Xian01
Legendary
Offline
Activity: 1652
Merit: 1067
Christian Antkow
|
|
August 06, 2013, 07:50:47 PM |
|
What's that on the top left and on the middle shelf?
Top left is a little USB fan pointing down on a couple Erupter USBs hosted on a DLink 7 port hub. Rest of top is more Erupter USB's plugged into more DLink 7 ports. Shelf below that is two ~13gH/s Block Erupter Blades. Shelf below that is 5 x Anker 10 port hubs, fans, and USB Erupters. Bottom shelf is power supplies.
|
|
|
|
BatmansBitcoins
Newbie
Offline
Activity: 5
Merit: 0
|
|
August 06, 2013, 08:46:12 PM |
|
Kramble, that's an impressive piece of kit. I haven't read the entire ~150 pages , but that's about one of the coolest pieces of kit I've seen on this thread. It's not about the MH/s, its about getting stuff to work. Hat's off to you! Way beyond my "Electrotech 101" which is some 20 years old now back in college where we'd blow up a capacitor/inductor or two :-) To the person who was interested in 10 port USB hub, this one works good (and is cheap) http://www.amazon.co.uk/TeckNet%C2%AE-Power-Adapter-Cable-Chipset/dp/B00DCO9YVM/ref=sr_1_5?ie=UTF8&qid=1375821594&sr=8-5&keywords=10+port+usb+hubDon't know if you can get it in the US, but 21 pounds works out about 30 USD - but shipping from England would be heavy I would think.
|
|
|
|
Trongersoll
|
|
August 06, 2013, 09:39:59 PM |
|
Kramble, that's an impressive piece of kit. I haven't read the entire ~150 pages , but that's about one of the coolest pieces of kit I've seen on this thread. It's not about the MH/s, its about getting stuff to work. Hat's off to you! Way beyond my "Electrotech 101" which is some 20 years old now back in college where we'd blow up a capacitor/inductor or two :-) To the person who was interested in 10 port USB hub, this one works good (and is cheap) http://www.amazon.co.uk/TeckNet%C2%AE-Power-Adapter-Cable-Chipset/dp/B00DCO9YVM/ref=sr_1_5?ie=UTF8&qid=1375821594&sr=8-5&keywords=10+port+usb+hubDon't know if you can get it in the US, but 21 pounds works out about 30 USD - but shipping from England would be heavy I would think. that looks like the anker with a different name and half the price.
|
|
|
|
btcee
Newbie
Offline
Activity: 24
Merit: 0
|
|
August 07, 2013, 12:15:49 AM |
|
Don't know if you can get it in the US, but 21 pounds works out about 30 USD - but shipping from England would be heavy I would think.
When trying to order from the US: We're sorry. This item can't be shipped to your selected destination.
|
|
|
|
Philj6970
Newbie
Offline
Activity: 29
Merit: 0
|
|
August 07, 2013, 05:26:13 AM |
|
@Kramble: Only half wave bridge rectifier? Another two diodes would go nicely Amazing work, wish I had the knowledge for that.
|
|
|
|
mazedk
|
|
August 07, 2013, 06:52:52 AM |
|
Don't know if you can get it in the US, but 21 pounds works out about 30 USD - but shipping from England would be heavy I would think.
When trying to order from the US: We're sorry. This item can't be shipped to your selected destination. Same when trying to get it to DK
|
|
|
|
kramble
|
|
August 07, 2013, 07:30:49 AM |
|
Only half wave bridge rectifier? Another two diodes would go nicely The diodes are to drop the 5V DC power brick output down to around 3V for input to the FPGA power supply (1.2 and 2.5V LDO regulators in the foreground), just to keep the regulator chip temperatures down (its also why the board is mounted vertically, improves air flow). Since the contraption is currently in my bedroom, I'm using passive cooling as the fans were keeping me awake. Most of the hi-tech is inside the FPGA chips, the breadboard is really just used as a patch panel. The only active components are a 74HC244 buffer for the byteblaster-clone that programs the FPGA's (driven bit-bang style from the raspberry pi gpio's), a couple of opto-isolators for the serial comms (again to the raspi, but passing through a DE0-Nano for obscure historical reasons), a 20MHz oscillator module (attached dead-bug style since its a SMT part) and a smattering of transistors/leds for monitoring. I always enjoyed the Heath Robinson style cartoons as a kid, seems to have rubbed off on me somewhat
|
|
|
|
rupy
|
|
August 08, 2013, 08:07:04 AM |
|
@Xian01 my guess is those powerbricks wont last 2 months like that...
|
BANKBOOK GWT Wallet & no-FIAT Billing API
|
|
|
Xian01
Legendary
Offline
Activity: 1652
Merit: 1067
Christian Antkow
|
|
August 08, 2013, 02:07:28 PM |
|
@Xian01 my guess is those powerbricks wont last 2 months like that...
Why do you say that ? Decent air-flow and temperature in that room, and the bricks are not hot to the touch at all.
|
|
|
|
Lyddite
Member
Offline
Activity: 98
Merit: 10
|
|
August 08, 2013, 07:40:34 PM |
|
Its a bit embarassing (like my electronics-fu is now totally blown), but seeing as the chip on the left just found a block (and at a ridiculous 1 billion difficulty), its just so much of an unlikely event that I just had to share. BTW that's an EP4CE10E22C8N mining at 12 MHash/sec !!
Cool, I have a 3 DE0-Nanos and was running them at 6.67 MHash a sec, powered by USB. I thought it would be cool to try to build something with the largest non-bga FPGAs from Altera but I guess it would be far from profitable at this point.
|
- Lyddite -
|
|
|
kramble
|
|
August 08, 2013, 09:04:17 PM |
|
Cool, I have a 3 DE0-Nanos and was running them at 6.67 MHash a sec, powered by USB. I thought it would be cool to try to build something with the largest non-bga FPGAs from Altera but I guess it would be far from profitable at this point.
I've got mine up to 35Mhash/sec, albeit with a custom power supply hack (bypassing the onboard regulators and directly supplying 1.2Volts at around 2 amps). Even if you keep the USB connection for comms you can supply 5V externally at around an amp or so (if you blow a lot of air over the regulators) and push it up to around 15-20MHash/sec. BGA packaging is a huge problem for the hobbyist. I managed OK with the 144 pin TQFP's (using a commercial breakout board), but it really limits you to the very smallest devices. I completely balked at BGA (I picked up a Xilinx LX9 in BGA really cheaply, just to see what it looked like, but I haven't even tried to mount it on a PCB, its just not going to work with what I can produce at home). I'm currently waiting for bitcoin difficulty to explode sufficiently to pick up some LX150 boards cheaply to play with for my Litecoin project, though the ISE license will be a problem (only 30 days evaluation, and a full license costs an arm and a leg). If anyone's got a Ztex LX75 board going cheap I'd be interested to hear from you
|
|
|
|
dani
|
|
August 08, 2013, 09:07:32 PM Last edit: August 08, 2013, 10:09:31 PM by dani |
|
to get back to that sweet idea, I build a small setup with 4 fans, sadly different in power, speed etc here you can see the fans the upper left is really powerful, alot more than the others built this fired it up what can I say: it works, but the setup is not good for comparison. The entry is 24x24 cm² and the exit is 8x8cm². A1*c1 = A2*c2 -> c2 = A1/A2*c1 = 9*c1. So one would expect the air to flow 9 times as fast as it does at the entry. It doesn't, its actually kinda lame. The powerful fan just pushed the air right out of the weakest of my fans (yeah, who would have thought that? ). I think when letting the fans suck the air out I would not get that effect of air moving the wrong direction, or maybe not that bad. Maybe I should try this with just one fan, doing the same in a smaller scale or get 4 equal fans. Any thoughts? Edit: The fans are pushing the air into the narrowing tunnel, so the output is the small end. I didnt make the middle (constant) section nor the output (widening) part, I thought this would not make a difference or have a positive effect. I might be wrong
|
Hai
|
|
|
Magazine
Member
Offline
Activity: 112
Merit: 10
|
|
August 08, 2013, 09:14:40 PM |
|
I love going back and looking at all the awesome GPU operations people had going in the early days. They may have been big but they looked so damn awesome.
|
|
|
|
DennisD7
|
|
August 08, 2013, 09:48:20 PM |
|
Update!I recently ordered some new USB Erupters, so I got me another 10-port Anker hub and taped them together with some sturdy industry grade double adhesive tape. I also moved up the power bricks so they get some of the side winds from the fan. I am expecting another 6 erupters so I'll have a total of 18. Previous post about this rig: https://bitcointalk.org/index.php?topic=7216.msg2801722#msg2801722
|
Proud ASIC miner and long term Bitcoin investor
|
|
|
Jay_Pal
Legendary
Offline
Activity: 1493
Merit: 1003
|
|
August 08, 2013, 09:58:49 PM |
|
to get back to that sweet idea, I build a small setup with 4 fans, sadly different in power, speed etc here you can see the fans https://i.imgur.com/Qji5V1p.jpgthe upper left is really powerful, alot more than the others built this https://i.imgur.com/jmr1ZcO.jpgfired it up https://i.imgur.com/jTWBUcy.jpgwhat can I say: it works, but the setup is not good for comparison. The entry is 24x24 cm² and the exit is 8x8cm². A1*c1 = A2*c2 -> c2 = A1/A2*c1 = 9*c1. So one would expect the air to flow 9 times as fast as it does at the entry. It doesn't, its actually kinda lame. The powerful fan just pushed the air right out of the weakest of my fans (yeah, who would have thought that? ). I think when letting the fans suck the air out I would not get that effect of air moving the wrong direction, or maybe not that bad. Maybe I should try this with just one fan, doing the same in a smaller scale or get 4 equal fans. Any thoughts? Yay!!! You made it!!! I'll PM you, but shouldn't the output be larger than the input (the other way around?)? And where is it pointing to?
|
|
|
|
dani
|
|
August 08, 2013, 10:11:25 PM |
|
to get back to that sweet idea, I build a small setup with 4 fans, sadly different in power, speed etc here you can see the fans https://i.imgur.com/Qji5V1p.jpgthe upper left is really powerful, alot more than the others built this https://i.imgur.com/jmr1ZcO.jpgfired it up https://i.imgur.com/jTWBUcy.jpgwhat can I say: it works, but the setup is not good for comparison. The entry is 24x24 cm² and the exit is 8x8cm². A1*c1 = A2*c2 -> c2 = A1/A2*c1 = 9*c1. So one would expect the air to flow 9 times as fast as it does at the entry. It doesn't, its actually kinda lame. The powerful fan just pushed the air right out of the weakest of my fans (yeah, who would have thought that? ). I think when letting the fans suck the air out I would not get that effect of air moving the wrong direction, or maybe not that bad. Maybe I should try this with just one fan, doing the same in a smaller scale or get 4 equal fans. Any thoughts? Yay!!! You made it!!! I'll PM you, but shouldn't the output be larger than the input (the other way around?)? And where is it pointing to? finally a prototype The fans are pointing in the direction of the tunnel (pushing air in), fixed my post
|
Hai
|
|
|
Trillium
|
|
August 08, 2013, 11:55:05 PM |
|
Can I play the big fan game too? It needs to run off a car battery jump-starter (600 A peak current) because none of my bench PSU could handle it. It needs about 200 watts or more. Even my small desktop PC ATX PSUs were no match. I made a bag thing out of a painters dropsheet to show how much air it moves. http://www.youtube.com/watch?v=ihvaFVsB4G8
|
BTC:1AaaAAAAaAAE2L1PXM1x9VDNqvcrfa9He6
|
|
|
Cablez
Legendary
Offline
Activity: 1400
Merit: 1000
I owe my soul to the Bitcoin code...
|
|
August 09, 2013, 12:50:59 AM |
|
What is your plan for that monster?!?
|
Tired of substandard power distribution in your ASIC setup??? Chris' Custom Cablez will get you sorted out right! No job too hard so PM me for a quote Check my products or ask a question here: https://bitcointalk.org/index.php?topic=74397.0
|
|
|
Xian01
Legendary
Offline
Activity: 1652
Merit: 1067
Christian Antkow
|
|
August 09, 2013, 12:53:33 AM |
|
Haha ! That's AWESOME. Thanks for posting the video.
|
|
|
|
2GOOD
|
|
August 09, 2013, 12:54:36 AM |
|
lolololololololol freakin' awesome
|
|
|
|
|