GridSeed 5-chip USB miner voltage mod

[happydaze]

Thanks. I'm there!

[ZiG]

Thanks. I'm there!

Me too...

[Cire]

Hello All,
I've been catching up on this thread for the last week.  And jumped into voltmodding while teaching myself to solder.  I  ttstarted off with a 47k axial resistor and the r52/139 mod.  All went well on both of my newly acquired seeds.  Unfortunately I couldn't hit my target goal of 1MH combined hashrate with cgminer gridseed version.  I was able to hit it with sandors miner but barely.  So back to radio shack i went to pickup some 49k resistors,  which they didnt have.  I decided to get some 3k resistors to run in series.   30 minutes later one seed was remoded and hashing away.  2 hours later I gave up on the second one.  For some reason I am completely unable to get any solder to stick to the spot where i originally removed the resistor at r139/52.  Ive spent the better part of the last day looking for a solution and have came up empty handed.  So I turn to you guys hoping you have the answer.   My thoughts and questions are as follows...

Is there some trick to be able resolder to the original position that isn't taking solder?  Or is there another point i can solder to that will accomplish the same mod?

Can I use ZIG mod with the 47k+3k resistor series instead of the 16.9/17k?  I am under the impression that his mod is intended to work in series with the normally existing 33k resistor at r139.  Which in my case is no longer there, so will the circuit? be complete without it?

Also, how heat sensitive is the "mod area"  I'm concerned that I may have heated things up to much in the general area of the point i was attempting to solder.

Any input that can help me get this miner up and running would be helpful.  I'm a bit out of my depth.

Thanks
CIRE

[happydaze]

Try to get a little blob of solder on that spot to rebuild the pad.
I'd put a bit of flux on the spot then just touch the fine tip iron to it at the same time as putting the solder wire there.  All very quickly.

The spot might be concave now and, as was explained to me here, the PCB has a coating that prevents solder from sticking.  If the pad is a little below the PCB coat you need to rebuild the pad.

After you rebuild the pad then you should be able to solder the resistor.

[ZiG]

Hello All,
I've been catching up on this thread for the last week.  And jumped into voltmodding while teaching myself to solder.  I  ttstarted off with a 47k axial resistor and the r52/139 mod.  All went well on both of my newly acquired seeds.  Unfortunately I couldn't hit my target goal of 1MH combined hashrate with cgminer gridseed version.  I was able to hit it with sandors miner but barely.  So back to radio shack i went to pickup some 49k resistors,  which they didnt have.  I decided to get some 3k resistors to run in series.   30 minutes later one seed was remoded and hashing away.  2 hours later I gave up on the second one.  For some reason I am completely unable to get any solder to stick to the spot where i originally removed the resistor at r139/52.  Ive spent the better part of the last day looking for a solution and have came up empty handed.  So I turn to you guys hoping you have the answer.   My thoughts and questions are as follows...

Is there some trick to be able resolder to the original position that isn't taking solder?  Or is there another point i can solder to that will accomplish the same mod?

Can I use ZIG mod with the 47k+3k resistor series instead of the 16.9/17k?  I am under the impression that his mod is intended to work in series with the normally existing 33k resistor at r139.  Which in my case is no longer there, so will the circuit? be complete without it?

Also, how heat sensitive is the "mod area"  I'm concerned that I may have heated things up to much in the general area of the point i was attempting to solder.

Any input that can help me get this miner up and running would be helpful.  I'm a bit out of my depth.

Thanks
CIRE

If you could post some pictures...it'll be easier to help...

ZiG

[c275]

So, the ultimate mod is replace the resistor and soldering a 49.9K without the two solder like the image below?  

[wolfey2014]

So, the ultimate mod is replace the resistor and soldering a 49.9K without the two solder like the image below?  
Image went "POOF!"

Correct!
Do the 49.9K resistor mod 'ONLY'. It is all you need.... R139 or what ever it's being called these days
You may use 50K 5% tolerance also..it will be fine.

I'd caution you on trying to run your pods at maximum clock. I'm running mine at 1163 and have been for weeks now. It is THE sweet spot for all 49.9k modded pods IMO!
Highest percentage of Accepts / Valid Shares and nearly ZERO HW errors are being had at this setting.

Good luck!

Wolfey2014

[Invigiator]

How do you guys look up the component values? Can anyone give me the value of this resistor?

I don't even know if it is R66 or R106. Seems to have blown out on my overclocked gridseed, but not sure if it has anything to do with the overclocking or just random.

http://i61.tinypic.com/2ch9652.jpg

[MoonShad0w384]

How do you guys look up the component values? Can anyone give me the value of this resistor?

I don't even know if it is R66 or R106. Seems to have blown out on my overclocked gridseed, but not sure if it has anything to do with the overclocking or just random.

--image snipped--

I think that component is actually FB26, and R66 and R106 are the two empty spots for resistors immediately to its left in your pic.

[chMiner]

How do you guys look up the component values? Can anyone give me the value of this resistor?

I don't even know if it is R66 or R106. Seems to have blown out on my overclocked gridseed, but not sure if it has anything to do with the overclocking or just random.

Have a look to this post! https://bitcointalk.org/index.php?topic=519112.msg6556103#msg6556103

[Invigiator]

Ha!!  Common fault. Genius ... many thanks

Damn! I think I am stuck.. I figured this is just a suppressor across the ground for the USB, so soldered a bridge over it to re-make the connection but the gridseed is still dead with no lights and no USB device detection.

Any ideas? If this is a common fault has anyone fixed it successfully?

How do you guys look up the component values? Can anyone give me the value of this resistor?

I don't even know if it is R66 or R106. Seems to have blown out on my overclocked gridseed, but not sure if it has anything to do with the overclocking or just random.

Have a look to this post! https://bitcointalk.org/index.php?topic=519112.msg6556103#msg6556103

[happydaze]

One of my modded gridseeds quit yesterday. There's no real obvious signs of a short or any damage.
The fan doesn't spin.  When I plug in 12v the fan moves a fraction.  USB seems to work.
If I plug this gridseed into the psu with all the other gridseeds the psu shuts off.

Would a failed mofset mosfet cause symptoms like this?  I ask because there is a very small bump on the top of one of the mofsets mosfets.

Any ideas?
Are there any tests I can do?

[ZiG]

One of my modded gridseeds quit yesterday. There's no real obvious signs of a short or any damage.
The fan doesn't spin.  When I plug in 12v the fan moves a fraction.  USB seems to work.
If I plug this gridseed into the psu with all the other gridseeds the psu shuts off.

Would a failed mofset cause symptoms like this?  I ask because there is a very small bump on the top of one of the mofsets.

Any ideas?
Are there any tests I can do?

Probably 12V short by burned MOSFET...check for shorts ...

ZiG

[happydaze]

Thanks ZiG. 

-----------------------------

Anyone going to undo the vmod 3?  I plugged in 100kh/s (gain), 12 watts (extra watts), 10 cents kw/h (my cost) in coinwarz. 

For LTC:  Revenue $0.04 / Profit $0.01 / $0.03 for electricity

I think I'll just solder the two bridges for vmod1 to undo vmod3 on all my gridseeds when I have time.

[nst6563]

One of my modded gridseeds just smoked this afternoon.  The same 12v mosfet blew that others have had go.  Did some SERIOUS damage to the board too.  The VCC pad is completely gone and it burned out the board down to the 2nd layer of copper.  In order to fix it my gridseed will look like fricken frankenstein.  Trying to figure out if I would be better served to mount the mosfet vertically with a sizeable heatsink on the back or patch up the board and lay it flat again and run a wire from the vcc pin around to the back side of the board or jumper it to another location on the top.  Wire will be required no matter what.

[foz]

Hi all

This has been a great thread...thanks to all who have contributed.  Another data point to add.

I had two of my ferrite beads melt on one of my 49.9k volt modded units while it was running at 1200 mhz.  It had been running for weeks without an issue.  However the failure occurred just after I cycled the power on the unit (RM850 Corsair).

After looking at the damage does anyone think it is worth ordering the two ferrite beads and attempting a repair?

http://i58.tinypic.com/2pshkzq.jpg

The black wire in the bottom of the photo is the neg fan lead...the positive fan lead got hot enough to separate from the board.

Thanks

[wolfey2014]

Hi all

This has been a great thread...thanks to all who have contributed.  Another data point to add.

I had two of my ferrite beads melt on one of my 49.9k volt modded units while it was running at 1200 mhz.  It had been running for weeks without an issue.  However the failure occurred just after I cycled the power on the unit (RM850 Corsair).

After looking at the damage does anyone think it is worth ordering the two ferrite beads and attempting a repair?

image snip...

The black wire in the bottom of the photo is the neg fan lead...the positive fan lead got hot enough to separate from the board.

Thanks

Hello.
Sorry for your hassles. That's a bummer!
There is really no point in running the 5 chippers at 1200+ MHz! Doing so will not yield any better performance and profit than running it at less than 1170MHz...
I run all of my pods at 1163 as that seems to be the sweet spot for maximum performance.
1200MHz is just a waste of energy and....hardware ;/
If you are running them off of a really clean psu with very low ripple etc. I personally think that simply bypassing the FB's may be okay to do, but I have not tried this myself.
Good luck!
Wolfey2014

[foz]

Hi Wolfey :-)

Thanks for your input.

So your running the 49.9 mod at 1163?  Or are you doing the 47k mod now?  I lowered my freq to 1163 on the 49.9 mod and really didn't see any noticeable power usage difference from the 1200 freq

[wolfey2014]

Hi Wolfey :-)

Thanks for your input.

So your running the 49.9 mod at 1163?  Or are you doing the 47k mod now?  I lowered my freq to 1163 on the 49.9 mod and really didn't see any noticeable power usage difference from the 1200 freq

Hey Foz!
Yep, 49.9K and actually some are running 50K 5% as well.
I do see a noticeable difference between the two frequencies.
Of course there would have to be as PWM goes.
But the main difference is in the overall reward for work performed.
I see higher reward per hash at 1163 than at 1200.

If you don't see much difference between the two, I'd set them to the lower frequency if I were you.

I just had one of my 'latest' silver pods take a crap on me. The main FET made magic black smoke the other day. Gee! Thanks a lot chippy! POS!
Anyway, when I get the time, I'll see about replacing it with a good ol' discreet TO-220 type FET.
Maybe one of my left over 45Amp 5V logic gate ones or just a good ol' analog 12V gate type if needs be as I don't know what the gate drive Voltage is yet.
Maybe I'll pull out my gold ol' CRO and have a look at the PWM signal on the gate since I'll have it torn apart anyway..fun fun fun....puhh!
Woof Woof!

[wolfey2014]

One of my modded gridseeds quit yesterday. There's no real obvious signs of a short or any damage.
The fan doesn't spin.  When I plug in 12v the fan moves a fraction.  USB seems to work.
If I plug this gridseed into the psu with all the other gridseeds the psu shuts off.

Would a failed mofset mosfet cause symptoms like this?  I ask because there is a very small bump on the top of one of the mofsets mosfets.

Any ideas?
Are there any tests I can do?

Probably went bad because they're mofsets

You have a short! If there is a bump on the FET, rest assured, it's TOAST!

Time for a replacement or perhaps just buy another pod...I hear their selling for less than $50 on eBay now.

Or, just wait for Flower to 'FINALLY' release their 10MHs pods and buy them.
The seeds are pretty well all done as of now, no thanks to alt coins halving in value a few times by now..
I mean, mine are making me a whole 25c/day each in DOGE now! WOO HOO!
Actually, I'm mining fresh new coins like VirtaCoin which are paying a lot better, once it gets on the exchanges that is DOGE to the moon? I don't think so...
Woof Woof!