[GUIDE] GridSeed 5-Chip USB, Blade & Black Miner Support/Tuning

[sandor111]

So since any frequency is allowed now and not just 25Mhz steps, does autotune initially start with 25mhz steps and then fine tune it from that point down to something like 5mhz steps?

That would practically be impossible to manage, because for example 825 and 838 MHz are not equivalent, by that I mean that if 838 MHz is stable, there is no guarantee that 825 would also be stable. The commands are calculated from different PLL dividers and that can affect stability, by contrast if you use the same frequency stepping, it's almost 100% certain that a lower frequency will be stable if the current one is.

What?  You wouldn't like a challenge like that?  J/k

Makes sense.  But if they use different pll dividers would it be possible to create 2 autotune functions - each confined within its own pll divider?

That's kind of like the code I like to tinker with but unfortunately I don't know the language cpuminer is in and learning it takes more time than I have right now

That would mean autotune would take much much longer, because it would have to test all (or a selected set) of the frequencies with differing PLL dividers each time it stepped up or down.
Cpuminer is written in the most basic language available to man, C. Most of the modern languages today are written in C.

[jamieb81]

This is my result with latest Cpuminer on one Vmodded GS



Taken with Webpage Screenshot

[nst6563]

So since any frequency is allowed now and not just 25Mhz steps, does autotune initially start with 25mhz steps and then fine tune it from that point down to something like 5mhz steps?

That would practically be impossible to manage, because for example 825 and 838 MHz are not equivalent, by that I mean that if 838 MHz is stable, there is no guarantee that 825 would also be stable. The commands are calculated from different PLL dividers and that can affect stability, by contrast if you use the same frequency stepping, it's almost 100% certain that a lower frequency will be stable if the current one is.

What?  You wouldn't like a challenge like that?  J/k

Makes sense.  But if they use different pll dividers would it be possible to create 2 autotune functions - each confined within its own pll divider?

That's kind of like the code I like to tinker with but unfortunately I don't know the language cpuminer is in and learning it takes more time than I have right now

That would mean autotune would take much much longer, because it would have to test all (or a selected set) of the frequencies with differing PLL dividers each time it stepped up or down.
Cpuminer is written in the most basic language available to man, C. Most of the modern languages today are written in C.

Unfortunately the languages I'm familiar with are VB, VBScript, VB.NET, ADSI, PowerShell, and some C#.  I was thrown into VB and vbscript with my job years ago and haven't had the time to venture into much else outside what is directly job related  

IF such an autotune adventure were to be undertaken, one way to go about it would be limit the range of the autotune.  So if the --freq=850 is specified, it would test a range of say 200Mhz above and below (if needed but not likely).  It would certainly take longer....a lot longer, but if autotune is only run once then I really see no harm in that. 
Also, you could have autotune dump results to a file once completed - and if you're adding config file support, you could dump the autotune results into a JSON formatted config file that could be loaded on startup. 

Just ideas.  If I knew C now, I'd be playing with the code. 

[edonkey]

This is my result with latest Cpuminer on one Vmodded GS

Nice!

Which voltage mod did you do? Last I checked there were a bunch of variations.

[jamieb81]

This is my result with latest Cpuminer on one Vmodded GS

Nice!

Which voltage mod did you do? Last I checked there were a bunch of variations.

I used the 47K resistor mod ( non smd )

I have to say I had a bunch of them and this GS works really well, others were a tat lower in speed

[reactor]

So since any frequency is allowed now and not just 25Mhz steps, does autotune initially start with 25mhz steps and then fine tune it from that point down to something like 5mhz steps?

That would practically be impossible to manage, because for example 825 and 838 MHz are not equivalent, by that I mean that if 838 MHz is stable, there is no guarantee that 825 would also be stable. The commands are calculated from different PLL dividers and that can affect stability, by contrast if you use the same frequency stepping, it's almost 100% certain that a lower frequency will be stable if the current one is.

What?  You wouldn't like a challenge like that?  J/k

Makes sense.  But if they use different pll dividers would it be possible to create 2 autotune functions - each confined within its own pll divider?

That's kind of like the code I like to tinker with but unfortunately I don't know the language cpuminer is in and learning it takes more time than I have right now

That would mean autotune would take much much longer, because it would have to test all (or a selected set) of the frequencies with differing PLL dividers each time it stepped up or down.
Cpuminer is written in the most basic language available to man, C. Most of the modern languages today are written in C.

Unfortunately the languages I'm familiar with are VB, VBScript, VB.NET, ADSI, PowerShell, and some C#.  I was thrown into VB and vbscript with my job years ago and haven't had the time to venture into much else outside what is directly job related  

IF such an autotune adventure were to be undertaken, one way to go about it would be limit the range of the autotune.  So if the --freq=850 is specified, it would test a range of say 200Mhz above and below (if needed but not likely).  It would certainly take longer....a lot longer, but if autotune is only run once then I really see no harm in that. 
Also, you could have autotune dump results to a file once completed - and if you're adding config file support, you could dump the autotune results into a JSON formatted config file that could be loaded on startup. 

Just ideas.  If I knew C now, I'd be playing with the code. 

Sandor posted code for polling the api and someone else added on code for outputting a freq string with per-miner-per-chip freqs.  Try page 80/81, even for someone who is as rusty with linux as I am I was able to get the stats up and running on my rpi in ~10 minutes after installing a web server.  That's worked well for me, autotune for a while then when things look stable'ish grab the output freq string.

[RowanX]

Rejected shares have nothing to do with cpuminer-gc3355, proof:

As I said before, it's 50% you, 50% the pool you're on.

On that basis I switched pool to same as you, and after 11 hours I have these stats:

CPUminer v1.0a (11 hours)
Pool: multi.ghash.io (16 diff)
A:24635  R:371  H:13 (1.5% rejects)

Screenshot (CPUMiner): http://i.snag.gy/PLW9h.jpg
Screenshot (Pool): http://i.snag.gy/vgZcR.jpg

1.5% rejects is nothing to worry about I suppose, just wondering why you get 0 rejects.

[sandor111]

Rejected shares have nothing to do with cpuminer-gc3355, proof:

As I said before, it's 50% you, 50% the pool you're on.

On that basis I switched pool to same as you, and after 11 hours I have these stats:

CPUminer v1.0a (11 hours)
Pool: multi.ghash.io (16 diff)
A:24635  R:371  H:13 (1.5% rejects)

Screenshot (CPUMiner): http://i.snag.gy/PLW9h.jpg
Screenshot (Pool): http://i.snag.gy/vgZcR.jpg

1.5% rejects is nothing to worry about I suppose, just wondering why you get 0 rejects.

Most likely related to the quality and speed of your internet connection.

[RowanX]

Rejected shares have nothing to do with cpuminer-gc3355, proof:

As I said before, it's 50% you, 50% the pool you're on.

On that basis I switched pool to same as you, and after 11 hours I have these stats:

CPUminer v1.0a (11 hours)
Pool: multi.ghash.io (16 diff)
A:24635  R:371  H:13 (1.5% rejects)

Screenshot (CPUMiner): http://i.snag.gy/PLW9h.jpg
Screenshot (Pool): http://i.snag.gy/vgZcR.jpg

1.5% rejects is nothing to worry about I suppose, just wondering why you get 0 rejects.

Most likely related to the quality and speed of your internet connection.

Ah, that makes sense.. I'm on quite dodgy wifi.   My neighbours wifi.. can't afford my own since getting made redundant last year.

[sandor111]

Rejected shares have nothing to do with cpuminer-gc3355, proof:

As I said before, it's 50% you, 50% the pool you're on.

On that basis I switched pool to same as you, and after 11 hours I have these stats:

CPUminer v1.0a (11 hours)
Pool: multi.ghash.io (16 diff)
A:24635  R:371  H:13 (1.5% rejects)

Screenshot (CPUMiner): http://i.snag.gy/PLW9h.jpg
Screenshot (Pool): http://i.snag.gy/vgZcR.jpg

1.5% rejects is nothing to worry about I suppose, just wondering why you get 0 rejects.

Most likely related to the quality and speed of your internet connection.

Ah, that makes sense.. I'm on quite dodgy wifi.  My neighbours wifi.. can't afford my own since getting made redundant last year.

Well there you go. Now compare that against 1 Gbit fiber line.

[nst6563]

Sandor posted code for polling the api and someone else added on code for outputting a freq string with per-miner-per-chip freqs.  Try page 80/81, even for someone who is as rusty with linux as I am I was able to get the stats up and running on my rpi in ~10 minutes after installing a web server.  That's worked well for me, autotune for a while then when things look stable'ish grab the output freq string.

I've already pulled the autotune results out of the log file and have no quarrels with doing so, however my suggestion was for dumping to a json config file so it's ready to go - no further 'intervention' needed.  Much like cgminer/bfgminer, it could write the current configuration to a config file (including pools).    Much of my job is streamlining and automating processes so I saw a chance where some steps could be removed 

[RowanX]

I've recently switched to multi coin mining (which auto converts to LTC).

But does this look normal?
http://i.snag.gy/jcEKi.jpg (ghash.io pool screenshot)

For the last hour+ my balance has not been increasing, and in that pool stats you can see even for blocks where I have shares, the payout is 0. Perhaps with only 2 gridseeds it is not feasible to be coin switching? Better stick to one coin (LTC)?

[nst6563]

I've recently switched to multi coin mining (which auto converts to LTC).

But does this look normal?
http://i.snag.gy/jcEKi.jpg (ghash.io pool screenshot)

For the last hour+ my balance has not been increasing, and in that pool stats you can see even for blocks where I have shares, the payout is 0. Perhaps with only 2 gridseeds it is not feasible to be coin switching? Better stick to one coin (LTC)?

Do they only pay out after full confirmations?

[sandor111]

JSON config option has been added, binaries have been updated and the Windows zip includes a sample config.

Code:

{
	"gc3355" : "\\\\.\\COM3,\\\\.\\COM4,\\\\.\\COM5",
	"gc3355-autotune" : true,
	"pools" : [
		{
			"url" : "stratum+tcp://eu.wafflepool.com:3333",
			"user" : "1AMsjqzXQpRunxUmtn3xzQ5cMdhV7fmet2",
			"pass" : "x"
		},
		{
			"url" : "stratum+tcp://doge.ghash.io:3333",
			"user" : "user",
			"pass" : "x"
		}
	],
	"freq" : "850",
	"debug" : true
}

For non-Windows you want to replace

Code:

"gc3355" : "\\\\.\\COM3,\\\\.\\COM4,\\\\.\\COM5"

by

Code:

"gc3355-detect: true"

To include the config, simply run

Code:

/path/to/minerd -c name_of_config_file

[RowanX]

I've recently switched to multi coin mining (which auto converts to LTC).

But does this look normal?
http://i.snag.gy/jcEKi.jpg (ghash.io pool screenshot)

For the last hour+ my balance has not been increasing, and in that pool stats you can see even for blocks where I have shares, the payout is 0. Perhaps with only 2 gridseeds it is not feasible to be coin switching? Better stick to one coin (LTC)?

Do they only pay out after full confirmations?

Not really sure (its ghash.io). Since yesterday my balance was increasing every few minutes. But for last 2 hours, nothing. I will contact their support.

[chinatom]

JSON config option has been added, binaries have been updated and the Windows zip includes a sample config.

Code:

{
	"gc3355" : "\\\\.\\COM3,\\\\.\\COM4,\\\\.\\COM5",
	"gc3355-autotune" : true,
	"pools" : [
		{
			"url" : "stratum+tcp://eu.wafflepool.com:3333",
			"user" : "1AMsjqzXQpRunxUmtn3xzQ5cMdhV7fmet2",
			"pass" : "x"
		},
		{
			"url" : "stratum+tcp://doge.ghash.io:3333",
			"user" : "user",
			"pass" : "x"
		}
	],
	"freq" : "850",
	"debug" : true
}

For non-Windows you want to replace

Code:

"gc3355" : "\\\\.\\COM3,\\\\.\\COM4,\\\\.\\COM5"

by

Code:

"gc3355-detect: true"

To include the config, simply run

Code:

/path/to/minerd -c name_of_config_file

WINDOW CAN'T RUN. 

[wax7]

windows works. looks good.

how to set freq in the config?
only 25 steps?

[nst6563]

How do we specify individual chip frequencies in the config file?

Like this?:
"gc3355-freq" : "\\\\.\\COM9:1200:0,\\\\.\\COM9:1200:1,\\\\.\\COM9:1225:2,\\\\.\\COM9:1225:3,\\\\.\\COM9:1225:4",

[sandor111]

windows works. looks good.

only 25 steps?

No, any frequency can be set.

How do we specify individual chip frequencies in the config file?

Like this?:
"gc3355-freq" : "\\\\.\\COM9:1200:0,\\\\.\\COM9:1200:1,\\\\.\\COM9:1225:2,\\\\.\\COM9:1225:3,\\\\.\\COM9:1225:4",

Yes

[wax7]

mmmh..
"gc3355-freq" : "\\\\.\\COM9:1200:0,\\\\.\\COM9:1200:1,\\\\.\\COM9:1225:2,\\\\.\\COM9:1225:3,\\\\.\\COM9:1225:4",

does not work. caint open device error.

EDIT uh wait that is for a 5-chip gridseed?

How for a Blade?

EDIT2: Sorry its to late here...   Works now. thanks!