Topic: GridSeed 5-chip USB miner voltage mod - page 9. (Read 156980 times)

@Firebird2000 and @poopypants ...:

Could both of you measure the voltage on C42 ( DVDD ) and POST it back...?

Check the dead ones too...
 
ZiG

This seems to be hppening to me as well.
It started today. Out of my 10 units, 3 have dropped out on me.
All have the 49.9k Mod.
I am using a 1200W Gold rated power supply, PCIE connectors to run them.
They have been going along perfectly for a few weeks now, until today.
Now I have 3 dead soldiers.
I run them remotely, so I won't be able to look at them until tomorrow.
But...something is going on here....I'm a bit worried about the other 7......

Could just be fate

Are you using one of those crappy LED power supplies? Could be the cause and a sign that it's about to crap out and ruin all your miners.

I'm using high grade clean switch mode power supplies. Never use cheap crap power bricks or LED power supplies. LED's don't care about high ripple. Seeds do!

Remember, warm seeds are better than cooked seeds 

Good luck!

Is there a proper simple tutorial for noobs?

not really. you could try the simple bridge-voltmod with silver pcb paint

Bit of a noob question here. I have no idea how to solder and replace resistors and stuff. Is there a pencil mod for the gridseed?

and put your black prob to fan's black cable.

Poopy,
12.4V is too low compared to standard 'proper' 12V switch mode power supplies. 13.80 is the minimum voltage it should be outputting under load. More if not loaded i.e. not powering anything.

I will just add my 2 cents for the people thinking about this.... I have tested Vmod3 vs Stock and I GUARANTEE the voltmod pumps more accepted shares than stock.... I had them running side by side with the same pool my test Had 10 Stock vs 10 Vmod3 with the 49.9k swap @1200mhs for an EXACT 510KH NOTHING more or nothing less... Dont get confused by all the people claiming 600KH 550KH stable YES the pool shows what ever sometimes more sometimes less and sometimes VERY high, I listen to what the local hashrates as this is what the POD is ACTUALLY mining at. Just like travelling in a car and looking at the aftermarket GPS showing you how fast your going, you dont look at that right? you look at your cars speedo

 I do not get ANY disconnects now that I use a PROPER USB hub with enough power... I have noticed that if you have 10 port hub you need a 5amp POWER adapter if you dont want ANY Libusb errors or CGminer restarts << these happen when there is not enough power.. THink about it like this --- Each USB Port needs atleast 500ma of power, you can get lucky running less but don't be surprised if you get Cgminer shutting down or your hashrates aren't showing properly.

I purchased a 30day old used 49Port HUb for $80, This has a 40amp fuse and connects to a 24pin ATX power.

Anyway back to my test 10 vs 10, Started the EXACT same time Within 1 Hr this is where you see the Gap between the shares grow, and the gap keeps getting larger and larger, by 2 hrs the gaps show that EVERY single vmod pod beating the Stock and not by small amounts of shares by significant amount. Even the strongest stock do not have a chance...

Without a doubt Vmod trumps stock the only time you may think about not vmodding if your power costs are MEGA $$$ but even then it would have to be VERY Expensive not to take it...

a simple test is to use coinwars just to see if hashrate and powercosts increase or decrease your  daily profit you will see what I am talking about.

Do a test, goto profit calculater click on scrypt and enter your hashrate @ 510KH + 25w + how much your power is then click CALCULATE, check to see your profit then do the same thing but this time choose 361KH @ 7w at the power costs see the difference. PLEASE NOTE, this is ONLY to see the difference between the both not to see how much you will make.

Anyway as Of now I have modded every single pod to 49.9k and I run them at 1200mhz I hash tested 1275, 1250, 1225 & 1213   ---- DEFINITLEY 1200 i feel is the best but `1213 is OK also But id rather 1200 and seeing less HW.... I havent done extensive calculations between the HW error rates and the accepted between the above frequencies because I feel its a waste of time... 1200 @ 510KH is perfect for me..

Just my 2 cents.

Make sure ur gridseed is not powered on. Take off the 12v plug from gridseed. Set ur voltmeter to continuity test mod. Put the red prob from volt meter to blown FB pad(one at a time) and put your black prob to fan's black cable. Does it make any sance to u?

You should get 2 result. Both should indicate no continuity at all.

Could you explain a little further how to check the power.

First check a continuity test on ur board between 12+ and 12- and USB power +/- if you see any short between them , there will be no point to solder another FB, because it will blow again. If u do find a short between any of ur 12v and 5v rail , report back. I assume u own a basic multimeter .

PS: because of ur blown FB you will not be able to use USB socket pin to check the continuity . best thing probably to check from one of the pad for BLOWN FB to ground.

Poopy,
12.4V is too low compared to standard 'proper' 12V switch mode power supplies. 13.80 is the minimum voltage it should be outputting under load. More if not loaded i.e. not powering anything.
The FB's are acting as high frequency filters to compensate for crappy power being fed by crappy psu's like the LED type you are using.
I am sure it's the culprit!
You can temporarily bypass the unidirectional/non-polarized FB to see if your pod will power back up. There is a good chance that the FB blew because of an overload or dead short down the line.
It could already be a new brick
Good luck!

Both power supplies are the 30amp LED power supplies, I have checked with a multimeter the barrel plugs for each unit, and steady 12.4 volts.  The ferrite bead is the one next to the USB plug on the board labeled FB26.  Does anyone know if FB25 is the same as FB26, I would like to try to desolder FB25 from one of them and plug it into FB26 to see if it starts hashing again before I order parts.  Are ferrite beads directional or can they be placed any which way?  Any tips on a way to check to see if they will hash without the bead, IE jump the pads just to check if they will hash?  Thanks in advance for replies, I am frustrated about the death of 3 and am worried it may domino to the rest if it is something else not just a coincidence.

Both power supplies are the 30amp LED power supplies, I have checked with a multimeter the barrel plugs for each unit, and steady 12.4 volts.  The ferrite bead is the one next to the USB plug on the board labeled FB26.  Does anyone know if FB25 is the same as FB26, I would like to try to desolder FB25 from one of them and plug it into FB26 to see if it starts hashing again before I order parts.  Are ferrite beads directional or can they be placed any which way?  Any tips on a way to check to see if they will hash without the bead, IE jump the pads just to check if they will hash?  Thanks in advance for replies, I am frustrated about the death of 3 and am worried it may domino to the rest if it is something else not just a coincidence.

That sucks.  Very strange also.  3 different mods, 2 different power supplies.  Same part blown on 3 boards.  Tell us more about the power supplies you have.