GridSeed 5-chip USB miner voltage mod

[worldlybedouin]

Is anyone offering voltage mod services in the NY/NJ area?

I wouldn't mind getting my 10 Gridseeds volt-modded by a pro and get a little extra bang for the buck. 

Please PM me if you can help with this.  Thanks!

[1715231302]

1715231302

[Invigiator]

I need help! I carried out the voltage mod successfully for the first gridseed, but then when doing the second I accidentally damaged an adjacent component. I was replacing the resistor in this picture:
http://cryptomining-blog.com/wp-content/uploads/2014/04/gridseed-68-kohm-r52-resistor-mod.jpg

Can anyone please tell me what the (black) component is to the right of the one circled in green? There are two components to the right of the one circled in green and they are perpendicular to one circled in green. The one I damaged was the black one at the bottom, i.e. closest to the edge of the board.

Any help in telling me what component that is (resistor) and what value, would be very much appreciated.

[Invigiator]

I need help! I carried out the voltage mod successfully for the first gridseed, but then when doing the second I accidentally damaged an adjacent component. I was replacing the resistor in this picture:
http://cryptomining-blog.com/wp-content/uploads/2014/04/gridseed-68-kohm-r52-resistor-mod.jpg

Can anyone please tell me what the (black) component is to the right of the one circled in green? There are two components to the right of the one circled in green and they are perpendicular to one circled in green. The one I damaged was the black one at the bottom, i.e. closest to the edge of the board.

Any help in telling me what component that is (resistor) and what value, would be very much appreciated.

It's the one circled in blue here: http://tinypic.com/view.php?pic=330g13c&s=8#.U4NoGa1dVWQ

[happydaze]

Looks like another zero value ...........



If you have a multimeter measure it on your other Gridseed to be sure. 

The zero value looks to be confirmed in the comments here: http://cryptomining-blog.com/1839-another-new-voltmod-update-for-the-gridseed-5-chip-btcltc-asic-miners/

Just as a side note, on this new voltmod, we aren’t replacing the R52 resistor (that should be 0 ohm)

[Invigiator]

Thanks sooo much for your help! I really really appreciate it.

I soldered the broken component in with globs of solder to bridge the gap and it seems to have worked since it's running OK.
That's the last time I solder those microscopic components with a regular hobby soldering iron ...

Thanks again!

Looks like another zero value ...........

https://i.imgur.com/oTxJqEH.jpg

If you have a multimeter measure it on your other Gridseed to be sure. 

The zero value looks to be confirmed in the comments here: http://cryptomining-blog.com/1839-another-new-voltmod-update-for-the-gridseed-5-chip-btcltc-asic-miners/

Just as a side note, on this new voltmod, we aren’t replacing the R52 resistor (that should be 0 ohm)

[nj_m715]

I need fb23/24. the pdf says: 2300ohm@100MHz,3A
but I think the 2 is a typo, I can not find 2300ohm fb for sale anywhere. I found 300ohm beads, but don't know what size to order. Where did you guys order your ferrite beads?  thanks

[CryptKeeper]

I need fb23/24. the pdf says: 2300ohm@100MHz,3A
but I think the 2 is a typo, I can not find 2300ohm fb for sale anywhere. I found 300ohm beads, but don't know what size to order. Where did you guys order your ferrite beads?  thanks

AFAIK you need these fellows:

742792121, FERRIT, SMD 300OHM, 3A 1206, RoHs http://r.ebay.com/mDsXD1

[geetash]

Can I just do the VMOD3 here: https://bitcointalk.org/index.php?topic=575315.0#post_step4
And just used the modified cgminer?

yep why not

[geetash]

Looks like another zero value ...........



If you have a multimeter measure it on your other Gridseed to be sure. 

The zero value looks to be confirmed in the comments here: http://cryptomining-blog.com/1839-another-new-voltmod-update-for-the-gridseed-5-chip-btcltc-asic-miners/

Just as a side note, on this new voltmod, we aren’t replacing the R52 resistor (that should be 0 ohm)

measured it and its zero

[MoonShad0w384]

Did the 2-bridge mod and replaced the thermal pad with some paste and mine is running stable fanless at 950MHz (~403kh/s). Thanks!

I also had to shift the heatsink vanes slightly, so that the top half didn't collide with the bottom half. Otherwise it would not have made good contact with the chips.

[happydaze]

MoonShad0w384 - have you tried Sandors cpuminer-gc3355 ? https://github.com/siklon/cpuminer-gc3355

Note: You'll want your fan running before you do 

I have two unmodified pods doing 396.1 kh/s & 397.1 kh/s on cpuminer-gc3355.  I just haven't gotten around to volt modding these yet.  My other pods are vmod3

I curious to see what Sandors cpuminer-gc3355 + vmod1 would do.  Anyone here have that combo?

At some point we might all have to revert to vmod1 to save a few watts.  One profitability calculator I used today said my Fury wouldn't cover the cost of electricity to run it by next Feb!

[ZiG]

MoonShad0w384 - have you tried Sandors cpuminer-gc3355 ? https://github.com/siklon/cpuminer-gc3355

Note: You'll want your fan running before you do 

I have two unmodified pods doing 396.1 kh/s & 397.1 kh/s on cpuminer-gc3355.  I just haven't gotten around to volt modding these yet.  My other pods are vmod3

I curious to see what Sandors cpuminer-gc3355 + vmod1 would do.  Anyone here have that combo?

At some point we might all have to revert to vmod1 to save a few watts.  One profitability calculator I used today said my Fury wouldn't cover the cost of electricity to run it by next Feb!

Hey Happydaze,

How is the Fury...?

Is it overclockable...?

Time for a new Volt Mod...

Cheers,
ZiG

[happydaze]

The Fury needs a voltmod!   5% hardware errors is "normal" right now.
Had you asked me this morning I would have said forget it.  I'm not used to so many rejects and hardware errors.  The gridseeds run way more stable.
But, this afternoon someone on Litecointalk released a first attempt of BFGMiner and it's a big improvement.  Still 5% HW errors but hardly any rejects.

There should be some overclock room.  The 328 clock that most seem to run results in about 230kh/s per chip.  That's already a bit of an overclock from 210khs (300 clk) which would be what Zeus last spec'd the chips at.  These chips started off being promoted as 300kh/s chips way back.  The problem is power consumption of 8 watts per chip now at around 230kh/s.

I'm running the Fury at 340 clk which should be around 238 kh/s x 6 chips = maybe 1.4Mh/s with 5.7% HW errors.  48.75 watts for just the device itself.  I'm waiting for 24hr poolside results then I'll adjust up or down.

Remember how the voltmods really seemed to stabilize the Gridseeds?  That's what I think the Fury needs right now - get it stabilized. But, because it already using 50 watts there might not be much to gain. Who knows? Maybe half a volt more would stabilize these things.

I'm not 100% sure but I think Terry from Zeus mentioned 1.3 volts as the voltage in a post on Litecointalk.  Fury's & Blizzards might have potential for a volt mod.  Zeus needs to leak a schematic so the pros here can figure it out.  The 12v 5A power supplies that come with them are already maxed out though.  I used the same type of plug as I used with my gridseeds.

[nst6563]

We should start a new Fury/Blizzard thread for those of us with those devices.  I thought about doing it, but haven't had the chance to really sit down and compile the information and stuff.

These things do run hot, and if you look at the circuit board, there are some test points with voltage markings next to them.  I have a good idea that the points are labeled for core voltage, bus voltage, etc.  If you check some of the pics on this link you can see in the upper left there is a point labeled 1.2vc...I would assume that stands for 1.2v core.  There are some other points on the other side of the board labeled 3.3vb and two labeled 1.2vb.  I haven't measured them yet though.

So far I've been trying to reduce the number of errors with settings and cooling.  I found some discussion on litecointalk where they mentioned that the speed 336 combined with higher difficulty reduces the number of errors.  I set the speed to 336, and at least on Manicminer pool the diff is set to 256 (or close to it) and so far out of 650 accepted shares, I have 29 hardware errors and 0 rejects (4.4%?! if my math is right...I'm tired as hell)

I also added some extra cooling.  The back of the board is smeared with a VERY thin layer of thermal paste...and it actually was making hardly any contact with the heatsink.  They should have used more paste or a thermal pad.  So I put dabs of paste on the backside of the board where the chips are, and added some heatsinks to the tops of the chips as well as the mosfets.  After the added thermal paste on the backside of the board the heatsink is noticeably warmer indicating more heat is being transferred to it from the board....so that's good.  The heatsinks on the top of the chips don't do a lot of good by themselves unless there's a fan pushing air through the case over them but that's for another time.  I have a second Fury coming so I could slap them together and make sort of a blade configuration with a fan on one end blowing over all of it.

Anyway...here's a pic where you can see the voltage test points and added heatsinks.

[ZiG]

Thanks for the info, guys...

Nice picture...and a nice beginning of the next journey to the "VoltMod Land"...

Any ideas about the power section...chip(s), ...schematics...etc...

Let's start another thread...if interested...


ZiG

[MoonShad0w384]

-snip-

Anyway...here's a pic where you can see the voltage test points and added heatsinks.

What happened to your heatsinks? It looks like they got run over by a truck...

[happydaze]

here's the post re: Zeus voltage https://litecointalk.org/index.php?topic=16301.msg177074#msg177074

it's possible for a physical mod. if change some of the resistors to get it to work at 1.2 ish V, you save some more power there comparing to 1.3.

power consumption is related to voltage it's running on, which is tunable with hardware mods.

nst6563's photo shows 1.2v printed, I wonder if the voltage is already upped a bit?

[wolfey2014]

Thanks for the info, guys...

Nice picture...and a nice beginning of the next journey to the "VoltMod Land"...

Any ideas about the power section...chip(s), ...schematics...etc...

Let's start another thread...if interested...


ZiG

Smells like more magic black smoke is on the horizon! 

[nst6563]

-snip-

Anyway...here's a pic where you can see the voltage test points and added heatsinks.

What happened to your heatsinks? It looks like they got run over by a truck...

LMAO! Well...yes and no....I had a fight with them.  They weren't agreeing to be cut up and it was the last heatsink I had of that variety that would fit with the height I had to work with.  Busted my last dremel cutoff wheel, so had to resort to the air powered metal saw I have...which the blades aren't that great for.  Didn't have a vice hear it, so I has holding the heatsink with pliers and a death grip.  In the end though...it's not the looks but the functionality I was after, an the scratches and marks just add to the effective surface area.  

I measured the 1.2vc mark and it reads 1.26v.  So if the chips were originally spec'd at 1.3v, there's some room to play with.  

As for IC's....I don't have a camera other than the one on my htc....but around the choke are all the same type of mosfet with the markings:
2R030
PEM
1349A6
4214

The other chip that looks to control the voltage is marked with:
0193
391
CCON

Since this is a Fury, there are NO markings on any of the ASIC IC's themselves.  The only other IC with markings is the Silab's USB interface - which from my understanding can support up to 1Mbps transfer rate however the Fury/Blizzard have it limited to 115200Kbps (could there be some performance gained by upping this limit??  It appears that the chips themselves are daisy chained to a nonce solved by the last chip in the chain has to travel through each chip before it and then to the interface....it could be possible that more data is flowing through the port than 115200Kbps can effectively handle I suppose).

:EDIT:  So it looks like the 0193 chip is possibly this TPS40193:  http://www.ti.com/lit/ds/slus719e/slus719e.pdf 

[nst6563]

I started a Fury thread (I suppose the Blizzards would apply too).

https://bitcointalk.org/index.php?topic=644921.0