Pages:
Author

Topic: GridSeed 5-chip USB miner voltage mod - page 37. (Read 156991 times)

sr. member
Activity: 378
Merit: 250
April 22, 2014, 06:09:13 PM
On Crypto Mining Blog Jabberwock (April 6th, 2014 at 00:46) http://cryptomining-blog.com/1839-another-new-voltmod-update-for-the-gridseed-5-chip-btcltc-asic-miners/#comment-842 measured 33k for R139 (what we call  R52 for easy reference).  

Thanks Zig, this might be exactly what I've been hoping for.  For me, I think it is easier to add a new axial resistor than to remove that tiny smd resistor. And, it could be easily re-bridged later if a SHA coin rises to and stays at the top of the profitability lists and you want to mine it.

Ok Guys --- ZigMod is official  -

R46 Pads are bridged from factory

Image from Sandor11 & edited by Zig - R46 position is the pads on the far left all alone circled in black
https://ip.bitcointalk.org/?u=http%3A%2F%2Fi.imgur.com%2FKBfnITv.png&t=539&c=UEjhwR7D7EBeVw

Now you know where it is, unbridge the solder join at the pads with either a knife or sharp blade (GENTLY) . Once the join is cleary un-bridged you may now solder 17kohm 1% 0402 SMD resistor OR an Axial Resistor.

The Above according to Zig should give you 20w @ 510-515KH.

Thanks Zig, as this is your mod (Please correct me if I am wrong)

Guys the first person to try this please share with us your results and pictures preferably comparing the old mod side by side to see if both are to 2 and the same in performance and wattage


What should be the benefit of doing it ?

EDIT: I read through some pages. Aim was to get exactly 49.9kOhm as resistor value, is that correct ?
Zig can you verify this ?

Regards.

You read my findings as I figured out that 49.9K was the correct value and the easiest part to install is a 1/4W Axial 1% tolerance resistor. And the 5V USB fan mod was my work too. But if you can't find 49.9k easily, you can use a 50K 1% 1/4W too with the same results if not a hair more power usage, like less than .50W or something - I guess. I would not recommend going below 49.9k as it will lower the achieved efficiency and profit$!
I agree too, why all these other mods now?
Listen, go with what works. There is tons of evidence on here now that 49.9K is perfect for the job. Optimum speed, optimum hashing, minimum if not non-existent HW errors and very little added power usage.
sr. member
Activity: 378
Merit: 250
April 22, 2014, 06:01:01 PM
Guys help please!

I made an "old" mod, not the latest ZiG one, but all i get is HW, even on 850 Mhz

Here is the picture of my unit. I did not solder the voltage=1 pins (vertical yellow line in instructions). Should that be a reason, why I'm getting only HW errors?
Btw I used 47k resistor, but did not have 0402, so used a bigger one..




so all I'm getting now is (no matter which frequency):



Please help!

Why? I don't get it. Nemercry's replacement of R52 (I think it's his mod) is the only and final mod needed. I thought this was clear to everyone.
I believe the 5V Ultra Quiet Fan USB mod is my mod too. At least, that's what the sequence of events led me to coming up with and reporting, also eons ago.
I found and reported eons ago that the 49.9K 1% tolerance resistor and even at 1200MHz only and don't use anywhere near 20W of power - if I recall correctly. I wanna say it's around 12W or so. No HW errors in general and each miner has its own sweet spot/speed too. I can measure the power consumption again, if anyone wants me to.
Gotta love experimentation! It's fun but sometimes, so unnecessary.
I think we've go the pods mining as efficiently as possible.
I do agree as well that if one wants to go back to mining SHA, change the resistor value back to 33K.
Interesting too about R52 vs R139 and still, everyone replaced the right resistor Wink
sr. member
Activity: 339
Merit: 250
Vice versa is not a meal.
April 22, 2014, 05:54:17 PM
On Crypto Mining Blog Jabberwock (April 6th, 2014 at 00:46) http://cryptomining-blog.com/1839-another-new-voltmod-update-for-the-gridseed-5-chip-btcltc-asic-miners/#comment-842 measured 33k for R139 (what we call  R52 for easy reference).  

Thanks Zig, this might be exactly what I've been hoping for.  For me, I think it is easier to add a new axial resistor than to remove that tiny smd resistor. And, it could be easily re-bridged later if a SHA coin rises to and stays at the top of the profitability lists and you want to mine it.

Ok Guys --- ZigMod is official  -

R46 Pads are bridged from factory

Image from Sandor11 & edited by Zig - R46 position is the pads on the far left all alone circled in black
https://ip.bitcointalk.org/?u=http%3A%2F%2Fi.imgur.com%2FKBfnITv.png&t=539&c=UEjhwR7D7EBeVw

Now you know where it is, unbridge the solder join at the pads with either a knife or sharp blade (GENTLY) . Once the join is cleary un-bridged you may now solder 17kohm 1% 0402 SMD resistor OR an Axial Resistor.

The Above according to Zig should give you 20w @ 510-515KH.

Thanks Zig, as this is your mod (Please correct me if I am wrong)

Guys the first person to try this please share with us your results and pictures preferably comparing the old mod side by side to see if both are to 2 and the same in performance and wattage


What should be the benefit of doing it ?

EDIT: I read through some pages. Aim was to get exactly 49.9kOhm as resistor value, is that correct ?
Zig can you verify this ?

Regards.
sr. member
Activity: 339
Merit: 250
Vice versa is not a meal.
April 22, 2014, 05:52:02 PM
Guys help please!

I made an "old" mod, not the latest ZiG one, but all i get is HW, even on 850 Mhz

Here is the picture of my unit. I did not solder the voltage=1 pins (vertical yellow line in instructions). Should that be a reason, why I'm getting only HW errors?
Btw I used 47k resistor, but did not have 0402, so used a bigger one..




so all I'm getting now is (no matter which frequency):



Please help!

Hello Ajeto,

just remove the solder bridge and your mod will work fine.

Regards.
member
Activity: 97
Merit: 10
April 22, 2014, 02:44:40 PM
Vmod3 is a 47k or slightly higher resistor at the position most call R52.  That's it.  No other soldering anywhere. No solder bridges.  Did you solder any other places on the board?
I also soldered bridge on 3rd position in North East corner (blue circle). Nothing else.

EDIT: And now i see i made unnecessary bridge soldering grrr..

EDIT2: Thanks happydaze for your input, unsoldering that bridge solved my problem. I assumed VMOD3 is somehow complementary to VMOD1
full member
Activity: 140
Merit: 100
April 22, 2014, 02:42:17 PM
Vmod3 is a 47k or slightly higher resistor at the position most call R52.  That's it.  No other soldering anywhere. No solder bridges.  Did you solder any other places on the board?
member
Activity: 97
Merit: 10
April 22, 2014, 02:24:28 PM
Guys help please!

I made an "old" mod, not the latest ZiG one, but all i get is HW, even on 850 Mhz

Here is the picture of my unit. I did not solder the voltage=1 pins (vertical yellow line in instructions). Should that be a reason, why I'm getting only HW errors?
Btw I used 47k resistor, but did not have 0402, so used a bigger one..




so all I'm getting now is (no matter which frequency):



Please help!
ZiG
sr. member
Activity: 406
Merit: 250
April 22, 2014, 11:27:53 AM
To make your life mods easier...Another picture of possible point to soldering/placement/modding/...etc...LOL



With WHITE I marked the COMMON soldering pads of the VID resistors - lower row of pads...which is CONNECTED to the lower pad of R46...SAME electrical point...You CAN USE any of those points as your Solder point for your convenience...easy soldering...

The WHITE BIG circle under this row of pads is Digital GROUND ...DGND...SAME as the Upper pad of R46...same point that Wolfey2014 (nothing against him, except his non-stop adds)... was soldering his axial resistor...

Of course you can use 0402 SMD directly on R46 pads...

The Black rendering is one of the possible "easy to place" position for a axial 1/8 through-hole REGULAR resistor...Fill free to unlock your creativity for the best fit...position placement..."beauty contest"... whatever... Grin

Sorry for the bad quality of my drawings, ...but I am away from home...with small laptop only...

Will be glad to see an independent verification(s) and corresponding result(s) ...to validate the effects of this Voltage mod/Overclocking solution...

Thank you to all participants in this thread for their time to read... and contribute to this effort...

ZiG  
ZiG
sr. member
Activity: 406
Merit: 250
April 22, 2014, 10:52:42 AM
On Crypto Mining Blog Jabberwock (April 6th, 2014 at 00:46) http://cryptomining-blog.com/1839-another-new-voltmod-update-for-the-gridseed-5-chip-btcltc-asic-miners/#comment-842 measured 33k for R139 (what we call  R52 for easy reference).  

Thanks Zig, this might be exactly what I've been hoping for.  For me, I think it is easier to add a new axial resistor than to remove that tiny smd resistor. And, it could be easily re-bridged later if a SHA coin rises to and stays at the top of the profitability lists and you want to mine it.

Ok Guys --- ZigMod is official  -

R46 Pads are bridged from factory

Image from Sandor11 & edited by Zig - R46 position is the pads on the far left all alone circled in black
https://ip.bitcointalk.org/?u=http%3A%2F%2Fi.imgur.com%2FKBfnITv.png&t=539&c=UEjhwR7D7EBeVw

Now you know where it is, unbridge the solder join at the pads with either a knife or sharp blade (GENTLY) . Once the join is cleary un-bridged you may now solder 17kohm 1% 0402 SMD resistor OR an Axial Resistor.

The Above according to Zig should give you 20w @ 510-515KH.

Thanks Zig, as this is your mod (Please correct me if I am wrong)

Guys the first person to try this please share with us your results and pictures preferably comparing the old mod side by side to see if both are to 2 and the same in performance and wattage


Thanks gtraah for your clarification/reiteration of the my proposal for simpler Voltage mod and Overckock...!

I am experimenting with this since March 13 ...when I received my first Gridseed ...my guinea pig...with trim pot
The reason I did not fix the value ...I am still searching for the sweet spot of STABLE operation ...for me it is NO HW errors at all...every HW error means your gridseed is like OFF...dead for this run of hashing...5-10-15 min of waste...lowering your effective hash performance...we are paid for Accepted shares ONLY...at the pool

I am searching for a value...and the corresponding Voltage to the chips...which will give LONG term ...24/7 with the BEST # of accepted shared per 24 hours... STABLE, no HW (0) errors...At the moment it is the 1150-1175MHz speed range...  IMHO

ZiG

EDIT...:

I just put a name...to be able to make a reference to it...not to be confused with other/previous mods...

If I already shared this here...it is ALREADY OPEN SOURCE...It is already community effort/Mod ...it is OUR mod... Grin

Fill free to do/use as you may find fit...@ (at) YOUR own risk... Wink

Happy modding...
ZiG
sr. member
Activity: 406
Merit: 250
April 22, 2014, 10:24:49 AM
I measured R139 before I removed it so I could know the value if and when I restore my units back to stock.  R139 does indeed measure 33k.

Thank you poopypants for confirming R139 value...!

FINAL answer is...   => 33K

Rmod to be soldered to R46 pads ...  =>  17K

Like this...:
 
http://www.mouser.com/ProductDetail/Vishay-Dale/RN60D1692FB14/?qs=sGAEpiMZZMtlubZbdhIBIELo8qaDstI25WRHFu9kJp0%3d ...Regular Axial 1/8W 16.9KOhms ... $0.10 in Qty 50 = $5 + S&H

full member
Activity: 140
Merit: 100
April 22, 2014, 08:33:13 AM
Easier to access the R46 position, less chance of accidentally removing or damaging something else if you are an amateur at soldering like me.  You have that big open space in the fins and on the PCB over by the R46 lettering.  There might be enough room over there to lay an axial type resistor flat so less chance of a short caused by should the solder joint ever fail.  

Maybe bend the resistor ends up to create jumper posts so you can easily switch back to stock and dual mining mode.

Jumper on = stock / single or dual mining capable
Jumper off = single mode 500 Kh/s

Not to scale, ignore resistor value in this photo:

full member
Activity: 134
Merit: 100
April 22, 2014, 07:47:50 AM
I measured R139 before I removed it so I could know the value if and when I restore my units back to stock.  R139 does indeed measure 33k.
sr. member
Activity: 420
Merit: 250
April 22, 2014, 07:38:52 AM
On Crypto Mining Blog Jabberwock (April 6th, 2014 at 00:46) http://cryptomining-blog.com/1839-another-new-voltmod-update-for-the-gridseed-5-chip-btcltc-asic-miners/#comment-842 measured 33k for R139 (what we call  R52 for easy reference).  

Thanks Zig, this might be exactly what I've been hoping for.  For me, I think it is easier to add a new axial resistor than to remove that tiny smd resistor. And, it could be easily re-bridged later if a SHA coin rises to and stays at the top of the profitability lists and you want to mine it.

Ok Guys --- ZigMod is official  -

R46 Pads are bridged from factory

Image from Sandor11 & edited by Zig - R46 position is the pads on the far left all alone circled in black
https://ip.bitcointalk.org/?u=http%3A%2F%2Fi.imgur.com%2FKBfnITv.png&t=539&c=UEjhwR7D7EBeVw

Now you know where it is, unbridge the solder join at the pads with either a knife or sharp blade (GENTLY) . Once the join is cleary un-bridged you may now solder 17kohm 1% 0402 SMD resistor OR an Axial Resistor.

The Above according to Zig should give you 20w @ 510-515KH.

Thanks Zig, as this is your mod (Please correct me if I am wrong)

Guys the first person to try this please share with us your results and pictures preferably comparing the old mod side by side to see if both are to 2 and the same in performance and wattage
full member
Activity: 140
Merit: 100
April 22, 2014, 06:11:28 AM
On Crypto Mining Blog Jabberwock (April 6th, 2014 at 00:46) http://cryptomining-blog.com/1839-another-new-voltmod-update-for-the-gridseed-5-chip-btcltc-asic-miners/#comment-842 measured 33k for R139 (what we call  R52 for easy reference). 

Thanks Zig, this might be exactly what I've been hoping for.  For me, I think it is easier to add a new axial resistor than to remove that tiny smd resistor. And, it could be easily re-bridged later if a SHA coin rises to and stays at the top of the profitability lists and you want to mine it.
sr. member
Activity: 420
Merit: 250
April 22, 2014, 01:42:55 AM

@powersup,

Thanks for the confirmation, buddy...I am glad that you find the same thing...and the nice words are appreciated too...





I am a bit confused, why will I ant to change all these resistors if I am getting as we speak 517 and at time 600 KH/s at a 1200 MHz setting. All I did was remove and replace that 1 resistor and that's it. What will the benefit be on doing the last mods I see here where we need to change so many resistors?

I almost died on stress changing this 1 resistor per miner and still cant believe that they are working.

On my first modification I lost the resistor I just removed, after searching with a loupe I found it on top of another...when I removed it from the top of the other it came out...bringing with him the other one too! That's right,  I had to put that one back and start all over. I WANT TO DO THIS FOR A LIVING! I MISS THE ADDRENALINE RUSH. Send your miners my way, I'll Modify them for a buck!

I am a happy camper... this is their performance right now Smiley

cmilian.1 559 KH/s 128 Diff
cmilian.2 643 KH/s 128 Diff


I think Zig is saying that it is much easier to mod it by doing what he said Once people actually know if 139k is 27k or 33k and its more flexible if you want to revert back plus you can add a switch.. if this was known from the start i think this would now me the main mod.. Since 510k and 20 w is perfect ratio and many people have tested the vmod3 i think it will be hard to switch over unless more documented tests are done... and i wouldnt reccomend wasting to much time, I think 510kh Set & forget is the key right now. Unless offcourse someone figures a nice an easy way to get 600kh+ Stable no HW then Different story.

But saying all this  - Congrats on ZigMod Smiley


Also your figures:

cmilian.1 559 KH/s 128 Diff
cmilian.2 643 KH/s 128 Diff

 are not the actual hardware performance of the miner, they are pool averages. The real value if everything done correctly should be 508-515Kh from the actual miner itself which is Awesome!! And what I am after  (Cgminer & I think Bfgminer show real miner performance.)
ZiG
sr. member
Activity: 406
Merit: 250
April 22, 2014, 12:49:54 AM

@powersup,

Thanks for the confirmation, buddy...I am glad that you find the same thing...and the nice words are appreciated too...





I am a bit confused, why will I ant to change all these resistors if I am getting as we speak 517 and at time 600 KH/s at a 1200 MHz setting. All I did was remove and replace that 1 resistor and that's it. What will the benefit be on doing the last mods I see here where we need to change so many resistors?

I almost died on stress changing this 1 resistor per miner and still cant believe that they are working.

On my first modification I lost the resistor I just removed, after searching with a loupe I found it on top of another...when I removed it from the top of the other it came out...bringing with him the other one too! That's right,  I had to put that one back and start all over. I WANT TO DO THIS FOR A LIVING! I MISS THE ADDRENALINE RUSH. Send your miners my way, I'll Modify them for a buck!

I am a happy camper... this is their performance right now Smiley

cmilian.1 559 KH/s 128 Diff
cmilian.2 643 KH/s 128 Diff


Hey buddy...

This procedure is as simple as it is...:   (STARTING WITH A NEW, UNMODIDFIED, STOCK Gridseed Miner)

1...=> Cut R46 bridge...Grab "Exacto" knife...with the help of magnifying glass (at least for my eyes) CUT in 30 sec or less trace between R46 pads...gently...OK.

2...=> Solder 1 ( ONE) ONLY resistor to the now separated R46 pads  ...23K ... if the R139 is = 27K...
                                                                                         or  ...17K ...if  the R139 is = 33K...

The SUM of these  2 Resistors ...EXISTING and NOT removed R139 ...and the Rmod...ADDED to R46 pads ...should be EQUAL to your desired value...in that case 50K...(49.9 with 1% tolerance is between 49.40K to 50.39K)...

AGAIN  - 1 CUT...in 30 seconds...+   1 SOLDER...could be even REGULAR 1/8W axial resistor...

NOT "ALL THESE RESISTORS"...JUST ONE...Capisco...? ... Wink

But you...or anybody else... is NOT FORCED to do anything...everybody is free to make his own decisions/mods...I am just SHARING my observations and results...Fill free to ignore them...  

ZiG  

P.S. The reason that I don't know what the REAL R139 value is...I never measured it...At the moment I am traveling ...on business trip...So I am NOT in my lab...don't have the gridseed in front of me to measure the REAL value of R139...this why I ask here ...Sandor...or somebody... to MEASURE the R139 on the board and to confirm the value of R139...I'll measure myself, when I am back...  
member
Activity: 93
Merit: 10
April 21, 2014, 11:09:43 PM

@powersup,

Thanks for the confirmation, buddy...I am glad that you find the same thing...and the nice words are appreciated too...





I am a bit confused, why will I ant to change all these resistors if I am getting as we speak 517 and at time 600 KH/s at a 1200 MHz setting. All I did was remove and replace that 1 resistor and that's it. What will the benefit be on doing the last mods I see here where we need to change so many resistors?

I almost died on stress changing this 1 resistor per miner and still cant believe that they are working.

On my first modification I lost the resistor I just removed, after searching with a loupe I found it on top of another...when I removed it from the top of the other it came out...bringing with him the other one too! That's right,  I had to put that one back and start all over. I WANT TO DO THIS FOR A LIVING! I MISS THE ADDRENALINE RUSH. Send your miners my way, I'll Modify them for a buck!

I am a happy camper... this is their performance right now Smiley

cmilian.1 559 KH/s 128 Diff
cmilian.2 643 KH/s 128 Diff
ZiG
sr. member
Activity: 406
Merit: 250
April 21, 2014, 10:56:13 PM

@powersup,

Thanks for the confirmation, buddy...I am glad that you find the same thing...and the nice words are appreciated too...



ZiG
sr. member
Activity: 406
Merit: 250
April 21, 2014, 10:47:21 PM
I modified the ORIGINAL image of Sandor111 ...(Thanks Sandor) :

*IMG SNIP*

With black lines are the points of interest from my post...R46 bridge (0 Ohm = short from schematics)...

R139...33K according to Sandor111 ...on Gridseed schematics is clearly marked as 27K...I never took a measurement on the actual board resistor...

R52 marking...and the corresponding resistor on the board...

If R139 is 33K...replace in my calculations "27K" ...with "33K"...

Any comments or questions are welcome...

Cheers,

ZiG



I do have 2 -

1. By simply removing the bridge from r46 pads and soldering 49.9k their is this EXACtly like soldering the 49.9k to r139?

2. What do you mean adjust the MV by 20 at a time etc... how is this done? Please elaborate

No gtraah,

R46 is in SERIES with the R139...

R46mod value is ADDED to the R139 ...be it 27K ( by Gridseed's schematics )  ...or 33K ( by Sandor's posting - unconfirmed ) ...

It is NOT replacing so called R52 ...
 In order to make it 49.9K...or let's round it up to 50K...(IMHO it DOES NOT need to be 49.9 EXACTLY...does NOT make any difference in the mod resulting Voltage)

Rmod = 50K - R139

So you will need to solder (50K - 27K = 23K) ...=>  23K resistor to R46 pads if R139 is 27K

So you will need to solder (50K - 33K = 17K) ...=>  17K resistor to R46 pads if R139 is 33K

This is why I ask Sandor ...or anybody interested ...to measure the actual value on the board itself...


2 question...I am using trimm pot(enciometer) of 50K, soldered to R46 pads to fine tune the Voltage supplied to the greedseeds...with this setup you CAN adjust voltage in 10-20mV steps ...and test the performance @ that voltage...like 1.45V...1.50V etc...

 
sr. member
Activity: 420
Merit: 250
April 21, 2014, 10:30:14 PM
I modified the ORIGINAL image of Sandor111 ...(Thanks Sandor) :

*IMG SNIP*

With black lines are the points of interest from my post...R46 bridge (0 Ohm = short from schematics)...

R139...33K according to Sandor111 ...on Gridseed schematics is clearly marked as 27K...I never took a measurement on the actual board resistor...

R52 marking...and the corresponding resistor on the board...

If R139 is 33K...replace in my calculations "27K" ...with "33K"...

Any comments or questions are welcome...

Cheers,

ZiG



I do have 2 -

1. By simply removing the bridge from r46 pads and soldering 49.9k their is this EXACtly like soldering the 49.9k to r139?

2. What do you mean adjust the MV by 20 at a time etc... how is this done? Please elaborate


EDIT: AGain too quick to type, I noticed you stated that its 22k instead of the 49.k Grr I already have all the 49.9k coming, I am just trying to think what other benefits is rather than being alittle more convenient to perform since you do not have to remove a resistor.

This is where my 2nd question stands -
Pages:
Jump to: