Topic: Avalon ASIC users thread - page 106. (Read 438722 times)

I think that is the controller board.  It has an FPGA on it.  To get a replacement, you go do a lot of basic simple correspondence.  To start, ask Avalon and burnin if they will ship you a replacement.  Ask what damage trade-in price they will give you.  While you are in the midst of a misfortune, your device is nonetheless valuable to the community, so continue to treat it with care.  It documents a failure mode, and the process of examining it should give insight for design improvements, and practice for the soon to be important repair business.  

The rest of the community also wants to know how to get replacements, and who will be helpful when disaster strikes.  It will strike.  Yesterday we had 2" hail and violent weather.  I could easily have taken a lightening strike and had the same mess.  Other people will have different issues, perhaps a mouse seeking a warm place to sleep, a dog marking territory, or a wife disappointed with a relationship with a secretary.

I would be very conservative.  If the module is damaged, it could in turn damage a replacement part.  I would run the unit with 2 modules to be sure that the repair worked.  I haven't thought through when & how to bring the suspicious module back online.

Again, take photos of each item that you remove.  Perhaps on white butcher paper with a carefully lettered 3x5 card with date and description.
Especially, take some close up photos of the controller board, front and back.

You might create a new payment address and put it in your posting signature, perhaps one or two Avalon owners can send a bitcoin or two as a hand-up.

i did the exact same thing.

but lucky for me, i detected a burning smell and heard a ticking noise.

i eventually found the atx cable's wires had melted and caused arching, probably causing the noise.

only damage was the atx cables and the atx port (wire insulator) melted into the pins so i had to replace it.

decided to scavage it from my old spoilt atx motherboard.

the bigger problem is new atx power cable. down here, ocz psu arent very common so i couldnt find the replacement cable and i didnt want to risk it and i was forced to buy a new psu today.

It would be nice if some of the avalon clone manufacturers would offer such a service with any spare parts they might have.

Thank you for being helpful. I'm surprised the Avalon powered on at all without all the proper connections in place.

My comfort level is not far above consumer. Sure, I've built my own computers, but it's been many years. I've replaced laptop components and such, but I don't even know what the components inside the Avalon are. I don't have any instruments.

Damaged components:

The main board the ATX cable plugs into appears to be damaged. What is that called? Can I get a replacement? If so, how?

Is the module next to it damaged? I don't know. Can I find out other than repairing everything else and turning it on?

USB cable is fried.

ATX cable is fried.

Any help with getting replacement parts?

As for RMA, I was referring to the power supply, which is what their employee told me to do.

If the controller board is dead, it is reasonable to request the favor of a replacement at incremental manufacturing cost.  It is equally reasonable to be denied.

Only slightly off topic, an obvious business opportunity is to collect money and buy up inventory to provide replacements.  But as harsh as the bitcoin community is, who would want that sort of exposure?

holy shit! You totally roasted your avalon.

Are you sure about RMA? It was your fault, theres no right for replacement

Thanks for posting.
Please keep a log book and many photographs.

First, find out what is damaged. 
Next, replace the damaged part.
Next, hash as soon as possible.

.....

What is your comfort level with electronics, on a scale from pure consumer to experienced with good instruments?

I'm screwed. I incorrectly thought the ATX main cable was the only one needed (since, unlike a normal PC where the modular devices need to be powered individually and don't even work otherwise, the Avalon modules all connect directly to the board). I've had power supplies in the past that supported both ATX and AT and I thought this was similar.

I plugged it in yesterday, had a little trouble with the USB not being connected, but after a few times taking it apart and putting it back again, I got it working. It ran great all last night and this morning. When I got back after a meeting, this is what I found:







I emailed Avalon, and they asked me to post here. Talking with the vendor of my power supply (OCZ ZX Series 1250W), he said 600W is too much to go through one wire, but it wouldn't trigger any concerns on their power supply as it doesn't measure power output per wire. The wire itself just melted.

What do I do now? I'm starting the RMA process for the power supply but as for the miner... I don't know where to begin.

I feel discouraged. I feel like an idiot.

I have some network issue it's all god now. Lan was not pinned in menu wifi was...

Also had this before, I reset the avalon and it's working smooth now. If you're getting the same error, I don't think you reset it properly. Wait 5-6 seconds after turning the avalon on before hitting the reset button

Go to your cgminer config tab and clear the value behind more options and set the chip frequency to 256 and try again.

I have seen this before with my silver Avalon. I am running 20130703 firmware.

In the Cgminer Config tab, what does it say in the "API Allow" text box?

I have try flashing firmware, manual reset on button....
different broswers and still getting the same error?

When I import my worker and pool and click save and apply the process stops here

Applying changes
/etc/config/cgiminer

[Firmware Version] => 20130723
cgminer-37815e0
luci-56acbaa
cgminer-openwrt-packages-d9bac89
Socket connect failed: Connection refused[Firmware Version] => 20130723
cgminer-37815e0
luci-56acbaa
cgminer-openwrt-packages-d9bac89
Socket connect failed: Connection refused


? this is my error
when i reset machine to default values

Swap PSU?

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 would just go and grab an ATX PSU and plug and play tbh

I ask only because the Server supplies only output 12v.  I would have to make a custom cable.   But if something inside the Avalon unit needs 5v or 3.3v I will just buy a standard ATX supply.

Could it be due to the power board or controller making that noise?

If it's not the PSU then it has to be one of the fans.  Maybe the fan blade is hitting one of the plastic supports, and you can't hear it under load because the fan might shake less when it has more angular momentum.

Since you don't hear it under load, one way to deal with it for now would just be to run the fans 100%,  you can do that with in "--avalon-fan 100" or by lowering the max temp to something that'll keep the fans running 24/7 in the "More Options" box on the CGMiner config page

Obviously that'll mean a lot more sound anyway, but it might not be as annoying.

For a longer term fix you could try buying a spare fan and try swapping fans one by one until you find the problem one.

You could also examine the fans closely with the power off and see if any of them are 'wobbly'