Quote from: AniceInovation on June 21, 2013, 04:20:13 PM

Nice one burnin, you're nailing it.
Just 2 questions:
-Do you provide standoffs, or should i get them:https://bitcointalk.org/index.php?topic=234459.0
And if so, how high should they be?
-The CAN cable, do you provide, or must we build one? I searched for a DIY one but found no easy to do ones.

1. Standoffs are included .
2. The CAN cable can be ordered with the boards both in DIY-Kit form and completely assembled.

Quote from: iikun on June 22, 2013, 12:28:28 AM

Quote from: idee2013 on June 21, 2013, 08:20:45 PM

someting about higher clockrates
https://bitcointalk.org/index.php?topic=239713.0
https://bitcointalk.org/index.php?topic=239360.0

wow..stable at 350mh/s
If this speed can be achieved by Burnin that would be truly awesome

Well that does depend on both board and chips, but with overvolting those higher speeds might be more stable then in the originals.

Quote from: mike2kt on June 22, 2013, 01:46:43 AM

Quote from: Defkin on June 21, 2013, 10:35:36 PM

Quote from: AniceInovation on June 21, 2013, 04:20:13 PM

Nice one burnin, you're nailing it.
Just 2 questions:
-Do you provide standoffs, or should i get them:https://bitcointalk.org/index.php?topic=234459.0
And if so, how high should they be?
-The CAN cable, do you provide, or must we build one? I searched for a DIY one but found no easy to do ones.

The CAN cable is easy. Just ask for IDC Header 10 pin.

http://www.youtube.com/watch?v=q6t8s_7eA0w
http://www.youtube.com/watch?v=p1yZKT3Yock
http://www.ebay.com/bhp/10-pin-idc-connector
http://en.wikipedia.org/wiki/Insulation-displacement_connector

For short runs and/or slower speeds a flat ribbon cable may not pose issues on the bus. Having said that...

The CAN specification uses two wire twisted pair differential signaling. Since external interference tends to affect both wires together, and information is sent only by the difference between the wires, the technique improves resistance to electromagnetic noise compared with use of only one wire and an un-paired reference (ground).
Im not sure what Burnin has on the board in regards to can, but in addition to using a twisted pair, a terminating resister should be placed at each end of the daisy-chain of devices. Again, short runs may get away without issues. When you are using High-Speed CAN, a resister is placed at each end, bridging the wire pair. The resisters bleed off residual line voltage (aka reflections). The resister value is 120 ohms, which matches the cables nominal impedance of 120 ohms.

Low-Speed CAN is designed to have a resister at each node, which comes along with the hassle of calculating the values for the resisters.

/no I have never done extensive work with CAN busses in cars/
//go BMW, they switched over to Ethernet/IP, makes interfacing so much simpler//

10 Pin Ribbon cable is a quite common CAN-Cable when not used in automobiles or rough environments.
It has the right impedance, and i've included termination resistors on the board that need to be enabled on the Bus ends.

Quote from: Micky25 on June 22, 2013, 02:28:45 PM

Hi burnin,

do you have any plans on testing the bitfury chips? https://bitcointalk.org/index.php?topic=228677.msg2407907#msg2407907

Best regards,
Micky

Yes these would be absolutely perfect and i intend to use these for the next generation.
The BFL chips would require a lot more development because of the flip-chip package.

Quote from: dani on June 21, 2013, 04:11:58 PM

Quote from: burnin on June 21, 2013, 01:36:53 PM

Thinking about posting a picture of the test rig, but it is a truly horrible mess of wires.
I don't want any of you to think the end-product could look even remotely similar to that thing.

Do it. come one. Do it.

ok here we go:

Whole test setup:

Test Jig:

The Microcontroller:

Size reference for the chip: