Topic: Bounty[PAID OUT] : a bitstream for better utilizing the Cairnsmore1 157-294.5btc - page 6. (Read 21997 times)

apologies for the mash up...

2 boards flashed and now undergoing 48hr stability test...thnx ppl...it has been about 5 years since I touched any Linux flavoured OS hence some rustiness

@Cranky4u Please try to avoid quoting whole posts and/or images... this really clutters the thread.

No problems, I was just to make it clear one part is not one big command line and is 4 seperate commands where you wait for it to flash.
Also that is merely a photoshopped image of one on their site, modified for the right settings.

a work of art...THANKS!

I quickly did an image up for you.
Follow these instructions and you should flash okay.
All the steps are important, worst comes to the worst try do them again from the start.

Unplug the usb cable on all other boards if you have more than 1.
Set one board to the above setting (SW1 - #3 enables programming btw)
Plug it again (this should make sure the program detects the board as ready)

At the command line type:
xc3sprog -c cm1 -j
This should show the board with it's 4 chips are visible.

At the command line type:
xc3sprog -c cm1 -p0 -Ixc6lx150.bit shortfin190.bit
then
xc3sprog -c cm1 -p1 -Ixc6lx150.bit shortfin190.bit
then
xc3sprog -c cm1 -p2 -Ixc6lx150.bit shortfin190.bit
then
xc3sprog -c cm1 -p3 -Ixc6lx150.bit shortfin190.bit

This flashes each of the 4 chips individually. It takes a few minutes for each one, be patient.
(make sure you update it for different bitstreams and changes in filename, ensuring obviously the bitstreams are in the same directory as xc3sprog)

Unplug the cable on the board.
Move SW1 switch 3 to ON (labeled mine), so it's like all the others next to it.
Move SW1 switch 1 to OFF (labeled reset) then after a few seconds to ON again
When just the yellow leds next to the 4 chips are on all, plug the board again

Cranky,

they're the same of the first image in Dip Switch Settings section on this page

http://www.enterpoint.co.uk/cairnsmore/cairnsmore1_support_materials.html

BUT for SW6 switch 1 which has to be on the OFF position (to the right looking at the image).

spiccioli

flashed successfully and tried mining with cgminer 2.6.1 but not working...test fails on startup

can someone post a picture of the dip switch settings for 190MHz settings so I can make sure they are right

Put me in for 2...

Work looks good so far.

Good one, didn't notice that one.

The command line works the same in linux and windows I believe, so I think you may have a small error in how you are using yours.
I notice you have a space between "-p" and the "0", also it's "-I" after that. Then the name of the first file and the second file, there is no space between -I and the first filename.
While that does not match the error completely it would be good to correct. Also make sure the bitstreams using the names you use are in the same directory as the xcsprog file, so it can find it. Renaming files is fine, as long as it is accurately named when you use it in the program.

I use:

xc3sprog -c cm1 -p0 -Ixc6lx150.bit shortfin190.bit
xc3sprog -c cm1 -p1 -Ixc6lx150.bit shortfin190.bit
xc3sprog -c cm1 -p2 -Ixc6lx150.bit shortfin190.bit
xc3sprog -c cm1 -p3 -Ixc6lx150.bit shortfin190.bit

(I renamed the shortfin files to be abit shorter to type out)
That flashes all 4 of the Chips. That finishes the flash then you just need to get it back into a mining state.


The below are two notes I copied from the main CM1 thread that I use to remind myself how to flash the Makomk bitstreams.
They explain essentially the same thing two different ways.

I did, I'm using Cgminer 2.5 (tried 2.61 aswell, no improvement), using Makomk's shortfin190 bitstream.

I am having some technical difficulties flashing my cm1 boards (62-0432 & 62-0433). I can follow the guides and bitcointalk.org hint to get down to step 4 in the manual - "Programming the bitstream". I then put in the command "xcsprog -c cm1 -p 0 iIxc6lx150.bit *.bit" but get an error of;
Can't open datafile xc6lx150.bit: No such file or directory
JDEC: ff ff 0xff 0xff
unkown JDEC manufacturer:ff
ISF Bitfile probably not loaded
root@cairnsmore:~#

I am using a Win 7 64 bit PC...any help on solving?

are you able to list bitstream and miner app used to achieve these figures?

I have trialled a few other bitstreams of Makomk. All stable, and now settled back on the 190 (is their a 200?)
It does now appear to be performing a bit better, maybe all that re-flashing did it some good, U rate more like 11-12 per board, rather than 10 or lower.
Guess even though it verified as success, re-flashing it sometimes is needed.
The U rates are not all uniform yet, but it's an improvement over last time.

Either way, this according to Cgminer translates to a 750Mhash/s average, which does qualify for the bounty.
One of my boards certainly does mine better than the other, Need to find out if their is something wrong with the board, thus needs me to send it back to enterpoint or if further improvements in the bitstream can solve it.

If I remember rightly my boards are numbered 62-0418 and 62-0419, so I got one of the later boards.

Well done Makomk.

Donating 12.5 BTC. 25 if it's released before 31/08.

Did you even read the second line of what I said?


I'm happy with his work, I've made that clear when I've mentioned in this thread and Enterpoint's. It's better than what I had before, so I'm not complaining.
I have struggled to get my cm1's to cooperate with me from time to time, but not due his bitstream, it was software suggested by enterpoint (which is not necessarily their fault) and me having to ask the community the best way around it, which I tipped nicely for their help.

I don't have a grudge against Makomk, I of course expected glasswalker to win, as a early entry and had an advantage with his connection to enterpoint but I made it clear I didn't care who done it. But he appears to came close then ran into problems, bad luck I guess, it allowed others a chance at it. Makomk came in as the 2nd major entrant and provided an great bitstream that works all 4 chips just fine and has improved it regularly since to this point where I decided to try it.

But make no mistake, my numbers are not wrong or mixed just because they aren't the same as yours. It's feedback to show their might be a little variance in results, it's known that not all cm1 boards react the same, maybe mine is one of those. It's stable so I'm happy, but "no cigar" on it doing the results for the bounty yet for everyone. I have no doubt Makomk can pull that off in the next few days. Apparently their is a 200 version floating around, but I've not tried it yet.

Btw since I updated to 2.6.1, while it's early days, my average U went down from 10 to 9.5 per board. So I might be switching back, to 2.5 unless their is a tweak to be done that provides better results.

I for sure have no problem with the bounty, but when people post wrong/mixed numbers that are not related to any rule as far as I know. Posts like that let it misinterpret that you wont to grudge Makomk as a potencial winner...

What is when your internet connection have problems, pool have problems, stales etc. You can't count what your pool says to the rule. The 760Mh rule was just taken from the standard hashrate of 2x icarus which is hashing at 380Mh. The hashrate of the unit is importand to the rule, not what the pool shows you. Thats all what i want to say.

eb

You are right, the expectation (rule) of 760Mh/s does not have to be related to my pool, maybe because it's not stated how it is to be measured. It's also not mention that I have to use a very specific piece or version of software either, however I will update from 2.5 to 2.61, to see if it makes any difference, out of curiosity sake.

With how you structured your reply, it could be seen as a hostile retort, why?
I'm providing feedback and since many other bitstreams so far on CM1 can be a little off when it comes to the reported numbers in the software you use to mine with (cgminer / MPBM), what my pool reports is usually a reliable method to determine an accurate number. My U average per board is 10 in cgminer, so it is a little under double what I use to get before, that is about in line with my stated numbers above. I'm sure others get better results, I am however I am not one of those yet.

You choose not to be part of the bounty, you've already paid him directly. Fair enough. I however did choose to take part in the bounty, so what is the problem?

The rule for the 760Mh is not related to your Pool! As my Pool shows sometimes 8.2Gh with my 10 boards, but MPBM shows me ~7.55 at the moment...

Just use MPBM or the new cgminer and what this soft shows you is the correct Mh your boards running at.

O_o

I'm not achieving 760Mh/s, maybe at most 675Mh/s on the shortfin190.
So while it appears stable, it to me does not yet count towards the bounty, it however is a great achievement and I'm sure he is very close.