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

[eghoff]

How difficult would this really be to solder?
I admit that i have never done any QFN soldering, but looking at this https://www.sparkfun.com/tutorials/59#Hot Plate Reflowing gives me some hope that i could be done at home, with relative ease(?)
Would there be any problems in using this technique?

[allten]

How difficult would this really be to solder?
I admit that i have never done any QFN soldering, but looking at this https://www.sparkfun.com/tutorials/59#Hot Plate Reflowing gives me some hope that i could be done at home, with relative ease(?)
Would there be any problems in using this technique?

I did this with the Bitsafe and the results were very good. I was pleasantly surprised with the method.
Although, I didn't have any QFN packages.

[eghoff]

How difficult would this really be to solder?
I admit that i have never done any QFN soldering, but looking at this https://www.sparkfun.com/tutorials/59#Hot Plate Reflowing gives me some hope that i could be done at home, with relative ease(?)
Would there be any problems in using this technique?

I did this with the Bitsafe and the results were very good. I was pleasantly surprised with the method.
Although, I didn't have any QFN packages.

Awesome to hear.
I will try to find some QFN packages in the near future and test this out.
From what i read, QFN packages should give no problems with this technique (bridges only forming on the outside, and not under), but i will try and test it first so i don't mess something up

[ButchHashidy]

How difficult would this really be to solder?
I admit that i have never done any QFN soldering, but looking at this https://www.sparkfun.com/tutorials/59#Hot Plate Reflowing gives me some hope that i could be done at home, with relative ease(?)
Would there be any problems in using this technique?

lol, i like the author's humor on this site.. thumbs up. good find

[allten]

Low Cost DIY ASIC Miner

Purpose:
This thread is intended for the development of a very low cost DIY open source
ASIC miner based on Avalon's chips. Hopefully, it will inspire other Bitcoin ASIC
manufactures to offer their chips as well.

Objectives:
* Get ASICs in the hands of the masses ASAP.

* Create the cheapest solution possible. The lowest possible price is the goal.
The $ per hash ratio is not expected to be the best.

* Make it simple enough that anyone with a DIY spirit can assemble one of these with
some simple low priced tools.
https://www.sparkfun.com/tutorials/category/2
DIY manufacturing may be an avenue to get these out much faster.
Expect a few hours assembly time per board.

*Make the PCB design that is not only DIY friendly, but mass production friendly as well.
It will use surface mount technology.

*Inspire more ASIC Bitcoin Mining manufactures to pop up all over the world (Decentralization).
ASIC manufacturing in general is the most complicated production process in the world.
IMO, the best thing for Bitcoin would be for ASIC Bitcoin Manufactures to be decoupled from
final product production; or at least offer their chips as well on top of their final products.

Specifications (subject to change):
* USB power source is all that is required.
* Utilizes a single Avalon ASIC
* 256+ MH/s
* 2 Watt maximum power usage
* TYPE A USB Male option on either side of the PCB
* Type B USB Female option
* More to come

Sponsors:
*Burnin - He is a serious manufacture of Bitcoin mining equipment. He has committed to helping
with the open source software/firmware to make this project function. He has started an official
thread for his miners and it's worth checking out:
https://bitcointalk.org/index.php?topic=179769.0

*Zefir - He is preordaining Avalon ASICs in smaller quantities for anyone interested:
funds: https://bitcointalk.org/index.php?topic=177827.0
Zefir is committed to helping this low cost DIY project be successful by making a small quantity
of chips available for ordering and development.

Team:
All of you! An open source design also means its open to anyone. Feel free to take the initiative
to help out in your own way. Please don't get discouraged as it may take time to figure
out where your talents fit in best. I have received an overwhelming quantity of supportive emails
and also many request to help out. I wasn't prepared for that, but now, I would like to make
a list of anyone that wants to get involved here and what they would like to bring to the table.
Shoot me ANOTHER email and I will add you name here.

"Someone42" who was the inspiration and pioneer in hardware wallets has offered his help here.
"Burnin" has also offered his help, but is also very busy as he is preparing his own miner.

Allten's Commitment:
I will engineer the schematic and the PCB. As soon as the chip specifications are available, I will
release within a week a preliminary design for feedback. I will personally manufacture enough
of these to get them in the hands of all those that will be helping with firmware/software development.
After that, a detailed tutorial will be made on how to order and produce these by yourself.

Important note: I do not want to become a distributor for these for the simple fact I would like
to have time available for other projects I'm passionate about. This is a golden opportunity for anyone
or a few people to become distributors of DIY kits and already assembled miners. I will assist anyone
to get set up for this once I've completed the tutorial.

Funding:
Sending some coins would be very helpful and greatly appreciated.
1AqEzSiw7aqZ7T53XvXMqrcnUD5tKcvJxP
What will donations cover? Enough to order the parts and PCBs for the initial boards used for development.
All extra will be used to drive bounties for software/firmware development.
I get the feeling that many have ordered chips with confidence that there will be a PCB solution
readily available before the ASICs arrive. If that is the case, consider donating up to 5% of
what you invested in chips to this project. It would sure help hurry it a long! Much of the bounty portion
may go to "burnin" as his code development for the controller portion will be shared with this project and
he is most heavily invested to make it work along with his product ASAP.



~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ OLD - For Reference ~~~~~~~~~~~~~~~~~

Update: April 15th 2013

A huge thank you to Avalon for making their chips available. This is a great step in the
right direction for Bitcoin! Hopefully ASICMiner, BFL, and future ASIC manufactures will
follow their lead someday.

Looks like we will still be waiting a little bit longer for some more detailed spec's, but
thanks to Burnin's help, he has shown me that the design is adequate so I'm going
forward with a BOM (Bill of Materials) and a preliminary schematic for everyone's
review.

Here's the BOM:
https://www.dropbox.com/s/45698vk7bpvbdkp/BOM.xls


Update: April 3rd 2013

we are not looking for investors, and unfortunately batch #3 is sold out at the moment, this means you can do a few things

1. wait till when we finish installing our SMT line and see our production capacity and decide if we were to do another batch or not. ( late April )
2. wait for us to finish our documentation and start selling the Avalon ASIC chips which you can built your own mining machines with. ( mid April )

Currently we don't want any money.

Thanks for querying though,
Cheers.

Update: April 1st 2013

Project is at a stand still until more details are released regarding
AVALON ASICs. I was hoping to encourage BitSyncom to just
give us a few preliminary details in the forum so we could make
sure the hardware design was appropriate. Unfortunately,
That has not happened.

Instead of waiting until complete and official Spec's are available
online I would like to ask anyone with an AVALON that has skills
with an oscilloscope and ohm meter to please help with this.
Shouldn't take too long.

I wanted to offer a bounty, but then I remembered anyone with
an Avalon probably wouldn't be motivated for anything less than
a Bitcoin.

Here are the spec's available so far:
https://en.bitcoin.it/wiki/Avalon#Chip_Specification


What is needed?

1) Pin out for all 48 pins. What pins are ground? What Pins are power? Clock pins? Data pins? etc.

2) Do the ASICs have a frequency multiplier? What is the frequency on the core clock pin? Internal
    core clocked is spec'ed around 256 MHZ

3) What is the data bandwidth on the clock/data lines? Just need to know what the clock
    frequency is for the clock pins and that would be sufficient.

Bonus: are there any serial communication protocols being used: SPI, I2C, Proprietary, etc.?

Thanks!

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Building a DIY Open Source ASIC PCB based on AVALON ASICs

Time is of the essence [obviously] and hoping there will be
a lot of talent that can help make this happen with respect to Firmware
and software. So far its just me with one other who has expressed
interest in an email.

The design is kept as simple and cheap as possible and components were
selected that would be easy to solder. The only difficult soldering
task would be the ASIC themselves and would require oven reflow
(or a skillet). People could get the ASICs pre soldered by someone else
if they don't feel comfortable giving it a try. There will be a tutorial on how
to do a seemingly challenging ASIC soldering job with few tools for those willing to try.

Why do this?

"MiniMiner would be great, and an excellent way to introduce new people to Bitcoin. 
A low cost device with a nice, simple, user interface could really broaden the market. 
I think there's something about mining that really gets people hooked and interested
in learning more." - Anonymous

Also, it was a tragedy (even though inevitable and necessary) IMO when custom
hardware started to come forth. In the GPU and CPU phase where anyone could learn
about it and be mining at the end of the day was a unique period that hooked many
into Bitcoin. I would like to help lower entry to competitive mining with this project.

Why Avalon?
I was inspired by the article about Yifu.
http://motherboard.vice.com/blog/engineering-the-bitcoin-gold-rush-an-interview-with-yifu-guo-creator-of-the-first-asic-based-miner
They are the only ones I trust to be supportive of a project like this. Even with BFL
rolling out their hardware soon and a 28nm ASIC to hit the market later this year,
this is still essential and could pave the way for other ASIC manufactures to consider
releasing their ASIC interface designs and sell just the chips.

I'm much more of a hardware guy and have not been involved at all with any Bitcoin
mining hardware before so I do not know if the design is appropriate.
Please chime in and help me see what's missing and what needs changed.
A schematic will be available in a day or two depending on the feedback received.
I'm hoping the Avalon team will take a moment and give their feedback as well.

Here's the chip Specification that I have found so far:
https://en.bitcoin.it/wiki/Avalon#Chip_Specification

My two biggest design hurdles at the moment:
1)Frequency driver for the Avalon ASIC. What is the input frequency?
I assume it has a clock multiplier, but if not I will need to reconsider that portion of
the design.

2) Having no idea how work is passed to the chip, would 8 data lines with a frequency
of 48 MHz cause a bottleneck? If so, by how much?

Here's a diagram for inspiration. It will give an idea to everyone (technical and non-technical) of what I'm thinking:

[arklan]

ya know, i just realized something. for originally purely artistic purposes, i have the equipment and experience making etchings in salt water. i used it to make pretty brass plates. but, the technique i learned is from the how made PCB world. there by, all i need to get my hands on are blank pcb boards and i can etch them.

i can print out the design itself from a PC using a laser printer, then it's transferred as a mask via use of a heated iron. the mask protects the parts from etching, leaving the rest exposed. run a electrical current (i have a lab quality controlled watt/volt power supply for this) through it while submerged in saltwater and bingo. pcb's.

cheaper then going to a fab for small orders, eh?

if there's interest in me doing this, i'll get some blanks and start experimenting. been a while since i've done any of this.

[BkkCoins]

ya know, i just realized something. for originally purely artistic purposes, i have the equipment and experience making etchings in salt water. i used it to make pretty brass plates. but, the technique i learned is from the how made PCB world. there by, all i need to get my hands on are blank pcb boards and i can etch them.

i can print out the design itself from a PC using a laser printer, then it's transferred as a mask via use of a heated iron. the mask protects the parts from etching, leaving the rest exposed. run a electrical current (i have a lab quality controlled watt/volt power supply for this) through it while submerged in saltwater and bingo. pcb's.

cheaper then going to a fab for small orders, eh?

if there's interest in me doing this, i'll get some blanks and start experimenting. been a while since i've done any of this.

This isn't quite how it's done. There is no electrical current used, and it's not salt water. Usually ferric chloride or ammonium persulfate is used. Laser printer transfers are unlikely to be fine enough resolution for 0.5mm pitch SMD devices. You could try but you will end up with inferior boards. Not to mention that 4 layer boards are typical for this application (though I am going to attempt 2 layers myself to save money) and a solder mask is pretty much mandatory (which means screen printing on top), and then thru hole plating (which does use electrolysis) (especially for this board as it requires thermal vias). The cost of ordering small batch PCBs from China is very reasonable. 2 Layers about $1.25 per 8-10 chip board, 4 layer about 4x this.

Not meaning to put down your idea or efforts - just don't want others to get the idea this is a reasonable path.

For more info see, http://en.wikipedia.org/wiki/Printed_circuit_board

[arklan]

ya know, i just realized something. for originally purely artistic purposes, i have the equipment and experience making etchings in salt water. i used it to make pretty brass plates. but, the technique i learned is from the how made PCB world. there by, all i need to get my hands on are blank pcb boards and i can etch them.

i can print out the design itself from a PC using a laser printer, then it's transferred as a mask via use of a heated iron. the mask protects the parts from etching, leaving the rest exposed. run a electrical current (i have a lab quality controlled watt/volt power supply for this) through it while submerged in saltwater and bingo. pcb's.

cheaper then going to a fab for small orders, eh?

if there's interest in me doing this, i'll get some blanks and start experimenting. been a while since i've done any of this.

This isn't quite how it's done. There is no electrical current used, and it's not salt water. Usually ferric chloride or ammonium persulfate is used. Laser printer transfers are unlikely to be fine enough resolution for 0.5mm pitch SMD devices. You could try but you will end up with inferior boards. Not to mention that 4 layer boards are typical for this application (though I am going to attempt 2 layers myself to save money) and a solder mask is pretty much mandatory (which means screen printing on top), and then thru hole plating (which does use electrolysis) (especially for this board as it requires thermal vias). The cost of ordering small batch PCBs from China is very reasonable. 2 Layers about $1.25 per 8-10 chip board, 4 layer about 4x this.

Not meaning to put down your idea or efforts - just don't want others to get the idea this is a reasonable path.

For more info see, http://en.wikipedia.org/wiki/Printed_circuit_board

well, yes, acid is usually used, but the salt water etching works fine too. i'll agree the detail wouldn't be .5mm, and multiple layers... yea. points made.

[ashleysly]

I have bought PCB boards before and can get them at a steal (for the price) from China.

[CoinHoarder]

Will you accept chips from ragingazn628's group buy also? https://bitcointalk.org/index.php?topic=177994.0

I'm trying to figure out my options with the chips I'm going to order through him.

[stanke]

Hello I read you need some website. I'm offering a space on my server if needed.

[shibaji]

Will you accept chips from ragingazn628's group buy also? https://bitcointalk.org/index.php?topic=177994.0

I'm trying to figure out my options with the chips I'm going to order through him.

Be sure to consider the physical presence of zefir/allten/burnin vs ragingzan628 - this was the deciding factor for me, and I went with zefir. If I get someone assembling in US, I will consider another position in ragingazn's list. Dalkore mentioned briefly, but I did not see a follow-up.

[Bogart]

I'm interested in being a part of this effort.

I also want to make and populate my own boards, but I'm something of a noob when it comes to SMT.  I've done some SMT repairs with an iron, but I've never made or populated a board.  I'm looking at what gear to buy to be able to do this.

Would a cheap small reflow oven like this one be OK?

http://www.ebay.com/itm/400325274614

Reading about IR reflow on wikipedia, they say "With infrared reflow, the board designer must lay the board out so that short components don't fall into the shadows of tall components."

[BkkCoins]

I'm interested in being a part of this effort.

I also want to make and populate my own boards, but I'm something of a noob when it comes to SMT.  I've done some SMT repairs with an iron, but I've never made or populated a board.  I'm looking at what gear to buy to be able to do this.

Would a cheap small reflow oven like this one be OK?

http://www.ebay.com/itm/400325274614

Reading about IR reflow on wikipedia, they say "With infrared reflow, the board designer must lay the board out so that short components don't fall into the shadows of tall components."

I've looked at those on eBay and wondered if they're any good, and if there is any gain over using a converted toaster oven for small scale projects. No doubt if you want to be in business doing this a proper unit would be good. There are a few projects around for adding more accurate temperature control curves to toaster ovens, which for small boards would have enough heat capacity.

[eghoff]

Would a cheap small reflow oven like this one be OK?

http://www.ebay.com/itm/400325274614

This would probably be ok, but if you just want to make them for own use, and not populate a great amount of boards, you could probably get by using the hotplate technique i linked above. At the moment i'm not 100% sure that it would work flawlessly, but i'm going to test this. If it does, we would have no problem populating the boards by our self, without expensive equipment.

[BkkCoins]

Would a cheap small reflow oven like this one be OK?

http://www.ebay.com/itm/400325274614

This would probably be ok, but if you just want to make them for own use, and not populate a great amount of boards, you could probably get by using the hotplate technique i linked above. At the moment i'm not 100% sure that it would work flawlessly, but i'm going to test this. If it does, we would have no problem populating the boards by our self, without expensive equipment.

It would be very useful if whoever makes boards could supply stencils as well.

[Knecke]

I would be pleased if anyone could make detailed pictures of the avalon board with the asics, power stages, etc.

I would like to start developing as soon as possible.

Regards

Knecke

[RoadStress]

I would be pleased if anyone could make detailed pictures of the avalon board with the asics, power stages, etc.

I would like to start developing as soon as possible.

Regards

Knecke

https://en.bitcoin.it/wiki/Avalon

[BkkCoins]

I would be pleased if anyone could make detailed pictures of the avalon board with the asics, power stages, etc.

I would like to start developing as soon as possible.

Regards

Knecke

https://en.bitcoin.it/wiki/Avalon

These aren't adequate. I looked carefully at the board photo and you can identify some pins but some parts go out of focus and it's impossible to follow traces to figure out most pins. A direct overhead flat shot showing a larger area of the board would be ideal.

I don't think any Avalon owner is likely to start showing off photos as it's counter productive to them. Still it would be nice.

[SebastianJu]

I would be pleased if anyone could make detailed pictures of the avalon board with the asics, power stages, etc.

I would like to start developing as soon as possible.

Regards

Knecke

https://en.bitcoin.it/wiki/Avalon

These aren't adequate. I looked carefully at the board photo and you can identify some pins but some parts go out of focus and it's impossible to follow traces to figure out most pins. A direct overhead flat shot showing a larger area of the board would be ideal.

I don't think any Avalon owner is likely to start showing off photos as it's counter productive to them. Still it would be nice.

You could offer a Bitcoin or a half. Still enough to stop the machine for an hour. And the ones that want to make a business out of it could invest the BTC.