Pages:
Author

Topic: GRIDSEED G-BLADE Overclocking 7Mh/s, improvements and repair - page 4. (Read 74040 times)

hero member
Activity: 840
Merit: 1000
I thought the same other day, to use to-220 you can cool it better and still can be almost at the same spot so you don't have dissapation on wires there is high current here.

Why they chaneg chips which are almost the same, only the price can be factor. They get chips for much lower price.
Did you find replacement for IRFH5301 also, to-220 housing?

I think I will order chips tomorrow because it takes almost 3 weeks to get it.

CryptKeeper's pod with IRFB7437?
he also did some modifications?
where is his topic?

He has blown ferrite beads, so he sent the pod to me to try to revive it.

one power mosfet was shorted, so I replaced it with IRFB7437.

You can see on the picture that when exploding, the ferrite beads have badly damaged the PCB near the solder points for the fan.

If you do TO-220 conversion, beware the Gate/Source/Drain could be misplaced.

So here comes Frankenpod







It works just fine, I disconnected it after a few seconds to avoid overheating the GC3355, but even the 7437 was hot. It will need active cooling with dissipator.



I have a gridseed pod with this exact chip burnt out, it's a MOSFET? I'm fairly handy with a hot-air gun/soldering iron, can you tell me what part to buy / where to get it so I can replace this frazzled chip please? Smiley

BTW the unit is still recognized as a USB device and seen by cpuminer, it just won't hash.

Thanks in advance for the info! Smiley

Hi, sorry for the later answer. The upper gate power MOSFET on older blades is IRFH5302. It can be swapped with a csd16321q5c that will run coller. Add small dissipators as I postes on top of the heatsinks to. It will help.
sr. member
Activity: 412
Merit: 250
People still think buying new miners will make them money, but will not, because more and more powerfull miners come out and dificulty raise exponentially. So what people do, buys miners and mine and when get something they buy more powerfull with that money and so on and so on.
At the end they don't have any profit from miner just buying new miners and that is all and they get no money out of it for them self.

Only one which benefits here and makes big money are companies which sell miners and that's why tere is more and more companies which sell them.

So the only way to get some money out of mining is to build your own miner, miner which can be upgraded to get more hash out of it.
Anything else is just wasting time and electricity.
newbie
Activity: 28
Merit: 0
Hi ! Again!  Smiley

I have to spend now more time on my study Undecided , so no time more for the g-blades...
Too much time I spent on it , with no benefit Embarrassed .
There are now Prototypes of 1024 MHS Miners in Q4 planed , and i think even if we push a Gblade to 10 MHS , it´s pointless. I have spend over 1800 USD on my both Gblades and countless hours in repairing and tweaking and for now I have only maybe 200 USD back if I´m lucky, and they allready reach the point to consume more than  they produce, with 0.365 USD / kWH in Germany, its allready painfull to mine. The only hope that I have is to hold some Coins like Doge and Litecoin in the hope , that they rise up at the end of year, but when the "monsters" are introduced in this game soon, like the 500 MHS and 1 GHS Miner, I don´t know if someone bother for something less than 100 MHS. And the most coins will die if they are only mined and imediatly sell. Only sell orders and no buy orders let the prices drop and drop until the coin is valueless. Like the Dogecoin , there are now some plans to realize merged mining with Litecoin , to save Dogecoin and stop it from slowly dying.
 
It was a pleasure for me to "talk" with you guys and I´m pretty sure that is the best thing to build a own miner , its a interesting plan to realize.
But for the moment i´m really disappointed of the situation and it´s hard for me to realize that i play a game with my own hard earned money , a game with no chance to win.
It´s time to face the truth. I have tried to get some investment back, in offering a repair of gblades or overclocking , but the blades are allready forgotten , so no chance to get something back.   

newbie
Activity: 4
Merit: 0
Replaced the 33k resistor with a 39k and the miner is holding steady at 6Mh/s! Decided to go for a moderate overclock to keep power costs down and reliability up!

http://img.photobucket.com/albums/v293/war6763/ScreenShot2014-08-07at12453AM_zpse07c7f42.png

An old heatsink I found inside a plasma TV. I also replaced the fan to keep the noise down and removed the barrel connectors. In their place I soldered 16 gauge wire and added some Anderson PowerPole connectors.

http://img.photobucket.com/albums/v293/war6763/2014-08-07012733_zpsd16a6a3a.jpg

MOSFET Heatsink Temperature:

http://img.photobucket.com/albums/v293/war6763/2014-08-07012742_zpsbf342e2c.jpg

Switching Capacitor Temperature (Swapped for Low ESR Capacitors)

http://img.photobucket.com/albums/v293/war6763/2014-08-07012749_zpsfba4eb1c.jpg
full member
Activity: 182
Merit: 250

Looks like you might be able to Jtag the microcontroller right on the board.
I would guess its locked, did you try and dump any code?

stm32f103rct6 short circuit.
I soldered a new microcontroller.
Firmware with Gridseed 5-chip not suitable.

Who has the opportunity to merge with the microcontroller firmware stm32f103rct6. This can be done using UART, through software Flash Loader Demonstrator



1 - 3,3 v
2 - RST
3 - Rx
4 - Tx
5 - Boot0
6 - Boot1
7,8 - GND
sr. member
Activity: 346
Merit: 260
Hello
On Gridseed G-Blade burned microcontroller stm32f103rct6, can anyone have firmware for it

Looks like you might be able to Jtag the microcontroller right on the board.
I would guess its locked, did you try and dump any code?
full member
Activity: 182
Merit: 250
Hello
On Gridseed G-Blade burned microcontroller stm32f103rct6, can anyone have firmware for it
newbie
Activity: 5
Merit: 0
i think the ferrite beads are 300ohm @ 100MHz, 3A, 805 type, that's what I used to replace a burnt out one anyway and it worked. Smiley

Thank you.  This has been sucking for me since the g-blade this happened to was best to OC of the two I have and had the fewest HW errors.

Just out of curiosity, did your performance improve any after this repair?
hero member
Activity: 655
Merit: 504
You wan chili saus?
I don't suppose somebody could tell me what FB28 and FB29 are right next to the barrel connector? FB28 just blew on me, I have a friend who might be able to fix it for me and I'd like to be able to approach him halfway intelligently on this since pcb are not my forte.  Thanks.

i think the ferrite beads are 300ohm @ 100MHz, 3A, 805 type, that's what I used to replace a burnt out one anyway and it worked. Smiley
hero member
Activity: 655
Merit: 504
You wan chili saus?
I thought the same other day, to use to-220 you can cool it better and still can be almost at the same spot so you don't have dissapation on wires there is high current here.

Why they chaneg chips which are almost the same, only the price can be factor. They get chips for much lower price.
Did you find replacement for IRFH5301 also, to-220 housing?

I think I will order chips tomorrow because it takes almost 3 weeks to get it.

CryptKeeper's pod with IRFB7437?
he also did some modifications?
where is his topic?

He has blown ferrite beads, so he sent the pod to me to try to revive it.

one power mosfet was shorted, so I replaced it with IRFB7437.

You can see on the picture that when exploding, the ferrite beads have badly damaged the PCB near the solder points for the fan.

If you do TO-220 conversion, beware the Gate/Source/Drain could be misplaced.

So here comes Frankenpod







It works just fine, I disconnected it after a few seconds to avoid overheating the GC3355, but even the 7437 was hot. It will need active cooling with dissipator.



I have a gridseed pod with this exact chip burnt out, it's a MOSFET? I'm fairly handy with a hot-air gun/soldering iron, can you tell me what part to buy / where to get it so I can replace this frazzled chip please? Smiley

BTW the unit is still recognized as a USB device and seen by cpuminer, it just won't hash.

Thanks in advance for the info! Smiley
newbie
Activity: 5
Merit: 0
I don't suppose somebody could tell me what FB28 and FB29 are right next to the barrel connector? FB28 just blew on me, I have a friend who might be able to fix it for me and I'd like to be able to approach him halfway intelligently on this since pcb are not my forte.  Thanks.
hero member
Activity: 840
Merit: 1000
Y0 J4!

What's new?
Anything more on those chips and that Super Duper home made 500MH monster?

I'm going to repair my pod that has the blown FET on it. I believe it's possible that it was over stressed when installed - hence the pre-mature failure that shouldn't happen,ever! At least not to me dammit!

I am going to use my CRO to see what the PWM signal on the gate looks like before I operate.
I'll post some pics of that if you guys want.

Let me know.

Wolfey2014
Pod Master

He, still hard to get chips at a fair price.
There are "no" chips available for scrypt that meets my price.
For SHA256, BE200 and Avalon gen 3 could fit, I'll check the BOM.

For the pod and blades, summer will be the final test to see if the MOSFETs are cooled enough after over clocking.
I reversed the fan on my pods so the airflow is cooling the top of the MOSFETs. They should stand higher room temp this way.
sr. member
Activity: 378
Merit: 250
Y0 J4!

What's new?
Anything more on those chips and that Super Duper home made 500MH monster?

I'm going to repair my pod that has the blown FET on it. I believe it's possible that it was over stressed when installed - hence the pre-mature failure that shouldn't happen,ever! At least not to me dammit!

I am going to use my CRO to see what the PWM signal on the gate looks like before I operate.
I'll post some pics of that if you guys want.

Let me know.

Wolfey2014
Pod Master
hero member
Activity: 672
Merit: 500
http://fuk.io - check it out!
hero member
Activity: 672
Merit: 500
http://fuk.io - check it out!
sr. member
Activity: 346
Merit: 260
I am starting to wonder if it would just be easier to take the damn gc3355's off the boards, create a new pcb that can handle the power needed for SHA and start a password cracking service Huh  Grin

I haven't checked, but I'm not sure they are that efficient for sha256. The dedicated sha256 chips would probably work better if they could crack passwords. Not sure they can crack passwords, even if the algorithm is the same.
Undervolting the GC3355 for scrypt will still make it usable for some time, but it's clear that ROI will be hard if you haven't hit it yet. As usual, new coins, instamining, luck... Could help.

The scrypt asic market is currently moving too fast, and prices are dropping in the same way. We should have something more stable when 28nm will be widely available. prices for hardware will level around 6-7$/MH and after that, it will only slowly decrease.
It'll be the best time to buy bare chips and build custom hardware, since the prices won't be cut by 10 or 15% every week.

The market is moving so quick that by the time you test your design, order parts, build circuit boards and then populate you might
be too far behind in hash power to get a good ROI. You would have to push the limit and go for max hash power out of the gate
with the current 28n chips in the marketplace and not be spoon fed with little increments in hash power and less watts like the
manufactures currently do. I would think it would be a greater expense, and time especially if you tried to build yourself but if you
could fork out a bit more money for a pile of chips and go and create a 500Gh+ unit or boards that can array to create a larger
hashing powered unit (Similar to old ASIC blades). The only drawback is the higher wattage need to drive the current chips that
would require you to have multiple 15A services feeding the unit unless under voltage and max amount of chips are used to help
decrease the power draw.

I have my old blade running at 1khz & 41k on my pot and getting 6.8Mhs @ 154deg on the mosfet heat sink.
Will try to crank her up a bit on the weekend and see how much freq she can take.
sr. member
Activity: 412
Merit: 250
this is why buying scrypt asics is not profitable and it wont be anymore, but only building them by ourself
hero member
Activity: 840
Merit: 1000
I am starting to wonder if it would just be easier to take the damn gc3355's off the boards, create a new pcb that can handle the power needed for SHA and start a password cracking service Huh  Grin

I haven't checked, but I'm not sure they are that efficient for sha256. The dedicated sha256 chips would probably work better if they could crack passwords. Not sure they can crack passwords, even if the algorithm is the same.
Undervolting the GC3355 for scrypt will still make it usable for some time, but it's clear that ROI will be hard if you haven't hit it yet. As usual, new coins, instamining, luck... Could help.

The scrypt asic market is currently moving too fast, and prices are dropping in the same way. We should have something more stable when 28nm will be widely available. prices for hardware will level around 6-7$/MH and after that, it will only slowly decrease.
It'll be the best time to buy bare chips and build custom hardware, since the prices won't be cut by 10 or 15% every week.
newbie
Activity: 20
Merit: 0
I am starting to wonder if it would just be easier to take the damn gc3355's off the boards, create a new pcb that can handle the power needed for SHA and start a password cracking service Huh  Grin
Pages:
Jump to: