Quote from: Jemison on March 25, 2014, 11:07:16 PM

Quote from: wolfey2014 on March 25, 2014, 10:28:14 PM

Quote from: Jemison on March 25, 2014, 09:11:15 PM

Quote from: wolfey2014 on March 25, 2014, 02:23:54 PM

"Thanks for the photos. How safely can we mix and match these mods? Right now I don't have an iron tip fine enough to do the soldering mods. I either need to modify something or order something. I am interested in the pencil mod and the fan mod tonight though. Would this be safe? "

No, I don't believe the pencil mod will do any good without the other 3 mods.
Pencil mod is to adjust PLL voltage which needs to be corrected due to the offset caused by the two solder jumpers and the resistor is a timing adjustment to smooth out clock transitions at higher frequencies I believe.

You can make the 5V fan mod any time but remember, it's recommended for Scrypt only operation.
It has not been tested in SHA or SHA / Scrypt mining mode!

As usual, do these mods at your own risk!

Wolfey2014

Wolfey,

Somewhere in this thread is mention of a cgminer version that allows you to set the voltage. I realize you're using cpuminer, but do you think all 4 mods are required if we used the voltage mod cgminer? Value your opinion.

Thank you!

You ask a GOOD question.

Well, from what I understand - the current mods bypass having to soft set the PLL and Core Voltages.
Personally, I can't say one way or the other if soft adjustment is beneficial or not. I have heard 'rumored' that soft fine adjustment of 1000MHz up to 1048MHz stable is possible.

I'm going to have to convert over to one of the other programs to be able to read client side hash rates.
I just don't like all the tweaking that is still going on over the course of cg and bfg during it's 'use as it evolves' phase. I am not interested, at this stage of their evolution, in becoming an unwilling guinea pig Wink

I'm happy with cpuminer and what the pool's report in spite of their second by second inaccuracy. A bit of an understanding of odds, how they are avaraged out and common sense helps me figure out what the 'not so rough' improvement figure is for a given pod's clean hash rate.

Some sites like litecoinpool give a 24 hour average speed and that's close enough for me.
Besides and 'primarily' it only really counts what performance stats one sees pool side as that directly correlates to how much moolah it's gonna make you Wink

- the end product, the result, the win! $$$$ Wink

I guess I'm gonna have to learn one or the other 'unless there is a better alternative' because I could give a hoot what else the program does. All I want it to do is show me an accurate local hashrate as a separate program from the rest of it. And it MUST not be memory hungry like cpuminer 'isn't!' Thank you very much GridSeed! Wink

It would be cool if not great if GC would take back up updating / tweaking their very good and very stable program! It lives! Grin

IMO, cpuminer is very optimized and user friendly compared to the others IMO.
If the others were less complex and learning curve intensive, I'd of bothered to test them too by now.
Lots of other people like/love them though. So, to each their own.

Peace!
Wolfey2014

I created my own local database and query the pool api hourly to get their reported hashrate for my miners.
On average, over 48hrs, the pool is reporting 20% more shares than cgminer, and I'm okay with that, lol.

I use cgminer's api to query the health of my miners and to re-start them if necessary. cgminer has the ability to run all or a selection.
Its api can be set on different local network ports so if you have multiple instances running on one box you can query them separately.

I've only had one hiccup with cgminer in the 7 days I've been running - but I think that was caused by my linux attempting to self-update,
which I've now disabled.

Thanks, that's good data to know.
I'm considering setting up a separate laptop to run Linux Ubuntu.
I'll have to learn it from scratch since the last low level language I messed with years ago was Assembly to program a PIC Wink

Wolfey2014