X6500 Custom FPGA Miner - page 18. | Bitcointalksearch.org

shackleford

full member

Activity: 281

Merit: 100

Quote from: li_gangyi on April 03, 2012, 02:08:55 PM

R2 and R8 are 0805 sized resistors. They are not terribly hard to solder, compared to say 0402 sized parts, but replacing R8 might be more of a challenge because of it's location.

A small soldering iron, good light, some soldering (not plumbing) flux, desoldering braid/wick and a pair of tweezers (failing that, a toothpick) should make short work of this.

Hi, sorry to be a pain but can you link me to the right resistor? There are some options and I want to be sure I am getting the right one. I checked Radioshacks website and looks like I am going to have to order them. Amazon has this http://www.amazon.com/Thin-Film-Tech-Precision-Continuous/dp/B0015DZIGC . Or would you guys just consider selling a few and shipping usps priority for some btc (would like to attempt this weekend)? Most the other stuff I have but I may need to get a smaller tip for the iron or upgrade.

I am at 0 invalids on the 200mhz firmware with copper and air but would love to push it a little after I get really good cooling so figured I might as well make this change. What kind of head room are you guys seeing?

li_gangyi

full member

Activity: 157

Merit: 100

R2 and R8 are 0805 sized resistors. They are not terribly hard to solder, compared to say 0402 sized parts, but replacing R8 might be more of a challenge because of it's location.

A small soldering iron, good light, some soldering (not plumbing) flux, desoldering braid/wick and a pair of tweezers (failing that, a toothpick) should make short work of this.

fizzisist

hero member

Activity: 720

Merit: 525

Quote from: shackleford on April 02, 2012, 03:54:26 AM

Quote from: fizzisist on March 29, 2012, 09:42:00 PM

In the last couple weeks, we've been experimenting with different FPGA core voltages (VCCINT). Our measurements have shown that on most rev3 boards, performance can be increased significantly by slightly increasing this voltage. Because of this, all rev3 boards starting today will be shipped with VCCINT set to 1.23V (instead of the original 1.20V). This applies to all boards shipped on or after 3/29/2012.

While this increases the board's power consumption slightly (data here), it also increases the potential limit on clock rate, which is especially useful with the new "overclocker" bitstreams.

If anyone would like to increase the core voltages on an older board, it simply means swapping two resistors (R2 and R8). You can either attempt this yourself or send it back to us and we'll do the work free of charge. We only ask that you pay the return shipping charge.

I am contemplating this… Just to be clear it is taking the resistor from the R2 spot and putting it in the R8 spot (both by the upright voltage regulators?) and vice versa. I think that is what you are saying but better safe than sorry. I don’t have a magnifying glass but it looks like it says 1911 (or 1161) on both of them and I did not see them in the schematic. They are pretty small and in a semi tight spot. If they were one of the bigger ones I would not hesitate to do it. I am going to do some research on how to solder this myself because the shipping and downtime sounds like a costly pain. Do you have any tips off hand for working with these tiny resistors? I may look to find a local place to do it if I can’t or just not rock the boat at all.

Sorry, "swapping" was a bad word choice. Those two resistors need to be replaced with 1.82k resistors (they are 1.91k). Unless you are getting a lot of invalids at higher clock rates, you should probably not make this change. It will reduce invalids (or increase the clock rate at which invalids start appearing), but if you have 0% invalids now, you can't get any lower than that!

shackleford

full member

Activity: 281

Merit: 100

Quote from: fizzisist on March 29, 2012, 09:42:00 PM

In the last couple weeks, we've been experimenting with different FPGA core voltages (VCCINT). Our measurements have shown that on most rev3 boards, performance can be increased significantly by slightly increasing this voltage. Because of this, all rev3 boards starting today will be shipped with VCCINT set to 1.23V (instead of the original 1.20V). This applies to all boards shipped on or after 3/29/2012.

While this increases the board's power consumption slightly (data here), it also increases the potential limit on clock rate, which is especially useful with the new "overclocker" bitstreams.

If anyone would like to increase the core voltages on an older board, it simply means swapping two resistors (R2 and R8). You can either attempt this yourself or send it back to us and we'll do the work free of charge. We only ask that you pay the return shipping charge.

I am contemplating this… Just to be clear it is taking the resistor from the R2 spot and putting it in the R8 spot (both by the upright voltage regulators?) and vice versa. I think that is what you are saying but better safe than sorry. I don’t have a magnifying glass but it looks like it says 1911 (or 1161) on both of them and I did not see them in the schematic. They are pretty small and in a semi tight spot. If they were one of the bigger ones I would not hesitate to do it. I am going to do some research on how to solder this myself because the shipping and downtime sounds like a costly pain. Do you have any tips off hand for working with these tiny resistors? I may look to find a local place to do it if I can’t or just not rock the boat at all.

NM found some good Youtubes on surface mount soldering .. still looks pretty small to me though.

Hawkix

hero member

Activity: 531

Merit: 505

OK, OK, next time I will explicitly append this: Wink

TheSeven

hero member

Activity: 504

Merit: 500

FPGA Mining LLC

Quote from: Hawkix on April 01, 2012, 03:49:38 PM

Quote from: fpgaminer on April 01, 2012, 06:27:46 AM

X8000 Sneak Peak

A Sneak Peak at the first 3MH/s/$ FPGA Miner!

X8000: Deep Purple Prototype:

Preliminary Details:

1.8 GH/s guaranteed, 2.4 GH/s typical.
$595.99 USD + Shipping/Tax.

Must be FAKE. All seems legit, but you made a mistake which reveals you - I never saw purple board from any suplier .. ![/list]

...even very popular boards:

BitBlitz

sr. member

Activity: 285

Merit: 250

Turning money into heat since 2011.

I almost spit coffee on my monitor! Too funny!

tgmarks

donator

Activity: 490

Merit: 500

You had me, I was almost ready to send in my pre-order

kakobrekla

hero member

Activity: 714

Merit: 500

Psi laju, karavani prolaze.

Quote from: Hawkix on April 01, 2012, 03:49:38 PM

Must be FAKE. All seems legit, but you made a mistake which reveals you - I never saw purple board from any suplier .. !

Hawkix

hero member

Activity: 531

Merit: 505

Quote from: fpgaminer on April 01, 2012, 06:27:46 AM

X8000 Sneak Peak

A Sneak Peak at the first 3MH/s/$ FPGA Miner!

X8000: Deep Purple Prototype:

Preliminary Details:

1.8 GH/s guaranteed, 2.4 GH/s typical.
$595.99 USD + Shipping/Tax.

Must be FAKE. All seems legit, but you made a mistake which reveals you - I never saw purple board from any suplier .. ![/list]

shackleford

full member

Activity: 281

Merit: 100

That was just mean.... Respect

bulanula

hero member

Activity: 518

Merit: 500

Almost got me a heart attack. A FPGA like that really will change the mining landscape.

Happy now that my GPUs are still going to be humming happily after all.

terrytibbs

hero member

Activity: 560

Merit: 501

You got me. I confess.

needbmw

legendary

Activity: 1302

Merit: 1008

Quote from: fpgaminer on April 01, 2012, 06:27:46 AM

following April 1st.

LOL

roomservice

full member

Activity: 199

Merit: 100

you almost got me haha

nedbert9

sr. member

Activity: 252

Merit: 250

Inactive

funny

fpgaminer

hero member

Activity: 560

Merit: 517

X8000 Sneak Peak

A Sneak Peak at the first 3MH/s/$ FPGA Miner!

Howdy guys. Fizzisist asked me to fill you guys in the upcoming design for the X8000, code named Deep Purple. This is the next revision of our board. It's a little ways off from being complete, since we plan to revamp the USB interface to handle the increased hashing performance. But I've got a picture from li_gangyi of the prototype:

X8000: Deep Purple Prototype:

Preliminary Details:

1.8 GH/s guaranteed, 2.4 GH/s typical.
$595.99 USD + Shipping/Tax.
Dual Virtex-6 FPGAs.
USB interface for configuration and mining communication.
Power supplied through 4-pin Molex connector or 5.5 x 2.1 mm barrel connector (wall wart style).
50 W power consumption measured at 2.2 GH/s (more measurements pending).
Independent protection fuses on each power connector.
JTAG available for development and debugging (headers are not populated by default).
Independent temperature monitoring for each FPGA.
Heatsink mounting holes, compatible with many heatsinks designed for the Northbridge chipset.
Additional clearance to allow for larger heatsinks.
3-pin headers to power 2 standard 12V fans.
Draws from the 12V supply at the Molex connector.
LEDs to indicate that the FPGAs are configured properly.
Mounting hole size diameter increased to allow for #6 size screws.
98 x 85.5 mm (3.86 x 3.37 inch) overall size

Again, this is just a prototype, but we're hoping to cram a few more features in there in the weeks following April 1st.

fizzisist

hero member

Activity: 720

Merit: 525

Quote from: freshzive on April 01, 2012, 12:46:03 AM

Quote from: TheSeven on March 31, 2012, 10:02:07 PM

Quote from: freshzive on March 31, 2012, 09:40:45 PM

wow, didn't realize there was a fizzisist fork of the modular miner....runs much more smoothly, way fewer stales on gpumax.

cheers Cheesy

I'm slightly puzzled...

Fizzisist's fork only has some minor overclocking-related changes, and is based on a discontinued version?

Yup. After running for a while, stales went back up... Guess it was just some early luck. Oh well.

Heh, yeah, that would have been really odd...

TheSeven's master branch should really have all the features of mine now, and have the awesome web UI, so I strongly recommend that.

freshzive

sr. member

Activity: 447

Merit: 250

Quote from: TheSeven on March 31, 2012, 10:02:07 PM

Quote from: freshzive on March 31, 2012, 09:40:45 PM

wow, didn't realize there was a fizzisist fork of the modular miner....runs much more smoothly, way fewer stales on gpumax.

cheers Cheesy

I'm slightly puzzled...

Fizzisist's fork only has some minor overclocking-related changes, and is based on a discontinued version?

Yup. After running for a while, stales went back up... Guess it was just some early luck. Oh well.

TheSeven

hero member

Activity: 504

Merit: 500

FPGA Mining LLC

Quote from: freshzive on March 31, 2012, 09:40:45 PM

wow, didn't realize there was a fizzisist fork of the modular miner....runs much more smoothly, way fewer stales on gpumax.

cheers Cheesy

I'm slightly puzzled...

Fizzisist's fork only has some minor overclocking-related changes, and is based on a discontinued version?

Topic: X6500 Custom FPGA Miner - page 18. (Read 219829 times)