So; after doing a thermal paste job I'm stuck with the no LED issues (beagle boots just fine). As I don't have that J11 wire on either board, any suggestions for troubleshooting? I think I'll be doing a proper clean and inspect later of the boards.
I've applied thermal paste before, and thinking I may have made a mistake there I took the water block back off one of the board; I was impressed with my work.... so I'm not sure where the issue is.
Topic: Cointerra Hardware Support **Unofficial - page 18. (Read 56948 times)
Here is a pic I circled J11. If you dont have that maybe your version of the boards is different. These boards are printed with model numbers and revisions
I have GSA CARD ASSY: COIN-500-013 REV.B. I had a leak also, it was a small crack in the hose for me. You should be able to locate that leak and fix it just how I did
I have GSA CARD ASSY: COIN-500-013 REV.B. I had a leak also, it was a small crack in the hose for me. You should be able to locate that leak and fix it just how I did
Scriber.. when was this unit shipped as I have 2 of them, first of April, and I have no J11 wires...
I am curious what this power is going to, and why it is needed on your boards. Following the leads on the board I have pulled now..
glad your making progress and fixing your problem. Now we just need to figure out our issue with the red LED light
Ok, to summarize this. If you roll back the firmware to 6.32 and pay attention to the actual beep codes before it goes all red are they the same. Basically, When you first fire, all leds light up, and start a countdown from right to left. Pay attention to where it fails and the tone the boards makes. Different firmware will do it at different times and lengths. Please try and be specific as I am getting close to figuring them out as well as the push button series.
Then try 6.48 and see what results are... 6.48 seems to be the most stable for unstable boards. Did that make sense? Actually it did. 7.6 is for a board that is fully functional, and will push it to the max straight out the box. Sadly, they have as of yet released new firmware. I am of the belief they could give a shit about anyone over 30 days or broken sticker.
Cointerra really fucked this pooch in this aspect. They had a chance to be the American leader in this market and shipped shit and will not support it. BFL screwed folks by delivery date.. Cointerra shipped shit.
glad your making progress and fixing your problem. Now we just need to figure out our issue with the red LED light
finally got it!! took out both white rubber things from the inside of each pump/waterblock. Both were gunked up with white "butter"/paste type of sustance cleabed it all out and now the pumps are pumping. All hoses are same temp. I left both the rubber/silicon out. But the temps are very high still. But atleast its hashing. I think its time for liquid pro. thats my last hope to cool these down. that or building a custom watercooling system and seeing if that does it. It seems like these pumps dont move alot of water.
http://i.imgur.com/UL3mCx8.jpg
http://i.imgur.com/UL3mCx8.jpg
Here is a pic I circled J11. If you dont have that maybe your version of the boards is different. These boards are printed with model numbers and revisions
I have GSA CARD ASSY: COIN-500-013 REV.B. I had a leak also, it was a small crack in the hose for me. You should be able to locate that leak and fix it just how I did
http://i.imgur.com/zHI53vg.jpg
I have GSA CARD ASSY: COIN-500-013 REV.B. I had a leak also, it was a small crack in the hose for me. You should be able to locate that leak and fix it just how I did
http://i.imgur.com/zHI53vg.jpg
Hey guys thanks for starting this thread, seeing as Cointerra shutdown the forums this is the only place i can find help. I've been running an April batch for two months now and today one board did not come online. I opened it up to see why only one board was going online and noticed that LED D45 was red. Does anyone know what this means? and is there is a way to fix it?
Is D44 (yellow) also lit next to D45(red)? Did you ever fix your problem? see my latest comment what Cointerra told me, it might fix your problem!
I only have D45(red) on. I tried all the suggestions that have been given on this thread (used compressor to clean the board, submerged with 99.99 alcohol and lightly cleaned with toothbrush, disconnected the J11 cable) so far none have worked. The board powers up with D37 through D44 (yellow) and D45(red). When the miner starts working D37 through D44 turn off but D45 stays lit. I got a reply from cointerra asking to send them a pdf of my stat page which i sent and I'm waiting for a reply to see what they say.
i havent seen anything connected to j11 on any of the rigs ive seen either, can anyone who has these cables tell us where theyre connected to?
J11 is on both boards. It is the 3pin black and white wire that plugs in directly behind the intake fans. This J11 wire runs into a small board next to the power supplies; a board located in the side compartment with the power supplies. One of my J11 plugs had the black rubber crap over it to hold the plug in so yours might be covered but its there. Here is the PDF they sent me. I have not tried pressing the reset button and I dont hear any clicking. I have been keeping this board powered off since its not working. Not need to have power running to it. I do have two other units running also below this one. I noticed there is a reset button on the outside front center top of the case and each internal board also has a button but I dont know what this one does. As far as one board having a jumper and one not I think this is just to signify and master and slave board
There are no J11 cables in either of my units. It is simply a dead plug as far as I can tell. The beaglebone has it own wire that can be plugged into the top of any PSU module.. Is This what you refer to? The board itself a self check, or built in bios, if anything fails, you will see red 45 led. I actually had my secondary board up for an hour, power stepped it 10 times, and turned the power off to see if other core would hit, and lost it. I have since pulled the board and low and behold, I have oil under the board. A slow leak from the front pump. I am cleaning as we speak, and will not plug it back in till my liquid pro is here tmw. If I do not figure out the pump leak, I will dunk this bitch in mineral oil and fire it.
J11 is on both boards. It is the 3pin black and white wire that plugs in directly behind the intake fans. This J11 wire runs into a small board next to the power supplies; a board located in the side compartment with the power supplies. One of my J11 plugs had the black rubber crap over it to hold the plug in so yours might be covered but its there. Here is the PDF they sent me. I have not tried pressing the reset button and I dont hear any clicking. I have been keeping this board powered off since its not working. Not need to have power running to it. I do have two other units running also below this one. I noticed there is a reset button on the outside front center top of the case and each internal board also has a button but I dont know what this one does. As far as one board having a jumper and one not I think this is just to signify and master and slave board
http://i.imgur.com/5rb2Jen.png
http://i.imgur.com/5rb2Jen.png
I got a reply from Cointerra and they have been emailing me trying to help. They first replied with I could not send the board or unit in for repair because I opened the unit and voided the warranty. They attached a pdf for trying to fix my problem with instructions that said to shutdown the unit, remove the cable from J11 on the offline board and then power the unit up. If the LEDs start the unit is all good. Leave the cable off and im done. I did this and the LEDs lit up then they immediately turned off in sequence and only D44 and D45(red) LED's stay lit. Cointerra asked me if the ground cable was connected. Yes it is. Cointerra says he has note seen many units with this d44,d45 led issue and is trying to figure out what is causing it. I have not heard anything else.
UPDATE: Just tried this process a few more times again and noticed only D45(red) stays lit solid
UPDATE: Just tried this process a few more times again and noticed only D45(red) stays lit solid
One more thing, have you played with the push button towards the back of the boards. I can tell you this, even if you are led45 dead, push it once and it changes the error codes. I am starting to believe it has to do with the check sequence of the board.
I got a reply from Cointerra and they have been emailing me trying to help. They first replied with I could not send the board or unit in for repair because I opened the unit and voided the warranty. They attached a pdf for trying to fix my problem with instructions that said to shutdown the unit, remove the cable from J11 on the offline board and then power the unit up. If the LEDs start the unit is all good. Leave the cable off and im done. I did this and the LEDs lit up then they immediately turned off in sequence and only D44 and D45(red) LED's stay lit. Cointerra asked me if the ground cable was connected. Yes it is. Cointerra says he has note seen many units with this d44,d45 led issue and is trying to figure out what is causing it. I have not heard anything else.
UPDATE: Just tried this process a few more times again and noticed only D45(red) stays lit solid
UPDATE: Just tried this process a few more times again and noticed only D45(red) stays lit solid
BTW, are you hearing the beep tones now scriber? or the throttling of the dead boards fans?
I got a reply from Cointerra and they have been emailing me trying to help. They first replied with I could not send the board or unit in for repair because I opened the unit and voided the warranty. They attached a pdf for trying to fix my problem with instructions that said to shutdown the unit, remove the cable from J11 on the offline board and then power the unit up. If the LEDs start the unit is all good. Leave the cable off and im done. I did this and the LEDs lit up then they immediately turned off in sequence and only D44 and D45(red) LED's stay lit. Cointerra asked me if the ground cable was connected. Yes it is. Cointerra says he has note seen many units with this d44,d45 led issue and is trying to figure out what is causing it. I have not heard anything else.
UPDATE: Just tried this process a few more times again and noticed only D45(red) stays lit solid
UPDATE: Just tried this process a few more times again and noticed only D45(red) stays lit solid
Ok, now we are getting somewhere. Problem is I have no j11 plug on either board. Where does your j11 cord go? What does it connect to? I am solid red 45 as well now. But get this, an hour ago I reset the machine, and half the board was working! I power stepped it 5-10 times, all was well but only half the board. I then shut the power down in an attempt to get the other half working and bam. Back to dead board. I will say this though, it seems I can only get the board working with 6.48. I have not seen it at all in 6.32 or 7.6 in over a week.
Please post more info or copy the pdf scriber.. we are getting there, I know we are!
Hey guys thanks for starting this thread, seeing as Cointerra shutdown the forums this is the only place i can find help. I've been running an April batch for two months now and today one board did not come online. I opened it up to see why only one board was going online and noticed that LED D45 was red. Does anyone know what this means? and is there is a way to fix it?
Is D44 (yellow) also lit next to D45(red)? Did you ever fix your problem? see my latest comment what Cointerra told me, it might fix your problem!
I got a reply from Cointerra and they have been emailing me trying to help. They first replied with I could not send the board or unit in for repair because I opened the unit and voided the warranty. They attached a pdf for trying to fix my problem with instructions that said to shutdown the unit, remove the cable from J11 on the offline board and then power the unit up. If the LEDs start the unit is all good. Leave the cable off and im done. I did this and the LEDs lit up then they immediately turned off in sequence and only D44 and D45(red) LED's stay lit. Cointerra asked me if the ground cable was connected. Yes it is. Cointerra says he has note seen many units with this d44,d45 led issue and is trying to figure out what is causing it. I have not heard anything else.
UPDATE: Just tried this process a few more times again and noticed only D45(red) stays lit solid
UPDATE: Just tried this process a few more times again and noticed only D45(red) stays lit solid
My temps at the hottest part of the day (85°C) are around 62°C. So its not getting hot. You can see my temps on page 2 of the thread. Funny thing is it seems to throttle more the cooler it gets. What firmware are you running? I'm running 0.7.6.
I have 0.7.6 too.
I am having the same issue as you guys. One cta goes to 120c right away and shuts down. it was working for for the past 3 months. the other CTA runs amazing over 800gh all the time
the "bad" one always ran hotter. i replaced the paste about 4 times now and like 4 cores go to the sky and it stops. at its different ones just seems like something is shorting out or i dont know what the hell is happing. any input would be great. i used ic diamond themal paste on all chips the "good booard temp are way down and it hashes amazing the "bad board is still HOT!
edit did some tinkering. found that letting the board cool down and then just start the one sounds like a fan in one of the pumps is ticking... also found that the "out" side of both cores on the "bad board" are really hot and the "in"side is cold. On the "good" board both side are equally Warm cound it be a pump issue? both pumps show speed but not as fast as the other board. if the fins inside the pump broke off the the pump would stiill spin just be clogged/not going anyhting?
im really tring to figure this out. already had one of the power blocks that the soldered on the board break off that's why i opened it up in the first place. fixed that and now this heat/power/idk/short issue
the "bad" one always ran hotter. i replaced the paste about 4 times now and like 4 cores go to the sky and it stops. at its different ones just seems like something is shorting out or i dont know what the hell is happing. any input would be great. i used ic diamond themal paste on all chips the "good booard temp are way down and it hashes amazing the "bad board is still HOT!
edit did some tinkering. found that letting the board cool down and then just start the one sounds like a fan in one of the pumps is ticking... also found that the "out" side of both cores on the "bad board" are really hot and the "in"side is cold. On the "good" board both side are equally Warm cound it be a pump issue? both pumps show speed but not as fast as the other board. if the fins inside the pump broke off the the pump would stiill spin just be clogged/not going anyhting?
im really tring to figure this out. already had one of the power blocks that the soldered on the board break off that's why i opened it up in the first place. fixed that and now this heat/power/idk/short issueDoes sound like a pump issue from what you described. Perhaps the impeller is damaged especially if its rpm is not up to speed like your good board. Contact the company who makes the coolers, they are located in canada I believe. Tell them how terrible Cointerra support is and you noticed they are using their product which might have failed on you. Btw I would not use thermal paste on these units, probably a good reason not to is that it continues to overheat and youve done the job 4 times. Use liquid pro its tried and proven here by numerous people including myself
The bad board pumps are 300rpm higher then the good board. why did the paste work so well with my other board? i will have to give the liquid pro a try...
i ended up taking one of the bad board coolers off and took the copper plate off the bottom and found a white rubber/cyclone piece inside. its suppose to work as a channel for the water to go on the copper plate fins i guess. thing is it was gunked up with this white residue and the water had no way to get to the fins... ended u p taking the whole white rubber piece out and now have one core up and hashing at 400gh. still hot as helll but at least its hashing. we will see for how long... o and the hoses on that core are now the same temp..
i will do the second one tomorrow and take some pics for reference.
@mattster - what are your ambient temps? Also, what were your cpu temps before LP? thanks.
I usually don' turn on the A/C till it gets 85° or so. Before Liquid Pro my temps were high, 105°+.
Now?
I am having the same issue as you guys. One cta goes to 120c right away and shuts down. it was working for for the past 3 months. the other CTA runs amazing over 800gh all the time
the "bad" one always ran hotter. i replaced the paste about 4 times now and like 4 cores go to the sky and it stops. at its different ones just seems like something is shorting out or i dont know what the hell is happing. any input would be great. i used ic diamond themal paste on all chips the "good booard temp are way down and it hashes amazing the "bad board is still HOT!
edit did some tinkering. found that letting the board cool down and then just start the one sounds like a fan in one of the pumps is ticking... also found that the "out" side of both cores on the "bad board" are really hot and the "in"side is cold. On the "good" board both side are equally Warm cound it be a pump issue? both pumps show speed but not as fast as the other board. if the fins inside the pump broke off the the pump would stiill spin just be clogged/not going anyhting? im really tring to figure this out. already had one of the power blocks that the soldered on the board break off that's why i opened it up in the first place. fixed that and now this heat/power/idk/short issue
the "bad" one always ran hotter. i replaced the paste about 4 times now and like 4 cores go to the sky and it stops. at its different ones just seems like something is shorting out or i dont know what the hell is happing. any input would be great. i used ic diamond themal paste on all chips the "good booard temp are way down and it hashes amazing the "bad board is still HOT!
edit did some tinkering. found that letting the board cool down and then just start the one sounds like a fan in one of the pumps is ticking... also found that the "out" side of both cores on the "bad board" are really hot and the "in"side is cold. On the "good" board both side are equally Warm cound it be a pump issue? both pumps show speed but not as fast as the other board. if the fins inside the pump broke off the the pump would stiill spin just be clogged/not going anyhting?
im really tring to figure this out. already had one of the power blocks that the soldered on the board break off that's why i opened it up in the first place. fixed that and now this heat/power/idk/short issueDoes sound like a pump issue from what you described. Perhaps the impeller is damaged especially if its rpm is not up to speed like your good board. Contact the company who makes the coolers, they are located in canada I believe. Tell them how terrible Cointerra support is and you noticed they are using their product which might have failed on you. Btw I would not use thermal paste on these units, probably a good reason not to is that it continues to overheat and youve done the job 4 times. Use liquid pro its tried and proven here by numerous people including myself
Is anyone else's Terraminer throttling often? At first I figured it was caused by the high temps but now that I have the temps down it still throttles.
What temp does it reach before it throttles? what's the performance before/after?
I'm running on ICdiamond for about 2 weeks and I haven't had any throttling so far.
Previously the MX4 was good for 3 weeks, so I'm not sure how long the ICdiamond lasts. (the MX4 degraded into "dust" after 3 weeks).
When it disables/throttles the core(s) the temps go down obviously so it might not be evident that it's caused by the degradation of the thermal compound.
If a pump fails then it also disables the cores ( I had one pump fail on me - it produces weird noise before failure)
Having said that, I'm interested in buying problematic Terraminers. Pm me with the stats if any of you consider selling them.
I am down to thinking that I may have a bad pump. Anyone know how to test that? Obviously something is telling the board not to start.
Bad pumps show as 0 RPM in the admin panel, iirc the boards will start with a pump failure.
No offense, but that is obvious. If you have a board that will not start, and it is because of a dead pump, how would you know?
Jump to: