Topic: [Work in progess] Burnins Avalon Chip to mining board service - page 84. (Read 624200 times)

Hi Burnin

Can you explain how you perform the overvolting currently? Is it something you can port to users in the next few days?

Thanks!

If cgminer already would support bitburner xx fully fine kano's request wouldnt be needed. But it doesn't look like. Burnin is using a workaround for overvolting. So the only way to ensure that cgminer really works fine is that a coder gets a miner. Everything else would be guessing how it works, eat way more time for coding and would have errors at the end. So i dont see another way.

Oh, really? Tell that to Canonical or Red Hat, they are still unaware 

giving something for free, and asking for something for free are entirely different things. is that difficult to understand?!

Out if this thread please

edit : not for this topic, sorry

Excellent outcome for all interested parties. Well done guys!! 

QG

 

Excellent!

  thanks

Indeed, the firmware update function of the boards is working.


1. I will see what i can do, we have full control over the usb endpoint in that micro, the FTDI chip is emulated
2. Each one will definitely get a unique serial-number programmed into them (Also used for CAN)
3. There is a command for that in the CAN protocol and thus it could be supported

Even though the problem is cleared now, burnin will provide boards, i wouldnt the suggested. People write me they were in hospital or moved and had no internet, were travelling and so. I dont want to tell them that their chips, they waited months for and hoped for a miner, are gone now. I think that would be the opposite of trust.

Sorry about that - yep I seem to ignite pretty easily when people tell me off about not doing stuff for free for them

This was the point why I made the request, it's no longer a standard Avalon

Thanks indeed - having the hardware does indeed mean we can do proper support and can ensure future support.
My usbutils has had a lot of little (but major) changes lately - I test changes across everything I have.

Yes the Avalon design is definitely not optimal ...

Not to bring up the CM1 argument again, but the CM1 is a good example here where someone wrote a different firmware with extra features and thus there is indeed at least two different versions of it ... the second one clearly better.
However, the first step is indeed to get something working and delivered, then later to improve it.

A few comments about the boards you send to others:
1) Please, please, please, put something in the USB iProduct and iManufacturer - the Avalons have that blank and it causes trouble due to the current Avalon code looking at any and all devices with the FTDI 0x6001 chip.
2) Would be good if they do have a serial number in USB iSerial - I'll be adding code to use that and display that (in the API) soon, and being different on each board is of course the best way to tell them apart (and possibly a blank sticker on the board for people to write their iSerial ?)
3) Maybe a future firmware update would add an identify command to say flash one of the leds ... directly related to 1) so people can tell which is which more easily (the API has this option already for hardware that supports an identify command)

no need for that, i will use chips from the "broken-chip and warranty reserves" and send Kano&ckovilas boards from the first batch.
(they'll need more then one so they can do CAN-bus testing too)

Again: Mining works, overclocking works, the only thing not in cgminer (yet) is the voltage adjustment.

I would like to donate one of my boards to Kano, so he can do his magic.
Burnin, is it possible to send one of my boards to Kano ?
If yes, can you contact Kano to ask for his address so you can make it happen ?
I will pay the shipment too.

If others would like to step in, just donate some mBTC to 1LYzxJU5xK675qQj4EMbTTSuJUJGkPwonY

EDIT: Taken care of by Burnin!

Still want to try a bare PCB. I wonder if that PCB is compatible with 10 chips? 

To get that right: cgminer does support overclocking, but no core voltage changes (yet).


Important: if any of you have questions or requests regarding your order: send me an email not a forum pn.

Great news burnin!
Let's hope the other batches won't be delayed, as the avalon batch2/3.

I'm in hope for a cgminer version that supports overclocking.

What about SebastianJu or Zefir, perhaps they might want to support the network in this way.

I hope I don't have to hide and run now... 

To clear up the cgminer matters:
1. i dont know what disputes you guys have over the cairnsmore support but please keep your hate out of this thread - thank you

2.
Initially I decided to emulate the Avalon protocol so there would be no forks or changes needed to be maintained.
And it works perfect with a 100% unmodified cgminer version. (without the voltage adjustment)

3.
Then, i decided to add the voltage adjustment feature which the original avalon protocol doesn't support.
Therefore changes will have to be made to get that into cgminer.
Currently I use a second application to do the Voltage adjustments.
I have yet to decide how to continue from here.

Nevertheless think it would be a good idea to provide kano with at least two boards so he can do debugging in the case problems arise.
Development/debugging without the actual hardware sucks.

In the long run I aim for a custom protocol that is more adapted to the BitBurner architecture, and allow individual per-board tweaking.


general project updates:
Chips received from Zefir, production will begin on Monday.
Shipping could start 5th of August (one week later).

Fair enough just excited

As I understand it from Zefir's thread the chips are still on their way to Burnin. 

I'm sure Burnin will announce the arrival of the first batch but I think it's a bit much to send every chip-holder a personal email. You might see a change in the status of your order but lets let him do his work for a few days, which will get the boards in our hands faster