wolfey2014
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 14, 2014, 03:28:55 PM |
|
Just curious....but is the 5v USB fan mod on these gridseeds that have the failed USB ferrite bead? It could be the added current usage of the fan stresses it enough over time to the point of failure as I'm sure the USB circuitry wasn't designed with the load of the fan in mind. I'm not saying it can't handle it, because there is plenty of power from the USB port for a fan, but given the build quality of these things it really wouldn't surprise me.
Accidently overheating the FB while soldering the + fan wire to it can cause pre-mature failure. It only takes a few seconds of too much heat to burn one up, I am sure. Besides the fan is only pulling 78mA, if that while running and maybe 140mA startup which lasts less than 2 seconds to spool up and stabilize. By the way, I forgot to mention that I revised the 5V USB Fan power mod recently. Only days ago actually. So I'm late posting it. Sue me! ![Tongue](https://bitcointalk.org/Smileys/default/tongue.gif) On the last few batches of pods, the fan + is now connected to and powered by 12V input via 5V regulator on the PCB. See pic. snip You can make the change easily if you wish to. I've been powering 6 of my miners via 5V USB for months now, for hundreds of hours straight now. No problems with the FB's! ![Wink](https://bitcointalk.org/Smileys/default/wink.gif) Just be very careful not to overheat the components you are keeping on the pcb. I know I didn't overheat any of them when soldering to them. It takes me less than a second to solder the wire to the tinned connections. All is well. By the way, all 20 were the latest plain aluminum colored pods. All heat sinks were evenly touching their respective chips. No components were on a different plane i.e. higher than the 5 main uP chips. Looks like they got the components mounted right on this run. See pic. snip Look like the crystal is still high, you can clearly see the componet imprssion in the thermal pad. and several other componet impression as well. Overheating could be a big factor. just curios thoose that have blown FB how are you running them? wiht 5v fan mod?, wiht no fan? top heat sink removed? its possible that the 5v fan mod isnt providing enough air flow for the FB causingthem to overheat. likewsie if there is no fan. if it overheating though lack of air flow, take all facotrinto account, ie climate, room temperture humidity, location of grids . I know runnign them wihtout the top sink can cause them to overheat, i was testing them had top heat sink off, and fans (120mm) running next to them, which got knocked and i didnt notice and im pretty sure i fried a few grid chips, tried reflow to no use. the chips are recongised by cpuminer all of them but only 2 are hashing. anyone know where i can get a few grid chips from to replace the cooked ones? The impressions the other components make are not keeping the main chips from being contacted hence cooled properly, is the point. NONE of my original 6 miners of months old 24/7 operation have had ANY problem whatsoever with the 5V USB mod! NONE! Where they are located right now, they are running at a normal 98*F to 115*F and are just fine. If you suspect your failed FET's, which are most likely the cause of the blown FB's, is due to lack of cooling, you are certainly welcome to revert back to 12V powered fans. It will be noisy but it will cool the pod better, obviously.
|
I Modify Miners Professionally! PM me for details!
|
|
|
amix
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 84
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 14, 2014, 04:28:58 PM |
|
im on 12v thermal take 120mm fans so got cooling and no noise! ![Smiley](https://bitcointalk.org/Smileys/default/smiley.gif) i was just bouncing ideas and suggestions around wolfey, trying to see a common factor wiht the blown FB that all.
|
|
|
|
wolfey2014
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 14, 2014, 04:59:05 PM |
|
im on 12v thermal take 120mm fans so got cooling and no noise! ![Smiley](https://bitcointalk.org/Smileys/default/smiley.gif) i was just bouncing ideas and suggestions around wolfey, trying to see a common factor wiht the blown FB that all. I meant the fans that come with the pods.... ![Roll Eyes](https://bitcointalk.org/Smileys/default/rolleyes.gif) They ARE noisy @ 12V!
|
I Modify Miners Professionally! PM me for details!
|
|
|
cmilian
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 93
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 15, 2014, 05:57:40 AM |
|
Guys after doing the Volt mod and running at 1200Mhz for 3 weeks with no problems all the sudden I am getting lots of Nonce errors.
Any ideas on why? would it be a small change in room temperature?
|
|
|
|
gtraah
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 15, 2014, 11:22:29 AM |
|
Hey guys I fixed my other miner with the blown Ferrite bead, This was easy to fix because I knew what the problem was, now all miners are fine until I noticed 2 miners in the back of the group with the green lights flashing but the fan not spinning , Im like HUUUUHHHH So I pulled it apart and guess what was blown
2 Ferrite beads the ones Close together Not the ones that a far apart The ones that are Right next to each other sitting on the board like 2 ferrite twins, infact I saw another user on here claiming the same thing happening. I have a strong feeling that the cause for this was a bad power lead / Barrel ending, I now have thrown this in the bin.
I am sure the people who may give me advise will know what I am talking about,, luckily I have the same ferrite beads as I have ordered these to repair the 1 x ferrite bead that I blew because I shorted out the red wire accidently.
Anyway I have now wasted 4 ferrite beads ... Basically the copper pads where these ferrite beads are meant to go are fukd, although they are fucked I was able to solder onto as I think there is still contact to the main board.. problem is when I did this and plugged the an on I instantly see sparkes and the ferrite bead goes RED hot I mean this all happens in 1 second and then I unplug it. Now I would like to ask the wizards in here, instead of throwing away 2 seeds as I feel I have 10 CHIPS that can mine but just need a way too, Maybe I can lower the frequency I dont know can I get some help here on a way I can salvage these 2 pods?
Visible issue 2 ferrite beads on each board blown, 2 boards & 2 ferrite beads = 4 ferrite beads Gone.... I do have spare ferrite beads but if there is a way to bypass these beads somehow I feel it wold be better, Even If I run these 2 seeds at 800Mhz And 5v Fan Or something...
Please some advise would be great
Hello, are you based in EU? thanks I am Based in OCeania - Australia - Sydney ![Smiley](https://bitcointalk.org/Smileys/default/smiley.gif) ![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2FA6nOstW.jpg%3F1&t=665&c=4zk_yivGtzr7Uw) check the continuity here and if it is shorted, take off the mosfet with hot air gun. make sure you take the pcb off from bottom heatsink and pad, otherwise it will not come off. i used 300 c with full air speed. took around 2 min and the mosfet comes off . after take this off check your continuity between 12v+ and gnd. you shouldn't have any short. replace the parts and put your FB back on , you will be good to go(unless there any other parts demaged!) The top yellow arrow is the first pin on top which is number 4(gate) and bottom 1-4 pin is Drain i think. if you find a short between gate and drain in a N channel mosfet, it known to be faulty. Dammm I dont have a multimeter to check the continuency... By the way, after I modded each pod I tested them, I waited for 3 accepted shares and then assembled them back together. If the mofset was faulty from birth dont you think this would have burnt up on the moment I plugged in to test.... One other thing, lets just say I burn the contact pretty bad basically between both FB the ones very close together , you can see the copper on the board all around these 2... I was wondering if because of this situation does these mean that its a constant short between 2 FB, I tried my best to make the solder not touching the other FB so each Dab of solder is on its own with each side of the FB, the solder is not connecting to one another but like I said underneith it is bare copper does this that both FB are shorting out via the copper? Also is there a way to do what you said without a multimeter?? And are you saying that the Gridseeds can do without the mofset? Does anybody know what I can do if I do not have a multimeter? I have got 2 pods just sitting there, These few posts above tell the story, so far I have been told to do something but I do not have a multimeter to check continuity, is there anything else I can do? I or is what is said above the absolute only thing i can try at this point?
|
|
|
|
nst6563
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 15, 2014, 12:26:51 PM |
|
Yes, you really need a multimeter in order to test continuity. If you're doing anything with electronics, you should have a multimeter that does at least basic testing including continuity. It's an investment in a tool you'll find very handy quite often.
|
|
|
|
amix
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 84
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 15, 2014, 02:29:42 PM |
|
Does anybody know what I can do if I do not have a multimeter? I have got 2 pods just sitting there, These few posts above tell the story, so far I have been told to do something but I do not have a multimeter to check continuity, is there anything else I can do? I or is what is said above the absolute only thing i can try at this point?
in time of need ive done a a diy tester. basically battery two bits of wire and a light bulb, or led and resistor. or if you have a tourch thats metal unscrew the bottom one wire onto the case the other onto the battery termanial. granted not great but for testing continutity works. few corc clips help here, or paper clips to hold things in place. you could use a buzzer as well. you just want somethign that will show you theres a complete circuit, light, led (use resistor in series!) buzzer EXCIRSE CAUTION AND COMMON SENSE THOUGH. ALSO BE AWARE USING TOO HIGH A VOLT BATTERY IE 9v COULD CAUSE DAMAGE.
|
|
|
|
toxic0n
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 413
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 16, 2014, 04:47:03 AM |
|
Well, I think I've got another case of blown FB24 and FB23. Modded 10 pods last night and one refused to hash although it was recognized by cpuminer. Opened it up and here is what I saw ![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2FV4HjDxYl.jpg&t=665&c=AgKu9xRXSyTfBg) ![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2FAMAvMYkl.jpg&t=665&c=0rQDNbo1ZVrK8A) I also noticed some bubbling on the centre chip and brown residue on the thermal pad. I'm guessing the chip is gone, what are the odds the rest will work if I replace the beads? Anything I can check to confirm? Should I avoid using that barrel plug? It was working fine before the mod. I've got a multimeter and am not afraid to use it. Thanks in advance!
|
|
|
|
gbyg
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 71
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 16, 2014, 04:55:38 AM |
|
Well, I think I've got another case of blown FB24 and FB23. Modded 10 pods last night and one refused to hash although it was recognized by cpuminer. Opened it up and here is what I saw ![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2FV4HjDxYl.jpg&t=665&c=AgKu9xRXSyTfBg) ![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2FAMAvMYkl.jpg&t=665&c=0rQDNbo1ZVrK8A) I also noticed some bubbling on the centre chip and brown residue on the thermal pad. I'm guessing the chip is gone, what are the odds the rest will work if I replace the beads? Anything I can check to confirm? Should I avoid using that barrel plug? It was working fine before the mod. I've got a multimeter and am not afraid to use it. Thanks in advance! Just check the continuity , if there is no short between 12+ and gnd, it should work.
|
|
|
|
happydaze
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 16, 2014, 11:10:45 AM Last edit: May 16, 2014, 11:28:12 AM by happydaze |
|
toxicOn - that one looks like the red fan wire got pinched on the edge of the heatsink and caused a short. You can see that the thermal pad did not make contact with the gridseed chips on the side where the missing red fan wire should be as if the red wire was caught under the edge. And you can see the burn mark on that edge.
You may be able to replace the other burnt parts and get it working, I don't know. The pods have dropped in price so much that you might be better off just replacing it.
I have one pod where the center chip ended up burning up (because of a short) and it continues to hash away (center chip removed) on two chips at 1200MHz, 200+ kh/s. Be cautious though, not worth starting a fire. At first mine worked but the amps were way to high for me to leave it unattended. It's around 10 watts on 2 chips at 1200MHz now so I feel better about it.
If you are not 100% confident in it don't risk causing a fire. In the photo, it looks to me like there may be other shorts along the path of the missing red wire.
|
|
|
|
toxic0n
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 413
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 16, 2014, 02:25:44 PM |
|
Hey, thanks for the response! I don't think the red fan wire got pinched, it examined it closely and see no marks on it of any kind. I think it got blown off when the ferrite bead exploded and that's what left the burn mark on the edge of the thermal pad as well.
|
|
|
|
wolfey2014
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 16, 2014, 02:44:02 PM Last edit: May 16, 2014, 02:59:21 PM by wolfey2014 |
|
Well, I think I've got another case of blown FB24 and FB23. Modded 10 pods last night and one refused to hash although it was recognized by cpuminer. Opened it up and here is what I saw ![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2FAMAvMYkl.jpg&t=665&c=0rQDNbo1ZVrK8A) I also noticed some bubbling on the centre chip and brown residue on the thermal pad. I'm guessing the chip is gone, what are the odds the rest will work if I replace the beads? Anything I can check to confirm? Should I avoid using that barrel plug? It was working fine before the mod. I've got a multimeter and am not afraid to use it. Thanks in advance! I think that your heat sink was not seated properly on the center chip and that is why it fried after you over-volted it. You can clearly see that there are no indents where the chips are supposed to contact with the top sink plate via the pad. The center chip over heated (as it would be hotter when overclocked) shorted out and the 2 FB's released their magic black smoke. That's all that happened. No big mystery. Try re-seating the top half heat sink so it touches the chips like it's supposed to. Otherwise it will just act like a air flow shield and keep the chips from being cooled at all. Of course, replace the bad chips first ![Wink](https://bitcointalk.org/Smileys/default/wink.gif) Good luck!
|
I Modify Miners Professionally! PM me for details!
|
|
|
toxic0n
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 413
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 16, 2014, 05:29:48 PM |
|
Thanks, Wolfey, that makes sense. Do you know if the rest of the chips should work if the blown one is removed? Would like to test them first before trying to replace the chip.
|
|
|
|
gtraah
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 16, 2014, 05:48:33 PM |
|
|
|
|
|
wolfey2014
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 16, 2014, 06:05:27 PM |
|
Thanks, Wolfey, that makes sense. Do you know if the rest of the chips should work if the blown one is removed? Would like to test them first before trying to replace the chip.
Well, you need a good DVM so you can test for shorts between 12V + and - after you remove the bad chip If it's open (not shorted) then you may have a good chance at the other 4 being good. Only way to tell is to try it. But I'd put a 2 or 5 amp fuse on the 12V input power just in case it's still shorted somewhere. Others have already done this surgery so maybe they can chime in and tell you what they did specifically. I haven't had any problems with my pods so I can't help you much other than what I've suggested thus far. Good luck and god speed!
|
I Modify Miners Professionally! PM me for details!
|
|
|
wolfey2014
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 16, 2014, 06:23:38 PM |
|
I haven't had to salvage any popped seeds yet so.... But I think Amix or Zig or HappyDaze can give you the correct answer. God speed!
|
I Modify Miners Professionally! PM me for details!
|
|
|
nst6563
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 16, 2014, 08:02:35 PM |
|
Some of the specs on that part look excessive. You'll want to match the specs as closely to this as possible: https://www.dropbox.com/s/w4m5vu4k5jgopk8/qm3022m6_ubiq.pdfIt's basically a 30v N-Channel mosfet with a 7mOhm RDSon and a 76A ID@25c (Continuous current drain at 25c). You'll want to match the gate charge as close as possible as well. I burnt out that same mosfet (3022m) and replaced it with the following: http://www.ti.com/product/CSD17552Q5A. I blew my mosfet when curiosity got the best of me and I tried to dual mine with the voltmod at 1100Mhz. Moral of the story...don't do that. Don't even think of doing it. I replaced my mosfet with the one from TI and it's back up and running but it's a little crippled. Chip 3 seems to generate some errors no matter what speed, but it still hashes and I still get around 439khs from that gridseed.
|
|
|
|
gtraah
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 17, 2014, 03:02:21 AM |
|
my mofset is not blown its the 2 FBs But everytime i place new FB, 5 seconds later They crackle and pop again. I was told it is most likely a faulty mofset causing this.. Man that site is awesome I actually ordered 3 mofsets as samples for free !!
|
|
|
|
Ferris419
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 17, 2014, 03:07:20 AM |
|
First looks like: LISTING: 1 element(s)
< COMPONENT INSTANCE >
Reference Designator: U29 Package Symbol: SOT23-6
Component Class: IC Device Type: PCF8574_P_0_SOT23-6_74HC2GU04 Value: 74HC2GU04
Placement Status: PLACED origin-xy: (1467.57 867.87) rotation: 0.000 degrees not_mirrored
Function(s): Designator: F8 Type: PCF8574_P_0_SOT23-6_74HC2GU04 Pin(s): 1, 6, 3, 4, 2...
Properties attached to component definition VALUE = 74HC2GU04
Pin IO Information: Pin Type SigNoise Model Net --- ---- -------------- --- 1 UNSPEC CLK 2 UNSPEC DGND 3 UNSPEC CLK 4 UNSPEC N31664886 5 UNSPEC OVDD33 6 UNSPEC N31664859
Second looks like: LISTING: 1 element(s)
< COMPONENT INSTANCE >
Reference Designator: U9 Package Symbol: SOT23_5
Component Class: IC Device Type: SN74AHCT1G125_0_SOT23_5_UP0111 Value: UP0111
Placement Status: PLACED origin-xy: (296.48 1497.43) rotation: 90.000 degrees not_mirrored
Function(s): Designator: F4 Type: SN74AHCT1G125_0_SOT23_5_UP0111 Pin(s): 3, 4, 1, 5, 2
Properties attached to component instance ROOM = OCD_CRT
Properties attached to component definition VALUE = UP0111
Pin IO Information: Pin Type SigNoise Model Net --- ---- -------------- --- 1 UNSPEC OVDD33 2 UNSPEC DGND 3 UNSPEC OVDD33 4 UNSPEC N31931394 5 UNSPEC DVDD_PLL
Hopefully you or one of the experts here know what they are because I don't. I have had 0 luck finding these. Can an expert here help me to find these or find something that will work to replace these? I can't believe I'm the only person here who had a problem with them getting hot and breaking.... These are 2 seperate pods and both chips look the same like the got hot and cracked ![Huh](https://bitcointalk.org/Smileys/default/huh.gif)
|
Bitcoin is gonna hit 100K usd
|
|
|
nst6563
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 17, 2014, 03:55:08 AM Last edit: May 17, 2014, 04:42:10 AM by nst6563 |
|
First looks like: LISTING: 1 element(s)
< COMPONENT INSTANCE >
Reference Designator: U29 Package Symbol: SOT23-6
Component Class: IC Device Type: PCF8574_P_0_SOT23-6_74HC2GU04 Value: 74HC2GU04
Placement Status: PLACED origin-xy: (1467.57 867.87) rotation: 0.000 degrees not_mirrored
Function(s): Designator: F8 Type: PCF8574_P_0_SOT23-6_74HC2GU04 Pin(s): 1, 6, 3, 4, 2...
Properties attached to component definition VALUE = 74HC2GU04
Pin IO Information: Pin Type SigNoise Model Net --- ---- -------------- --- 1 UNSPEC CLK 2 UNSPEC DGND 3 UNSPEC CLK 4 UNSPEC N31664886 5 UNSPEC OVDD33 6 UNSPEC N31664859
Second looks like: LISTING: 1 element(s)
< COMPONENT INSTANCE >
Reference Designator: U9 Package Symbol: SOT23_5
Component Class: IC Device Type: SN74AHCT1G125_0_SOT23_5_UP0111 Value: UP0111
Placement Status: PLACED origin-xy: (296.48 1497.43) rotation: 90.000 degrees not_mirrored
Function(s): Designator: F4 Type: SN74AHCT1G125_0_SOT23_5_UP0111 Pin(s): 3, 4, 1, 5, 2
Properties attached to component instance ROOM = OCD_CRT
Properties attached to component definition VALUE = UP0111
Pin IO Information: Pin Type SigNoise Model Net --- ---- -------------- --- 1 UNSPEC OVDD33 2 UNSPEC DGND 3 UNSPEC OVDD33 4 UNSPEC N31931394 5 UNSPEC DVDD_PLL
Hopefully you or one of the experts here know what they are because I don't. I have had 0 luck finding these. Can an expert here help me to find these or find something that will work to replace these? I can't believe I'm the only person here who had a problem with them getting hot and breaking.... These are 2 seperate pods and both chips look the same like the got hot and cracked ![Huh](https://bitcointalk.org/Smileys/default/huh.gif) U29 is a "dual unbuffered inverter" http://www.nxp.com/products/logic/buffers_inverters_drivers/series/74HC2GU04.html - see the ordering tab on that same page I posted that back on May 14, have you tried that link and through the Order tab? If so you should have seen that Mouser carries that part (direct link: http://www.mouser.com/ProductDetail/NXP-Semiconductors/74HC2GU04GW125/?qs=sGAEpiMZZMutVWjHE%2fYQwzJv2zzn7Nf7MNFQIOCd6fU%3d) That at least gets you one part. I'll try to see if I can locate the other (not likely since UPI is almost exclusively overseas) or see if I can find a close match as a replacement. :EDIT: I did some quick searching and it looks like ON Semiconductor has a very close match to the original part. It has a higher accuracy rating but a lower output current (by 100mA....may or may not be a concern). http://www.digikey.com/product-detail/en/CAT6219ADJTD-GT3/CAT6219ADJTD-GT3-ND/2271436Another possible match (Same specs except a slightly higher minimum input voltage) http://www.silergy.com/productsview/sy6306b Those are the closest matches I could find. Maybe someone else can try if they haven't already. If I have time tomorrow I can check some more.
|
|
|
|
|