => 1141
[match_work_count16] => 1188
[match_work_count17] => 1161
[match_work_count18] => 1110
[match_work_count19] => 1130
[match_work_count20] => 1157
[match_work_count21] => 1175
[match_work_count22] => 1179
[match_work_count23] => 1226
[match_work_count24] => 1165
[STATS] => 0
[ID] => AVA0
[Elapsed] => 1439
[Calls] => 0
[Wait] => 0.000000
[Max] => 0.000000
[Min] => 99999999.000000
[baud] => 115200
[miner_count] => 24
[asic_count] => 10
[timeout] => 35
[frequency] => 354
[fan1] => 1560
[fan2] => 2400
[fan3] => 2400
[temp1] => 26
[temp2] => 50
[temp3] => 50
[temp_max] => 50
[no_matching_work] => 81
[match_work_count1] => 1139
[match_work_count2] => 1077
[match_work_count3] => 1146
[match_work_count4] => 1112
[match_work_count5] => 1129
[match_work_count6] => 1151
[match_work_count7] => 1144
[match_work_count8] => 1147
[match_work_count9] => 1083
[match_work_count10] => 1157
[match_work_count11] => 1147
[match_work_count12] => 1099
[match_work_count13] => 1163
[match_work_count14] => 1175
[match_work_count15] => 1141
[match_work_count16] => 1188
[match_work_count17] => 1161
[match_work_count18] => 1110
[match_work_count19] => 1130
[match_work_count20] => 1157
[match_work_count21] => 1175
[match_work_count22] => 1179
[match_work_count23] => 1226
[match_work_count24] => 1165
I've been running my 4 module version Avalons at 350mhz using the stock PSU. Is this bad?
I have had 7 Avalons at 350mhz running on the stock PSU for about a week. I did have to replace one PSU, but otherwise they seem to be running fine.
In my previous experience when a PSU goes bad, the whole machine just shuts off, after replacing the power supply everything is good to go and no extra harm was done.
Am I taking a large gamble of the PSUs failing and potentially harming the equipment?
I read in the Batch 3 thread that a four module Avalon @ 350mhz runs at 930 Watts at the wall. I think the Avalons only ship with 850w PSUs. So obviously what I'm doing is probably risky, I'm just wondering how risky??
I have had 7 Avalons at 350mhz running on the stock PSU for about a week. I did have to replace one PSU, but otherwise they seem to be running fine.
In my previous experience when a PSU goes bad, the whole machine just shuts off, after replacing the power supply everything is good to go and no extra harm was done.
Am I taking a large gamble of the PSUs failing and potentially harming the equipment?
I read in the Batch 3 thread that a four module Avalon @ 350mhz runs at 930 Watts at the wall. I think the Avalons only ship with 850w PSUs. So obviously what I'm doing is probably risky, I'm just wondering how risky??
I would say you are very lucky. We have 2 4 module avalons with stock PSUs, and both are very unstable at anything over 300 MHz. Replaced one so far with a 1200w corsair PSU, and it runs solid at 350MHz static overclock. In fact it runs better with less error on static 350 than with 350 as a start point and --avalon-auto option.
Not sure why yours are OK, but consider yourself lucky!
was there any special way that you overclocked? I just got my 4 module running w a 1250 W OCZ PSU yesterday and I was happy that she was up and running. Haven't had time to mess around with OC.
One of my batch 3 avalons is performing very poorly due to a high HW errors. Unplugging all by one module, each module individually has between a 25 and 75% HW error rate at either 256 or 300Mhz.
This degraded performance starting after about 6 hours of mining at 300Mhz.
Is there anything I should try before trying to make a warranty claim?
I was calculating the hardware error rate wrong. It should be HW/LocalWork, which brings the error rates i'm seeing to a much more reasonable 1%.This degraded performance starting after about 6 hours of mining at 300Mhz.
Is there anything I should try before trying to make a warranty claim?
Regardless, 3 out of 4 of my batch 3 units end up with <10Gh/s after 3-12hrs of mining, some more frequently than others.
Could this behavior be due to the new temperature throttling feature? I have the default temperature limits of 70C target and 90C cutoff, but I don't see the temps going over 70C.
Sorry if this has already been discussed:
Should I be connecting the second EPS12V CPU connection on the PDU board? It seems to be working just fine with one, but I'm not interested in toasting anything!
And if I should connect it, can I just use an 8-pin PCI-e connection? I've got only one "CPU" connection on my PSU.
Should I be connecting the second EPS12V CPU connection on the PDU board? It seems to be working just fine with one, but I'm not interested in toasting anything!
And if I should connect it, can I just use an 8-pin PCI-e connection? I've got only one "CPU" connection on my PSU.
I connected *all* the power connectors on my b3's and they all work fine:
1 x 24-pin mobo connector
2 x EPS 12v
3 x PCI-E 8-pin
I looked around and didn't see any directions on the forums about which to connect, and Avalon shipped with no documentation, so I figured it's a power-hungry thing and why not just connect them all.
You cannot substitute the pci-e for the eps connector.
I feel like I'm doing something wrong.
I hooked up 3x 3-module units today and my temps seem a bit high. On all 3 of them my Temp2 sits at 60C with the fans running at 3500 RPM. Is this normal? Also what is a normal accepted/HW error ratio?
I hooked up 3x 3-module units today and my temps seem a bit high. On all 3 of them my Temp2 sits at 60C with the fans running at 3500 RPM. Is this normal? Also what is a normal accepted/HW error ratio?
They changed the temp sensor for B3, it's closer to the chips, should have a hotter readout.
Good to know
I was worried there wasn't going to be any room for overclocking.
Can someone explain to me what NMW is cgminer for Avalon please as i see increasing numbers in this and is it anything to be worried about ?
Sorry if this has already been discussed:
Should I be connecting the second EPS12V CPU connection on the PDU board? It seems to be working just fine with one, but I'm not interested in toasting anything!
And if I should connect it, can I just use an 8-pin PCI-e connection? I've got only one "CPU" connection on my PSU.
Thanks all
Should I be connecting the second EPS12V CPU connection on the PDU board? It seems to be working just fine with one, but I'm not interested in toasting anything!
And if I should connect it, can I just use an 8-pin PCI-e connection? I've got only one "CPU" connection on my PSU.
Thanks all
Thanks for Sharing this is good information to know. The good news is that it was only your avalon that was affected and nothing else. Making me think twice about electrical fires that this thing can cause 
I feel like I'm doing something wrong.
I hooked up 3x 3-module units today and my temps seem a bit high. On all 3 of them my Temp2 sits at 60C with the fans running at 3500 RPM. Is this normal? Also what is a normal accepted/HW error ratio?
I hooked up 3x 3-module units today and my temps seem a bit high. On all 3 of them my Temp2 sits at 60C with the fans running at 3500 RPM. Is this normal? Also what is a normal accepted/HW error ratio?
They changed the temp sensor for B3, it's closer to the chips, should have a hotter readout.
Only 2 of those 24 wires carry 12VDC.
See the pinout:
http://image.pinout.net/pinout_network_rj45_files/24-pin-atx-pinout.gif
See the pinout:
http://image.pinout.net/pinout_network_rj45_files/24-pin-atx-pinout.gif
If you're looking to buy/host some Avalon batch 3 modules or if you want to sell some batch 3 clone parts (specifically a PDU and some cables), please see this thread: https://bitcointalksearch.org/topic/m.2886697
So in retrospective, for those of us looking on what NOT to do and prevent accidents, why did this happen? That power supply should have handled current perfectly. Did you do custom cabling or what? I thought Avalon provided with proper cables. I still don't understand why this happened....
He ONLY installed the 24pin mainboard connector.
Sorry for my ignorance but what else was he supposed to connect? power only comes through one cable AFAIK.
I thought the same thing. And you if you try that, you'll end up with a melted miner. You have to hook up the 24pin ATX cable AND all of the power supply cables, like you would if you were supplying power to devices in your computer like a hard drive or a DVD player. Why it runs at all with only the single cable plugged in? I have no idea.
I've put this thread up as well to see how much interest is out there for buying Batch 3 modules: https://bitcointalksearch.org/topic/wts-avalon-batch-3-modules-sold-269845 I've gotten some serious offers, but most are lower than I was hoping.
Oh, I see. Seems kind of stupid that Corsair, OCZ and all those suppliers *assume* that the only application for their PSU is a PC.
I'd appreciate if you could point me to some thread that tells me how to "trick" the rest of the cables. I have a Corsair 750W and I read the manual but never did it say anything about connecting all of them.
However, I was planning to ONLY use the Molex connectors for my mining application anyway (I don't require the use of the 24-pin connector). Not sure if I'd be in trouble as well or whether the burning issue only happens when connecting the 24-connector cable alone...thoughts?
The problem had nothing to do with a single faulty wire or bad PSU.
The Avalon is not designed like a PC. All of its modules and controllers are all powered by that single PDU board.
Because the power supply has a single 12V rail, if you only have the ATX connector plugged in, the Avalon tries to pull all of the ~800W (66A) it requires from 2 tiny (18 gauge) leads.
That's ~33 amps per wire, but the maximum current they can safely handle is 6 amps each. The result is guaranteed melted wires, or worse as was the case here.
When all the plugs are attached to the PDU, there are 26x 12V leads sharing the load, a much more reasonable 2.5 amps each.
Doesn't the ATX cable itself consist of 24 cables not 2? It'd seem to me that the ATX connector would be enough to distribute all the load.
I feel like I'm doing something wrong.
I hooked up 3x 3-module units today and my temps seem a bit high. On all 3 of them my Temp2 sits at 60C with the fans running at 3500 RPM. Is this normal? Also what is a normal accepted/HW error ratio?
I hooked up 3x 3-module units today and my temps seem a bit high. On all 3 of them my Temp2 sits at 60C with the fans running at 3500 RPM. Is this normal? Also what is a normal accepted/HW error ratio?
They changed the temp sensor for B3, it's closer to the chips, should have a hotter readout.
I feel like I'm doing something wrong.
I hooked up 3x 3-module units today and my temps seem a bit high. On all 3 of them my Temp2 sits at 60C with the fans running at 3500 RPM. Is this normal? Also what is a normal accepted/HW error ratio?
I hooked up 3x 3-module units today and my temps seem a bit high. On all 3 of them my Temp2 sits at 60C with the fans running at 3500 RPM. Is this normal? Also what is a normal accepted/HW error ratio?
If you're looking to buy/host some Avalon batch 3 modules or if you want to sell some batch 3 clone parts (specifically a PDU and some cables), please see this thread: https://bitcointalksearch.org/topic/m.2886697
So in retrospective, for those of us looking on what NOT to do and prevent accidents, why did this happen? That power supply should have handled current perfectly. Did you do custom cabling or what? I thought Avalon provided with proper cables. I still don't understand why this happened....
He ONLY installed the 24pin mainboard connector.
Sorry for my ignorance but what else was he supposed to connect? power only comes through one cable AFAIK.
I thought the same thing. And you if you try that, you'll end up with a melted miner. You have to hook up the 24pin ATX cable AND all of the power supply cables, like you would if you were supplying power to devices in your computer like a hard drive or a DVD player. Why it runs at all with only the single cable plugged in? I have no idea.
I've put this thread up as well to see how much interest is out there for buying Batch 3 modules: https://bitcointalksearch.org/topic/wts-avalon-batch-3-modules-sold-269845 I've gotten some serious offers, but most are lower than I was hoping.
Oh, I see. Seems kind of stupid that Corsair, OCZ and all those suppliers *assume* that the only application for their PSU is a PC.
I'd appreciate if you could point me to some thread that tells me how to "trick" the rest of the cables. I have a Corsair 750W and I read the manual but never did it say anything about connecting all of them.
However, I was planning to ONLY use the Molex connectors for my mining application anyway (I don't require the use of the 24-pin connector). Not sure if I'd be in trouble as well or whether the burning issue only happens when connecting the 24-connector cable alone...thoughts?
The problem had nothing to do with a single faulty wire or bad PSU.
The Avalon is not designed like a PC. All of its modules and controllers are all powered by that single PDU board.
Because the power supply has a single 12V rail, if you only have the ATX connector plugged in, the Avalon tries to pull all of the ~800W (66A) it requires from 2 tiny (18 gauge) leads.
That's ~33 amps per wire, but the maximum current they can safely handle is 6 amps each. The result is guaranteed melted wires, or worse as was the case here.
When all the plugs are attached to the PDU, there are 26x 12V leads sharing the load, a much more reasonable 2.5 amps each.
It's only a power distribution board with melted connectors.
If somebody can post a good pic that shows all the traces, you should be able to test every single trace with a continuity tester.
Mark which one does not test good.
Then take it to a component level repair place and have them fix the traces. It won't be cheap, and it will take time, but it can be done. Resolder connectors on there, replace all capacitors, and you should be good to go.
If there is an actual chip on there that needs to be replaced, you might be screwed totally.
If somebody can post a good pic that shows all the traces, you should be able to test every single trace with a continuity tester.
Mark which one does not test good.
Then take it to a component level repair place and have them fix the traces. It won't be cheap, and it will take time, but it can be done. Resolder connectors on there, replace all capacitors, and you should be good to go.
If there is an actual chip on there that needs to be replaced, you might be screwed totally.
that;s assuming i have the pinouts
Oh, I see. Seems kind of stupid that Corsair, OCZ and all those suppliers *assume* that the only application for their PSU is a PC.
I'd appreciate if you could point me to some thread that tells me how to "trick" the rest of the cables. I have a Corsair 750W and I read the manual but never did it say anything about connecting all of them.
However, I was planning to ONLY use the Molex connectors for my mining application anyway (I don't require the use of the 24-pin connector). Not sure if I'd be in trouble as well or whether the burning issue only happens when connecting the 24-connector cable alone...thoughts?
I'd appreciate if you could point me to some thread that tells me how to "trick" the rest of the cables. I have a Corsair 750W and I read the manual but never did it say anything about connecting all of them.
However, I was planning to ONLY use the Molex connectors for my mining application anyway (I don't require the use of the 24-pin connector). Not sure if I'd be in trouble as well or whether the burning issue only happens when connecting the 24-connector cable alone...thoughts?
The PSU worked fine, it was the connecting wire that had a problem, it appears to have caught fire and spewed molten plastic all over the board.
Anyway, all you have to do is connect the 24 pin ATX connector along with the GPU connecting wires.
Not sure what are the other "connecting wires" you are referring . My mining board will only have molex for a fan but that's about it.
Oh, I see. Seems kind of stupid that Corsair, OCZ and all those suppliers *assume* that the only application for their PSU is a PC.
I'd appreciate if you could point me to some thread that tells me how to "trick" the rest of the cables. I have a Corsair 750W and I read the manual but never did it say anything about connecting all of them.
However, I was planning to ONLY use the Molex connectors for my mining application anyway (I don't require the use of the 24-pin connector). Not sure if I'd be in trouble as well or whether the burning issue only happens when connecting the 24-connector cable alone...thoughts?
I'd appreciate if you could point me to some thread that tells me how to "trick" the rest of the cables. I have a Corsair 750W and I read the manual but never did it say anything about connecting all of them.
However, I was planning to ONLY use the Molex connectors for my mining application anyway (I don't require the use of the 24-pin connector). Not sure if I'd be in trouble as well or whether the burning issue only happens when connecting the 24-connector cable alone...thoughts?
The PSU worked fine, it was the connecting wire that had a problem, it appears to have caught fire and spewed molten plastic all over the board.
Anyway, all you have to do is connect the 24 pin ATX connector along with the GPU connecting wires.
Jump to: