Topic: Cointerra Hardware Support **Unofficial - page 6. (Read 56761 times)

Thanks. I'll try that out. I don't think it's hitting 120 though because it seems to top out around 95 as best I can see. It could be the UI isn't keeping up though. I really don't like the cointerra UI. I have a couple pi's and a failing controller now so I may swap the failing beaglebone with a pi and see what happens.

With regard to the spacer I ground mine down to even but not below. I checked the pattern of the thermal paste after doing so and there was a significant difference in the contact pattern of the block against the chips. Before I had visible high spots in the paste but after grinding the spacer things were much more even. After running for a couple days I think this change yielded 10-15C improvement.

I am planning on either removing it with a razor or grinding it down further the next time I have them apart (probably this weekend). I will post results for everyone with pics.

So far with the changes I have made one board has dropped an average of 40C and the other 60C and that's on noctua thermal paste. I still have some work to do because I am only running at power level 6 and cannot run higher at this time without the boards exceeding 80C.

What is happening the cores hit 120C very fast and then either turn off or get the frequency turned down so it runs slower at a cooler temp.

If you want to see the fast error just connect the board to a PC with usb and run cgminer.  It will show which cores are getting hot.

Regarding the ring around the cores, what would happen if that was ground down well below the chip height?  Anyone tried this?

Has anyone tried immersion cooling these boards or has everyone just thrown them away?

Definitely not. I have clear lines so I would see any air. Plus I purged it all when I re-ran the lines on that board.

I'm going to take a straightedge and my feeler gauges to it as soon as the liquidpro comes in. I'm willing to bet the chips don't sit straight so sometimes I get lucky and they cool properly but most of the time they do not. That's probably why LiquidPro works so well for people. Until then I will deal with it because I'm getting tired of taking it apart every night. I will post my updated results as soon as my shipment arrives.

Air in the cooling system maybe?

Well I was all excited to get the now famously running terraminer off of the extension cords and into a more permanent location tonight but upon relocating the unit a couple chips decided they wanted to be hot again. I'm contemplating the possibility of faulty temp sensors. I was gentle when moving the unit and most of the cores are still reading great temps. It's not out of line but definitely hotter than I would like. I have the unit in a (near) freezing cold environment so it should hold until the LiquidPro arrives later this week (hopefully). Another interesting thing I have noticed is that the temps start off really high and then settle down significantly lower after running for a few minutes. Do the rest of you experience this behavior when mining starts?

Another interesting thing, the machine with the resurrected battery-shorted board was running 20C hotter last night than it is running today. I'm not complaining but I really wish I could understand the wild fluctuations in cooling performance (or supposed cooling performance if the sensors can be trusted)

More to follow after the LiquidPro treatment over the weekend.

Another update. I fired up the other board in the unit and I am now hashing at about 1.3TH total and my hottest chip is 66C. The fans are barely running and the errors are well within range. I'm calling this a win. If the unit runs stable overnight I will move it to a more permanent location and let it eat.

I think what happens is the coolant in the system degrades and little bits of solids hinder the pump performance. I initially drained the system into a cup and there flakes of solids in there. Small ones but solids just the same. We will see how much temps change after the ethylene glycol. I also expect cooling performance to improve after the thermal paste has time to burn in. I have already seen significant improvement in just this short time.

Alright everyone. I tried a couple things tonight and one of them yielded very encouraging results.

I tried this on the worst offending board with the overheating. At this point I believe that if you have a board with a chronic overheating problem it is an issue in the chips and you aren't going to get rid of it completely but the things I have listed above and what I list here will make it as good as it can be given what you have.

This board doesn't run stable at any power level over 6 regardless of how cool or hot it is. Whether this is because the prior owner abused it by running it too hot for too long or if it's just a manufacturing defect the world will never know. That being said I was able to run this board at around 80C consistently at power level 9 with these changes. It didn't make the board any more stable but it did keep the temps in check. I am running at level 6 now and 670GH from the board with no errors. It's sticking consistently around 70C on the hottest chip.

Here's the skinny. I removed the water coolers from the board and took the boards out. I got out my dremel tool and ground down the spacer so it was no longer holding the water block away from the chips. The metal ring around the perimeter of the chip is too thick and holds you (unevenly) away from the chips. One was particularly bad and uneven. You could see it in the heat paste. The other was pretty good but I ground it down anyway.

After doing this I tried it out. Temps still sucked. Overheating at anything over level 6. Back to the drawing board.

Now I decided to cut the water lines off and flush the system. I got some good petroleum rated 1/4 inch line and replaced them after I blew everything off. I refilled the system with distilled water (just for now to test it out. I will add some ethylene glycol later) and plugged int again. SUCCESS! I'm running nice and cool and getting good GH!

Next step will be to give the whole rig a treatment with liquidpro when it arrives. I will post results of that test as well.

Here are some pics of the process and completed rig.

http://i.imgur.com/VdznceH.jpg
http://i.imgur.com/vhU6ctY.jpg
http://i.imgur.com/sL0X0eR.jpg

Couple new developments in the quest to make a bad miner good.

I was working on the water blocks last night and I discovered something interesting with the inserts. See pictures below.

http://i.imgur.com/aGvNWaP.jpg

http://i.imgur.com/CZgSzVc.jpg

Note the difference in size and the distortion on the larger one. If your inserts look like the large one you will need to remove and discard them. If they look like the smaller one you can leave them in place and they will improve the performance of the water block vs having them out.

I managed to get the board that was rapidly overheating to run at about 75c last night by removing the inserts but considering the cold air it's pulling it should be operating around 50. Tonight I am going to try 2 new things.

1. replace the black hoses with transparent hoses and flush the pumps/radiators. Then I will refill and monitor flow and temps.

2. If the above does not yield satisfactory results I will attempt to find a replacement water block/pump combo and install it.

More to follow...

They are not by any means quiet or reliable. Buy an antminer.

I see one of these advertised on eBay. Wondering if its worth buying or are they all badly built. Power isn't an issue. Just need it to be quiet and reliable.

Alright. I got the pumps and inserts sorted on one of the boards. Temps went from fluctuating between 95 and 105C to sitting pretty at 77. Basically all I did was remove the inserts and top off the coolant. Seems to work well. Now I have another problem. One of the pumps just died. It was working great this morning, I unplugged the unit to switch with the one I had just repaired and when I plugged it in it died. These stinking things are going to be the death of me. I'm trying to find a water cooling kit that will work but the brackets are all wrong. Looks like I will have to purchase a block and pump separate. That should do the trick. Updates to follow...

Little more info on the LED lights and what they mean.

Front status LEDs: First is just machine status (CTR1). Basically this just lets you know the beaglebone is up and running and that it got an IP. If it's orange/red you either didn't get an IP or there is some other issue with the unit. If it's off the beaglebone is being a butthead and you need to restart your miner.

The second light (CTR2) is not used. Ignore.

Then there's miner 1 and 2 which show activity on the miners kind of like the HDD activity light on a PC.

If your CTR1 doesn't light up, open the case and look at the beaglebone controller. There should be a series of 4 blue lights on the left side (when looking at it from the back of the machine) and they should be flickering. That's how you know you are connected to the network. If not, hit the reset button on the beaglebone until it works. If it doesn't ever work then you probably need to reflash or replace the unit.

I have never managed to get the main reset button to work on the terraminer. Maybe it's just me but the reset instruction never seems to make it to the board. Not really an issue because the firmware seems to work OK.

Now for the status on the CTA boards:

There are a series of 9 LEDs on the boards numbered D37-D45. These are the chip status lights and board status light. D45 is indicative of a board fault and if everything is peachy it won't be lit. The others are status on the individual chips. If d37-44 are cascading on and off then everything is good. If some are being skipped then that core is offline. The most common cause for this is overheating on that core. You will need to check your temps in the UI.

Now for D45. If D45 is lit red it means a board fault. Most commonly this is caused by the board being unable to communicate with one of the chips. Most commonly this means you have a fried component on the board. D37-40 are for the back chip. D41-44 are for the front chip. I don't have any more detail than that at the moment. You will need to inspect the board (especially the mosfets) for a visibly damaged component. It may be very subtle and difficult to see. I suggest using a magnifying glass.

Status lights on the power supplies are AC OK and DC OK. AC is on the left, DC is on the right. If DC is orange you need to troubleshoot your fault per the method I mentioned in my last post. If AC is orange you probably have a power cord issue or a blown power supply.

If you have any additional questions or need more detail on any of this please let me know. I know a lot of people got screwed by cointerra. I want to help if possible.

Good morning all,

Long time reader but I garnered some benefit from reading this thread so I'm going to share my experiences and fixes with you. I have suffered nearly every failure/issue mentioned in this thread over the last 2 weeks and managed to repair about 90% of the issues so far.

The story beigins with me purchasing a used terraminer IV on ebay. I paid 250 plus shipping and was promised the unit was in good working order. It was not. I requested a replacement/refund and he sent me another even more broken terraminer. Now I have 2 that I am in the midst of bringing back to life.

I plugged the first miner in and set it up. First thing I notice is the temps are outrageously high on ALL chips. Even when just idling they would overheat and "shut down". Upon reading here I learn that the heat transfer paste is an issue so I pull the water blocks off to reapply. Turns out the previous owner had already applied LiquidPro but it looked like a pretty trashy application as there were gobs of the original thermal paste left all over the place. I thoroughly cleaned the chips and applied a quality thermal paste that I use when building PC's. I may use liquid pro in the future but bear with me on this. These units shipped with thermal paste similar to what I used and at least some operate properly with just that. I should at least be able to idle the machine without overheating my chips right? Right! So I reapply and put everything back together. This greatly improved the issue on one of the boards but not the other. This basically tells me that my theory was correct about the paste. Properly applied thermal paste is adequate as long as there is not some other issue affecting the cooling system.

Now it was time to dig a little deeper. I confirmed both my pumps were running but they were making some noise. Water pumps should make a quiet hum like the sound of a motor running or virtually no sound at all. You may be able to hear water moving through it you listen carefully. Mine were making a sort of groaning/creaking sound. Pump RPM was OK so that wasn't the issue. I decided to disassemble the water blocks.

Many water blocks use jet plates and rubber/silicone inserts to direct/even the flow of water through the block. That's all well and good (even effective sometimes) but the inserts on these machines appear to be a problem or at the very least they develop one over time.

Upon inspection the silicone insert was pretty warped but not too dirty or anything like that. I decided to remove it per some posts I read here. I also noticed that there wasn't much coolant in the block which seemed strange. I got some quality coolant and filled the block. Put it back together, ran the pump, shook the pump, repeated until the block wasn't empty after a run. Now it runs silently and as of this posting I can idle the chips without them getting hot.

In summary of heating issue 1:

1. Check your pumps. You can get the RPM from the UI if your unit is online but you should still check the sounds they make. Otherwise wait for your machine to completely cool, unplug the delta fans on the board you want to check, unplug 1 of the water blocks so only one will start and plug in the power to that board. You can also bench test the pump if you like. It's a little safer but then you don't get to learn if the pump plug on the board is working properly. They should make a quiet hum. Anything else is a problem and you need to investigate. Mine were low on water and the insert was warped so I believe it was restricting flow. Basically if you have good paste and the chips are still getting hot you probably have an issue here. Also make sure you don't have any leaks in the lines or radiators.

2. Check your paste. LiquidPro is the best but it's also difficult to apply. I will post pics later of my temps running on non-conductive paste. It's fine if everything else is working properly.

Now for issue 2. The first miner also was getting the D45 red light on the board furthest from the power supplies. I can't take all the credit for this one. Thanks here goes to numnutz2009 for his excellent post about shorted boards and mosfets. In my case the D45 red light was caused by a blown mosfet on whatever code the D44 light corresponds to. It's the second from the front on the right side when looking at the miner from the rear. There are 4 mosfets for each chip (2 on top, 2 on bottom) and one of them was blown. This one was easy to spot. It was burnt out on the top. This doesn't always happen though. Sometimes they blow out the bottom and you need to look for a gob of solder sticking out from under the side. More on that later. I ordered replacement mosfets from Texas Instruments and will replace as soon as my rework station comes in. You can order them directly from TI or you can look them up on DigiKey. The part number is CSD16342. If you are half way decent with soldering you should be able to use a rework station to replace them. Otherwise numnutz2009 has an open offer to repair them for you.

Issue 3. Another minor issue I had was the front status LED board. Upon opening the case I found a chip laying on the bottom of the case and one hanging by copper filaments from the status board. This was pretty simple, just solder them back on. You need a pinpoint tip but they go back on pretty easy. If your filaments are ripped up like mine were you may need to use some thin copper wire to remake the filaments. It took about 30 minutes for me to reapply both chips. I will confirm orientation later but I believe the little circle on the chips needs to be toward the bottom.

Now for miner #2 that was supposed to be a good replacement for the crap one I was sent.

I plug the miner in and one boards lights up just great. All 8 chaser lights, good temps, good fan speed, the works. This board is cherry. It will hash at a solid 800+ in the 50-60C range. Great! Plug the other one in. Orange light of death. The board won't even start.

There are 2 lights on the power supplies. Left is AC OK, Right is DC OK. I had the right one going orange. which meant a fault on the output side. Obviously something was shorted. Again thanks to numnutz2009 for pointing me in the right direction here. First thing I pulled out the power supply and plugged it in. Both lights green. Then I removed the terminals from the CTA board and placed them in a safe location so they would not short against the case and plugged in the power supply again. Still no fault which told me the little collector board in front of the power supply was also OK. Definitely a fault on the CTA board.

Now per the numnutz2009 post I attempt to hook up a bench power supply to blow out whatever component is shorted by providing unmetered power to the board. Short was still too great and it shut down my bench power supply. Now it was time to go mid-evil. I took the board out and hooked it to a car battery. One big flash and a pop later and my board was alive on 7 chips. D45 red light because of the mosfet I just blew out. Incidentally it was the same chip as above. Second from front on right when looking at the board from the back. This one however, did not show any signs of burning out from the top. This one actually blew out the bottom. It took me a while to find it but eventually I noticed the gob of solder sticking out the bottom of the mosfet. That's the culprit there and that mosfet will be replaced with the other when my rework station arrives. The board is working well on 7 chips though.

Couple notes about intentionally shorting your board.

1. Wear safety glasses. You are about to make an electronic component overheat and explode. Molten solder could be blown in your face. Wear safety glasses.

2. If using a bench power supply be sure it has overload protection or you could blow that out instead of your board. I'm amazed how much current that little mosfet was able to pull before blowing.

3. If using a car battery like I did, make sure the car is not on! Cars charge the battery at 13 volts or more and that MAY be enough voltage to damage something on your CTA board. I checked the power supplies on the miners I have and they put out a consistent 11.99. The battery I used was at 12.2 which I figured (very scientifically by assuming) was close enough to be safe.

4. Make sure you have the polarity correct. All of my boards have positive closest to the edge of the board and negative further in. This is likely the case on ALL of the boards. Reversing polarity is a BAD idea and will not under any circumstances fix your board. In fact it will probably destroy it. Electrical connections are polarized for a reason.

I ended up taking the one totally good CTA from each miner and putting them into one. The frankenstein miner is hashing at over 1500 on power level 8 and has run stable for 3 days straight. Tonight I will complete my initial repairs on the other 2 and I expect to see speeds around 13-1400 until I get the mosfets replaced.

That's all I have for now. If you have questions please let me know. I will post pictures later this week of everything and how it's all working. I will monitor this thread occasionally as well for others that need help.

If this post helped you and you want to send me some BTC love my address is: 1Q6JZq2gaVmAUyVWNuWgyWtNn3YTaTrygt

I can't post but I could send you, send me a PM
I believe it is not a pump, because a lot of troubles in 2 cta's, i don't know what can i do with that

Mine are all powered down since the price went below $300. Looking for $400 before I power up again.

can you post 8.15 firmware.  Obviously can not download it from their website.

I am guessing you have one bad pump.  Just unplug one at a time and boot.  If everything looks the same with the pump unplugged then that is your guy.  If nothing comes on then you have the good pump unplugged.

Just curious how many out there are still running terraminers or all they all shut off?

Hi
Now I have the same problem. I Bought 2 TMIV. Once is good always
Another worked for one week then it started with troubles, Cointerra support said me tO upgrade 8.15 firmware, that killed machine for one month then I minned through PC with USB, but Cointerra Support help me and after one month BBB worked again, but boards are with troubles, the CTA0 is working without D40 light 630 gh, another CTA1 is only mining 280 GH and D37 to D41 are working (4 Yellow Lights turned on), now I have 8.15 FIRMWARE WORKING but machine still with the same troubles. I touch Pumps; in CTA 0 are working 2 pumps but I felt only 3 hoses are working when I touched they. CTA1 is working only One hose from the one pump which is working. Of course Like you said D45 show that troubles. But when I turn on machine, I can see all lights are turning on for once time.
- I would like to know which is the problem. Pumps, Power sources, boards, cables?

I know this is a little old but I wanted to put my 2 cents in for future help. I also had a red D45 and a core not working. I was able to remove the pump and smack it with a screwdriver a few times and get the pump working again. If that pump does not work, the core will be shut off and that light will be red. Not very scientific, or high tech, but a pump is a pump. When they dont move, smack em!

does anyone know of another way to connect cgminer to the terraminer boards?  serial?

several boards of mine have usb connection issues.

thanks.  i guess i shoyld give google a try.