DIY PCB with AVALON: "The Quarter Stick" - Needs Help!

[daemondazz]

It shouldn't be too difficult for someone with SMT soldering experience. The chips are QFN, not BGA, which is a positive thing for homebrew DIY.

[BkkCoins]

Thank you Bkk, man this design sounds really great. I will keep an eye on your github.

This being a sort of DIY project, how difficult would it be for an average Joe with a small amount of chips and some soldering experience to put together? Too risky?

Not too hard for those inclined to technical work and good with building stuff.
or,
Very, very hard for those clueless about how to put things together.

To get some idea do some googling for tutorials on "toaster oven reflow", and "skillet reflow".

Here's two topics to give you some idea, but there is a lot written about it online.

http://www.seattlerobotics.org/encoder/200006/oven_art.htm
https://www.sparkfun.com/tutorials/59

(Sparkfun has many other tutorials - see the one on stencilling as well)

The hard part is what to do when something doesn't work. Cross your fingers and hope it works first time because if not then you have some debugging work ahead requiring deeper knowledge and/or test equipment.

[Signus]

It shouldn't be too difficult for someone with SMT soldering experience. The chips are QFN, not BGA, which is a positive thing for homebrew DIY.

It's not too difficult, just a hassle. If you haven't done SMT before, get a breakout board for some QFN chip like the MSP430 and practice soldering and desoldering the chip as well as the capacitors and resistors.

A hot air rework station is your best friend. I currently only have my hot air gun (which burns a little hot for rework if I'm not careful), but I'm going to order a new station soon hopefully.

Also get some RC digital tweezers, these will save your life if you don't keep your SMT resistors and capacitors separate from eachother. Plus they all look the same, so these give you a way of finding their value easily.

[ektwr]

ASIC's are not FPGA's so it's only for mining and nothing else. Is it possible someone to post a block diagram with the philosoply of the system?

[atcsecure]

Are there any plans to make a larger miner PCB design? I have about 480 chips on the way, but no knowledge of EE. I'm hoping for something I can take to a local plant for printing and assembly, but the aim as stated in the amended OP would result in too small of a board. I don't have 480 USB ports available...  

The design I'm working on is 16 chips on a 10cm x 8cm board. Each board has a PIC USB controller but also supports I2C. This allows chaining multiple boards together so they use only one USB port. I'm targeting a 2 layer board to keep costs very low. If I run into problems with 2 layers then I'll bump up to 4 layer but that increases pcb cost by 3-4x. Right now parts cost estimate for one board is about $20 including pcb, not including ASIC, heat sink, assembly or parts shpg. ($20/16 chips is $1.25 per chip.)

I'll post my design on github eventually after final specs come out. You could panelize 8 boards on a larger 20cm x 32cm pcb for production. Then you would have 4 BIG boards total. Or whatever combination works. On each panel you would tie the sections together with I2C bus (4-wire ribbon cable), have one USB port and make a power splitter lead to distribute the 12V. 8 units using 32W each is 256W per BIG board. You could run two boards off a good 650W supply. Smaller boards give a bit more flexibility in arranging them for cooling.

I'll post more as I progress or check here:  http://github/bkkcoins
Don't expect anything until after Avalon docs are released.

sounds good - we should organize a large PCB group buy to get those costs down

[FullFathom5]

Been following this thread with great interest.

Am a novice but can devote time and would like to pool resources.

Anyone in this group in CA, SF/SJ greater bay area willing to work together?

EDIT: If so PM me here and we can talk details.

[bassclef]

Not too hard for those inclined to technical work and good with building stuff.
or,
Very, very hard for those clueless about how to put things together.

To get some idea do some googling for tutorials on "toaster oven reflow", and "skillet reflow".

Here's two topics to give you some idea, but there is a lot written about it online.

http://www.seattlerobotics.org/encoder/200006/oven_art.htm
https://www.sparkfun.com/tutorials/59

Thanks for the answer, and links. So soldering the ASICs to the board could be accomplished in a toaster oven, or a skillet. Nice. Not for mass production, obviously, but a beginner with a handful of chips and some soldering equipment could potentially do this.

Any idea of cooling requirements for these? The units BFL released into the wild seem to run very hot with large heatsinks. At the risk of getting ahead of myself, once we know the heatsink size needed, custom orders could be placed with a company like, say Heatsinks USA. I have dealt with them in the past when building LED lighting and they are inexpensive and cut custom widths.

[Raymgiv]

Any idea of cooling requirements for these? The units BFL released into the wild seem to run very hot with large heatsinks. At the risk of getting ahead of myself, once we know the heatsink size needed, custom orders could be placed with a company like, say Heatsinks USA. I have dealt with them in the past when building LED lighting and they are inexpensive and cut custom widths.

The BFL units use a different chip, its not related to the discussion(DIY PCB with AVALON).  The Avalon chip specs you should refer to are here: https://en.bitcoin.it/wiki/Avalon#Chip_Specification

Avalon made a note about heat sinks. "there is a air gap between the die and package top, install a heatsink on chip is useless. and will cause overheating. because the top PCB copper act as a heatsink too. do not cover them."

If the statement applies to this proposed configuration is yet to be determined. I hope burnin is looking into it.

-R

[rassalas]

Hi all,
I'm interested in this project.   What about KiCad?   It's not what I'm familiar with. Nonetheless, if we want to make an open source board, shouldn't we agree on an Open Source development suite?  My rudimentary investigation into this aspect of the project brought me to KiCad.   I downloaded it.  Does anyone else on this subject have sentiments for or against any particular software?  Try to think Open Source so all could collaborate or anyone could take their own fork if necessary.. 
  I'm seriously considering ordering some of the Avalon Chip, even to breadboard with, but would like to prepare ahead and get as many component footprints and expected nets set up ahead of time as possible... not to mention part ordering for various connectors and other components.  Summary,  PCB Schematic, footprints, Parts selection, Netlist, Routing, preferably all open source.
 

[rassalas]

Not too hard for those inclined to technical work and good with building stuff.
or,
Very, very hard for those clueless about how to put things together.

To get some idea do some googling for tutorials on "toaster oven reflow", and "skillet reflow".

Here's two topics to give you some idea, but there is a lot written about it online.

http://www.seattlerobotics.org/encoder/200006/oven_art.htm
https://www.sparkfun.com/tutorials/59

Thanks for the answer, and links. So soldering the ASICs to the board could be accomplished in a toaster oven, or a skillet. Nice. Not for mass production, obviously, but a beginner with a handful of chips and some soldering equipment could potentially do this.

Any idea of cooling requirements for these? The units BFL released into the wild seem to run very hot with large heatsinks. At the risk of getting ahead of myself, once we know the heatsink size needed, custom orders could be placed with a company like, say Heatsinks USA. I have dealt with them in the past when building LED lighting and they are inexpensive and cut custom widths.

I'm thinking stencils, paste and a reflow oven.  I am open to collaborations and/or contract work.

[ektwr]

KiCAD seems fine to me. But i disagree about the "oven solution". Its not appropriate for IC's because the heat is not diffusing properly at the surface of PCB.
I prefer the clean and more secure method of hot air station with liquid solder paste.

[ragingazn628]

Hey there,

I've been organizing group buys for Avalon Asics. Let me know how I can be of service! I am also forming a team locally in my state to pursue the Asics project. Let me know if I can help somehow

[eghoff]

KiCAD seems fine to me. But i disagree about the "oven solution". Its not appropriate for IC's because the heat is not diffusing properly at the surface of PCB.
I prefer the clean and more secure method of hot air station with liquid solder paste.

What about using a hotplate?

[Signus]

KiCAD seems fine to me. But i disagree about the "oven solution". Its not appropriate for IC's because the heat is not diffusing properly at the surface of PCB.
I prefer the clean and more secure method of hot air station with liquid solder paste.

What about using a hotplate?

You can't set a PCB on top of a hotplate, it might burn the PCB itself.

It has too wide of an area to be efficient I imagine, but if you find a way to give it a shot, show us.

[allten]

OK, I have a preliminary Schematic for engineering discussion.
https://www.dropbox.com/s/9c51uivncr5hhnp/DIY_AVALON_Rough_Draft1.sch.pdf

https://www.dropbox.com/s/ypqbam0uxwtcav9/BOM_Rough_Draft1.xls

Please take the engineering with a grain of salt. I would love to have some EE's look it over and
offer suggestions for improvement. I'm still having difficulty solving the math to yield values for
the following components:
C2?
C3?
R4?
C6?
c7?
L2?

Any help would be appreciated.

thanks,
     Allten

[Dabs]

Anyone in Asia? I'm in the Philippines. Also organizing a group buy for ASIC chips. So we've got 2 in Europe, 1 in the United States, and 1 in Asia. Decentralization.

[Bicknellski]

Anyone in Asia? I'm in the Philippines. Also organizing a group buy for ASIC chips. So we've got 2 in Europe, 1 in the United States, and 1 in Asia. Decentralization.

Replied to your PM Dabs... chat with me on Gtalk. Sure some people in Indonesia welcome sharing resources some how.

[ektwr]

We can find the values empiracaly from other designs, no problem. What the led R_led G_led are used for anyway?

EDIT:http://www.ti.com/lit/ds/symlink/lm2742.pdf
Look this-->Cc1 = 4.7pF 10%, Cc2 = 1nF 10%, Rc = 229kΩ 1%. These values give a phase margin of 63° and a bandwidth
of 29.3kHz
(from datasheet example)
C2=Cc1 C3=Cc2 R4=Rc

OK, I have a preliminary Schematic for engineering discussion.
https://www.dropbox.com/s/9c51uivncr5hhnp/DIY_AVALON_Rough_Draft1.sch.pdf

https://www.dropbox.com/s/ypqbam0uxwtcav9/BOM_Rough_Draft1.xls

Please take the engineering with a grain of salt. I would love to have some EE's look it over and
offer suggestions for improvement. I'm still having difficulty solving the math to yield values for
the following components:
C2?
C3?
R4?
C6?
c7?
L2?

Any help would be appreciated.

thanks,
     Allten

[roybitcoin]

OK, I have a preliminary Schematic for engineering discussion.
https://www.dropbox.com/s/9c51uivncr5hhnp/DIY_AVALON_Rough_Draft1.sch.pdf

https://www.dropbox.com/s/ypqbam0uxwtcav9/BOM_Rough_Draft1.xls

Please take the engineering with a grain of salt. I would love to have some EE's look it over and
offer suggestions for improvement. I'm still having difficulty solving the math to yield values for
the following components:
C2?
C3?
R4?
C6?
c7?
L2?

Any help would be appreciated.

thanks,
     Allten

Hi,

For this project I would prefer a monolithic regulator. You could considere this: http://www.ti.com/product/lm3678. This one have an Enable input (EN) for control and Soft Start for the inrush requirement.

On the other hand, this is for one ASIC solution. I´m considering the 10 ASICs design and I´m looking to use the PIC32 too (I have used only the PIC18 before and I´m thinking about that).

For 10 ASICs powering I see this modules as a good option: http://www.murata-ps.com/en/products/dc-dc-converters/single-output-non-isolated/oky2-t-16-d12p-c.html

Regards

[samurai1200]

For an application like this, a monolithic PSU is the way to go. The one roybitcoin suggested looks good, is only $1.72 in 1-piece prices, but only puts out 1.5A and is a package that cannot be soldered by hand (would need a reflow method).

I will dive into your schematics later today and see if I can come up with some values for your PSU, and do some digging for a suitable monolithic equivalent.