More updates on my bad card:
- No matter what I do it will shut off within 5 minutes of beginning mining, although it does not throw a lot of errors while it is mining. - The shutdown is a slow taper in noncerate to about 10 GH/s, after which is hits 0 GH/s. After this the board does not restart. - Heatsinking the board like crazy has no effect and thermal diode readings all over the voltage regulators max out at 45C when you start it up, so it's not any kind of heat issue - Manually setting the clock speed has no effect - Board position has no effect
At this point I'm just going to say it's a bad board and needs to be RMA'd unfortunately. Dave, will you accept RMAs with heatsinks on them?
Can you please post a picture, front and back side? I had a card performing similarly, I carefully monitored the power consumed by the card. It was ~70 watts. Had to reduce the voltage a bit to make it reliable. Sure, just a minute. I checked the voltage (Inductor +, GRND -), 0.862 V for this card versus and average of 0.860 V for the other cards, so the voltage is apparently not the issue.
|
|
|
More updates on my bad card:
- No matter what I do it will shut off within 5 minutes of beginning mining, although it does not throw a lot of errors while it is mining. - The shutdown is a slow taper in noncerate to about 10 GH/s, after which is hits 0 GH/s. After this the board does not restart. - Heatsinking the board like crazy has no effect and thermal diode readings all over the voltage regulators max out at 45C when you start it up, so it's not any kind of heat issue - Manually setting the clock speed has no effect - Board position has no effect
At this point I'm just going to say it's a bad board and needs to be RMA'd unfortunately. Dave, will you accept RMAs with heatsinks on them?
|
|
|
Yeah so that one board is still bad. Keeps shutting itself off even if I downclock it so the onboard temps are only 30-40C. I guess I'll have to talk to dave about RMAing it. This happens no matter what slot it's in.
|
|
|
bad off per chip good cores 0: 848 35.892 37.522 2507 76 4 0 0 16 00 (2.243/chip) 34% 1: 848 36.307 37.491 2536 15 0 0 0 16 00 (2.269/chip) 35% 2: 848 27.431 27.946 1916 134 16 5 2 16 00 (1.714/chip) 30% 4: 848 35.720 36.708 2495 68 2 0 0 16 00 (2.232/chip) 34% 5: 848 37.180 37.152 2597 14 1 0 0 16 00 (2.324/chip) 35% 6: 848 38.898 38.917 2717 8 2 0 0 16 00 (2.431/chip) 36% 8: 848 33.000 34.309 2305 82 4 0 0 16 00 (2.062/chip) 32% 9: 848 33.529 36.698 2342 82 1 0 0 16 00 (2.096/chip) 32% A: 848 0.000 0.000 0 0 0 0 0 16 00 (0.000/chip) 9% speed down C: 848 35.648 36.793 2490 16 1 0 0 16 00 (2.228/chip) 34%
So, yeah, I have 1 bad H-board it appears. It throttles down for about 15 minutes and then just goes to 0 GH/s Uh, are these still covered under warranty if I heatsinked the back, Dave? Edit: This is interesting, a second board just did the same thing. :| Your Hboard maybe overheat that why Hboard shutdown and you need better cooling. Thermal diode reports chips are 51 C VRM is 54 C :| I swapped the cards into different slots and now they're both no longer throwing lots of errors or shutting down Clearly H-boards are delicate flowers Most recent stats: board-2 speed nrate hrate good errors spi-err miso-er duplic good bad off per chip good cores 0: 848 36.851 37.195 2574 75 0 1 0 16 00 (2.303/chip) 19% 1: 848 35.419 37.290 2474 9 1 0 0 16 00 (2.214/chip) 18% 2: 848 37.452 38.685 2616 18 1 0 0 16 00 (2.341/chip) 19% 4: 848 35.262 36.434 2463 60 2 0 0 16 00 (2.204/chip) 18% 5: 848 36.035 36.846 2517 18 1 0 0 16 00 (2.252/chip) 19% 6: 848 37.982 38.600 2653 8 0 0 0 16 00 (2.374/chip) 20% 8: 848 31.668 34.045 2212 77 1 0 0 16 00 (1.979/chip) 17% 9: 848 34.847 36.550 2434 69 0 0 0 16 00 (2.178/chip) 18% C: 848 35.777 36.518 2499 30 2 0 0 16 00 (2.236/chip) 19% D: 848 32.885 31.941 2297 13 1 2 0 16 00 (2.055/chip) 17%
|
|
|
Okay, this is interesting. There are two problem boards, and they keep shutting off as if they're thermocycling or something (go up to 36 GH/s, slowly go to zero, then suddenly go back to full speed). The chips on these boards are fine by temperatures, all less than 60 C.
|
|
|
bad off per chip good cores 0: 848 35.892 37.522 2507 76 4 0 0 16 00 (2.243/chip) 34% 1: 848 36.307 37.491 2536 15 0 0 0 16 00 (2.269/chip) 35% 2: 848 27.431 27.946 1916 134 16 5 2 16 00 (1.714/chip) 30% 4: 848 35.720 36.708 2495 68 2 0 0 16 00 (2.232/chip) 34% 5: 848 37.180 37.152 2597 14 1 0 0 16 00 (2.324/chip) 35% 6: 848 38.898 38.917 2717 8 2 0 0 16 00 (2.431/chip) 36% 8: 848 33.000 34.309 2305 82 4 0 0 16 00 (2.062/chip) 32% 9: 848 33.529 36.698 2342 82 1 0 0 16 00 (2.096/chip) 32% A: 848 0.000 0.000 0 0 0 0 0 16 00 (0.000/chip) 9% speed down C: 848 35.648 36.793 2490 16 1 0 0 16 00 (2.228/chip) 34%
So, yeah, I have 1 bad H-board it appears. It throttles down for about 15 minutes and then just goes to 0 GH/s Uh, are these still covered under warranty if I heatsinked the back, Dave? Edit: This is interesting, a second board just did the same thing. :|
|
|
|
The slots are no longer labeled "A1 A2" etc, rather now they are just 0 1 2 3 .. E F
My bad performing card is in slot 7 and I haven't tried to the other slots out yet
edit: Slot 7 card just died, I will move it to a new slot and see if it hashes any better there
2nd edit: Okay, there ARE banks, they just arent labeled on the PCB. Messing around with trying them in different banks now
Did you use fan on Hboards? 3x CM R4 @ 2k RPM, 120mm Hboards are also heatsinked
|
|
|
The slots are no longer labeled "A1 A2" etc, rather now they are just 0 1 2 3 .. E F
My bad performing card is in slot 7 and I haven't tried to the other slots out yet
edit: Slot 7 card just died, I will move it to a new slot and see if it hashes any better there
2nd edit: Okay, there ARE banks, they just arent labeled on the PCB. Messing around with trying them in different banks now
|
|
|
Still haven't heard back about the refunds :/
|
|
|
10x H boards heatsinked and running; getting 36 GH/s per board
Currently pulling about 420 w on a gold PSU
Pictures and thermals coming soon
Taco can you post log from admin page ? I'm curious how many chip fully work on new version H-board. And pictures boards with your heasink 100% of chips are returning some sort of valid nonce, impressively, and the startup dump says 100% good cores One board is a little bunk and throws a massive amount of errors in comparison to the others and the chips are tuning down to about 25 GH/s after a little while. Rest of the boards seem OK. edit: Restarted and it seems OK
|
|
|
10x H boards heatsinked and running; getting 36 GH/s per board
Currently pulling about 420 w on a gold PSU
Pictures and thermals coming soon
|
|
|
Fedex just got to me, I'm going to pick it up later tonight and I'll try and have some photos/thermal data up for you guys
|
|
|
Does anyone know how to downclock the chips on the new boards? I don't have enough cooling apparently and my heat sinks won't arrive for another week. The chips just keep cycling, the high side of 575GH for my whole rig, and then down to 100GH, slowly, as the Hboards turn the chips off one by one.
The DHCP script did not work for my network, I had to have a friend go in the Pi and help me set up a static IP manually.
Aside from that, the October rig is great. For anyone still waiting on theirs, make sure you have good cooling ready to go. I am looking at some 100cfm 120mm fans right now, but I might get ones a little stronger.
Static pressure is often more important than airflow... I'm going to heatsink my boards and then try 3x coolermaster r4 120mm 2000 rpm fans on my rig tomorrow and then check the chip and thermal regulator temps with a thermal diode and give you guys some data
|
|
|
Still haven't heard back. I'll check on my breaks at work today.
|
|
|
^^ Thanks.
Haven't heard back about the refunds. :/
I'll check tomorrow morning before I head into work.
Night guys.
|
|
|
![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2FDWpgeWT.png&t=663&c=R-5EnYWCNd2BPQ) perfect Expect 4-18 hours... its normal clearance Kay, thanks!
|
|
|
![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2FDWpgeWT.png&t=663&c=R-5EnYWCNd2BPQ) perfect
|
|
|
Cool, I'm gonna go drink a beer and not think about this toooo much until I hear back about the refund situation. Between a full time job, volunteering, and coming here, this has spanned an 11 hour day and I'm a little tired.
|
|
|
There was a typo in the Sierra size as suspected; e-mail as of a couple of minutes ago says: 17” x 16” x 7”
This has been corrected in the previous post
Thanks. I am assuming 7" H X 17" W X 16" D. Yeah, that's my guess too, but I can ask again if you and DnT wanna confirm
|
|
|
Yeah, those numbers don't make sense now that I look at the rack specs. 20.5” (482 mm) is the width of a normal server rack, which is given for the BJ. None of the other numbers conform to a normal #U size. I'll pass another e-mail over.
Well the dimension between rack posts is ~17.5" (technically a little wider but there isn't suppose to be a metal to metal fit between chassis wall and the rack post). Chassis widths (excluding ears or slides) are usually ~17.25" +/- 0.25". The 19" is the nominal dimensions and is larger than the space between posts. Still the numbers do appear to be complete nonsense. The width would need to be ~ 17" excluding rackmount ears to fit between the rack posts. The height would be in increments of 1.75" minus 1/8" for clearance 6.875" for 4U). In your clarification ask them to provide the H, W, D with each dimension. They just got back to me... Do you want to ask again to specify which dimension is which?
|
|
|
|