Block Erupter USB - Overclocking/ hacking ?

[ooxtcoo]

Had a BE go bad recently.  Troubleshooting: Not communicating; looked at voltages on the edge connector and found 7 to 6 on a good unit showing a little over a volt but on the faulty unit over 3 volts; cleaned a spot between a couple of pins on the inside side of the BE chip and now have the same voltage between edge pins 7 and 6 but not communicating.  The LED was flashing the correct sequence and duration so it wasn't the CP2102->crystal but found both the SUSPEND and SUSPEND-NOT pins high.  They should be complementary.  Not shorted so internal or elsewhere on the board.  Tried a manual reset of the CP2102 bringing the reset pin to ground but that was no help either.  The CP2102 seems to have the proper voltages aside from the SUSPEND and SUSPEND-NOT pins.  Since they're the states of an inverter input and output it's disallowed logic.  Anyone seen this and know of a fix aside from replacing the CP2102?  I note I had been getting strange high frequency noise on my network, it stopped around the time this went belly up.

first:
did you check the usb connector solder pins?
i had many of them with bad soldered usb connector pins...
try to resolder this 4 pins first.

or:
desolder the asic and check if it is detected.
is cp2102 getting hot?
if you desolder asic and it is detected again, you know your asic is demaged.

the cp2102 works without any external components.
i mean, if you desolder every component but keep only the cp2102 on board, it must be detected by your computer.

if not, maybe its bad soldered but mostly death cp2102.

i have some new cp2102 if you need.

Thanks for the reply.  I've inspected the board closely with a loupe  The USB pins seem quite well soldered visually inspecting.  The ASIC seems to be just a little askew from dead center but I'm betting on the CP2102 as dead - ordered one from Digikey earlier today.  I replaced one once before and it took me two days to get it right so I know they can tolerate soldering and desoldering, not sure about my tolerance for difficult work without the best (expensive) tools however. 

I've measured continuity between the data pins and the corresponding rear surface mount pins on the USB, that earlier today, but perhaps the signals just aren't reaching the fingers in the USB plug.  Think I'll grab a USB extension, plug it in and measure....yes, continuity.  It seems to be in a suspend state according to the suspend pin out (12) although the reverse is indicated by the suspend-not pin out (12) as they are both high.  I believe it wouldn't be capable of identifying itself if in a suspended state.  A good unit immediately gets a device reset notation in linux to std out when plugged in.  I would expect to see a similar notation when I bring the reset pin to ground but nothing. 

from where are you?
i can solder it for you in 1 minute.

get a cheap hotairgun, with it its easy to desolder and solder.
i would not know how to solder the cp2102 without hotair or irda.

and resolder the usb pins first, sometimes you cant see visualy bad contacts.
i had about 10 erupters with bad solderes usb pins.

same with asic blade..
i have ordered 2 of them, one had a problem with lan led, this was also a bad
soldered contact.
it seems to be a problem with bigger components, on these often 1 pin loose contact.



@all
what oszillator pinout is used in the erupter blade?
did someone overclocked the new blades which have only 1 oszillator soldered on board?

[sidehack]

AM Blade V2 overclock kits, and link to instruction page with more info.

https://bitcointalk.org/index.php?topic=310752.0

[qtronix]

i had some aluminum plates cut to use as heat sinks there 50x50x 3mm thick. think i am going to mod all mine to the max and just let them run until they burn out 

[/URL]

[soy]

This last was my second to go bad that worked well after replacing the CP2102.

Picked up a CP2102 from my PO box yesterday and this morning sat down to replace the chip. The eyeglass gremlin has been at work and my best fine work eyeglasses are misplaced. The replacement went well, first attempt after removing the old, the IC fell into place and I tried heating it where it fell. I had fluxed the edges but had forgotten to flux the center pad. Heated for 5 minutes with the hot air machine, let cool and inspection showed alignment was fine but one side was high off the board. This time I used the clothespin with a thru screw and it was particularly easy given more than half the legs were already soldered. Heated, let cool, tested and the CP2102 was reporting it was there. Cleaned well with brake cleaner then blew off with compressed air, installed the heatsink and it's now hashing away at AMU 5: | 334.4M/325.2Mh/s | DA: 32 DR: 0 HW: 0 WU: 4.72/m

[soy]

Had a BE go bad recently.  Troubleshooting: Not communicating; looked at voltages on the edge connector and found 7 to 6 on a good unit showing a little over a volt but on the faulty unit over 3 volts; cleaned a spot between a couple of pins on the inside side of the BE chip and now have the same voltage between edge pins 7 and 6 but not communicating.  The LED was flashing the correct sequence and duration so it wasn't the CP2102->crystal but found both the SUSPEND and SUSPEND-NOT pins high.  They should be complementary.  Not shorted so internal or elsewhere on the board.  Tried a manual reset of the CP2102 bringing the reset pin to ground but that was no help either.  The CP2102 seems to have the proper voltages aside from the SUSPEND and SUSPEND-NOT pins.  Since they're the states of an inverter input and output it's disallowed logic.  Anyone seen this and know of a fix aside from replacing the CP2102?  I note I had been getting strange high frequency noise on my network, it stopped around the time this went belly up.

first:
did you check the usb connector solder pins?
i had many of them with bad soldered usb connector pins...
try to resolder this 4 pins first.

or:
desolder the asic and check if it is detected.
is cp2102 getting hot?
if you desolder asic and it is detected again, you know your asic is demaged.

the cp2102 works without any external components.
i mean, if you desolder every component but keep only the cp2102 on board, it must be detected by your computer.

if not, maybe its bad soldered but mostly death cp2102.

i have some new cp2102 if you need.

Thanks for the reply.  I've inspected the board closely with a loupe  The USB pins seem quite well soldered visually inspecting.  The ASIC seems to be just a little askew from dead center but I'm betting on the CP2102 as dead - ordered one from Digikey earlier today.  I replaced one once before and it took me two days to get it right so I know they can tolerate soldering and desoldering, not sure about my tolerance for difficult work without the best (expensive) tools however. 

I've measured continuity between the data pins and the corresponding rear surface mount pins on the USB, that earlier today, but perhaps the signals just aren't reaching the fingers in the USB plug.  Think I'll grab a USB extension, plug it in and measure....yes, continuity.  It seems to be in a suspend state according to the suspend pin out (12) although the reverse is indicated by the suspend-not pin out (12) as they are both high.  I believe it wouldn't be capable of identifying itself if in a suspended state.  A good unit immediately gets a device reset notation in linux to std out when plugged in.  I would expect to see a similar notation when I bring the reset pin to ground but nothing. 

from where are you?
i can solder it for you in 1 minute.

get a cheap hotairgun, with it its easy to desolder and solder.
i would not know how to solder the cp2102 without hotair or irda.

and resolder the usb pins first, sometimes you cant see visualy bad contacts.
i had about 10 erupters with bad solderes usb pins.

same with asic blade..
i have ordered 2 of them, one had a problem with lan led, this was also a bad
soldered contact.
it seems to be a problem with bigger components, on these often 1 pin loose contact.

@all
what oszillator pinout is used in the erupter blade?
did someone overclocked the new blades which have only 1 oszillator soldered on board?

I live in the US state of Georgia.  Did smt board repair for Yokogawa for a couple of years recently.  Now retired.  Have just gotten it working this morning: AMU 5:|334.4M/325.2Mh/s | DA: 32 DR: 0 HW: 0 WU: 4.72/m.  And yourself?

[soy]

Had a BE go bad recently.  Troubleshooting: Not communicating; looked at voltages on the edge connector and found 7 to 6 on a good unit showing a little over a volt but on the faulty unit over 3 volts; cleaned a spot between a couple of pins on the inside side of the BE chip and now have the same voltage between edge pins 7 and 6 but not communicating.  The LED was flashing the correct sequence and duration so it wasn't the CP2102->crystal but found both the SUSPEND and SUSPEND-NOT pins high.  They should be complementary.  Not shorted so internal or elsewhere on the board.  Tried a manual reset of the CP2102 bringing the reset pin to ground but that was no help either.  The CP2102 seems to have the proper voltages aside from the SUSPEND and SUSPEND-NOT pins.  Since they're the states of an inverter input and output it's disallowed logic.  Anyone seen this and know of a fix aside from replacing the CP2102?  I note I had been getting strange high frequency noise on my network, it stopped around the time this went belly up.

first:
did you check the usb connector solder pins?
i had many of them with bad soldered usb connector pins...
try to resolder this 4 pins first.

or:
desolder the asic and check if it is detected.
is cp2102 getting hot?
if you desolder asic and it is detected again, you know your asic is demaged.

the cp2102 works without any external components.
i mean, if you desolder every component but keep only the cp2102 on board, it must be detected by your computer.

if not, maybe its bad soldered but mostly death cp2102.

i have some new cp2102 if you need.

Thanks for the reply.  I've inspected the board closely with a loupe  The USB pins seem quite well soldered visually inspecting.  The ASIC seems to be just a little askew from dead center but I'm betting on the CP2102 as dead - ordered one from Digikey earlier today.  I replaced one once before and it took me two days to get it right so I know they can tolerate soldering and desoldering, not sure about my tolerance for difficult work without the best (expensive) tools however. 

I've measured continuity between the data pins and the corresponding rear surface mount pins on the USB, that earlier today, but perhaps the signals just aren't reaching the fingers in the USB plug.  Think I'll grab a USB extension, plug it in and measure....yes, continuity.  It seems to be in a suspend state according to the suspend pin out (12) although the reverse is indicated by the suspend-not pin out (12) as they are both high.  I believe it wouldn't be capable of identifying itself if in a suspended state.  A good unit immediately gets a device reset notation in linux to std out when plugged in.  I would expect to see a similar notation when I bring the reset pin to ground but nothing. 

from where are you?
i can solder it for you in 1 minute.

get a cheap hotairgun, with it its easy to desolder and solder.
i would not know how to solder the cp2102 without hotair or irda.

and resolder the usb pins first, sometimes you cant see visualy bad contacts.
i had about 10 erupters with bad solderes usb pins.

same with asic blade..
i have ordered 2 of them, one had a problem with lan led, this was also a bad
soldered contact.
it seems to be a problem with bigger components, on these often 1 pin loose contact.

@all
what oszillator pinout is used in the erupter blade?
did someone overclocked the new blades which have only 1 oszillator soldered on board?

I live in the US state of Georgia.  Did smt board repair for Yokogawa for a couple of years recently.  Now retired.  Have just gotten it working this morning: AMU 5:|334.4M/325.2Mh/s | DA: 32 DR: 0 HW: 0 WU: 4.72/m.  And yourself?

Ah well.  Ran long enough to produce 690 Accepted shares then went Zombie.

[soy]

Had a BE go bad recently.  Troubleshooting: Not communicating; looked at voltages on the edge connector and found 7 to 6 on a good unit showing a little over a volt but on the faulty unit over 3 volts; cleaned a spot between a couple of pins on the inside side of the BE chip and now have the same voltage between edge pins 7 and 6 but not communicating.  The LED was flashing the correct sequence and duration so it wasn't the CP2102->crystal but found both the SUSPEND and SUSPEND-NOT pins high.  They should be complementary.  Not shorted so internal or elsewhere on the board.  Tried a manual reset of the CP2102 bringing the reset pin to ground but that was no help either.  The CP2102 seems to have the proper voltages aside from the SUSPEND and SUSPEND-NOT pins.  Since they're the states of an inverter input and output it's disallowed logic.  Anyone seen this and know of a fix aside from replacing the CP2102?  I note I had been getting strange high frequency noise on my network, it stopped around the time this went belly up.

first:
did you check the usb connector solder pins?
i had many of them with bad soldered usb connector pins...
try to resolder this 4 pins first.

or:
desolder the asic and check if it is detected.
is cp2102 getting hot?
if you desolder asic and it is detected again, you know your asic is demaged.

the cp2102 works without any external components.
i mean, if you desolder every component but keep only the cp2102 on board, it must be detected by your computer.

if not, maybe its bad soldered but mostly death cp2102.

i have some new cp2102 if you need.

Thanks for the reply.  I've inspected the board closely with a loupe  The USB pins seem quite well soldered visually inspecting.  The ASIC seems to be just a little askew from dead center but I'm betting on the CP2102 as dead - ordered one from Digikey earlier today.  I replaced one once before and it took me two days to get it right so I know they can tolerate soldering and desoldering, not sure about my tolerance for difficult work without the best (expensive) tools however. 

I've measured continuity between the data pins and the corresponding rear surface mount pins on the USB, that earlier today, but perhaps the signals just aren't reaching the fingers in the USB plug.  Think I'll grab a USB extension, plug it in and measure....yes, continuity.  It seems to be in a suspend state according to the suspend pin out (12) although the reverse is indicated by the suspend-not pin out (12) as they are both high.  I believe it wouldn't be capable of identifying itself if in a suspended state.  A good unit immediately gets a device reset notation in linux to std out when plugged in.  I would expect to see a similar notation when I bring the reset pin to ground but nothing. 

from where are you?
i can solder it for you in 1 minute.

get a cheap hotairgun, with it its easy to desolder and solder.
i would not know how to solder the cp2102 without hotair or irda.

and resolder the usb pins first, sometimes you cant see visualy bad contacts.
i had about 10 erupters with bad solderes usb pins.

same with asic blade..
i have ordered 2 of them, one had a problem with lan led, this was also a bad
soldered contact.
it seems to be a problem with bigger components, on these often 1 pin loose contact.

@all
what oszillator pinout is used in the erupter blade?
did someone overclocked the new blades which have only 1 oszillator soldered on board?

I live in the US state of Georgia.  Did smt board repair for Yokogawa for a couple of years recently.  Now retired.  Have just gotten it working this morning: AMU 5:|334.4M/325.2Mh/s | DA: 32 DR: 0 HW: 0 WU: 4.72/m.  And yourself?

Ah well.  Ran long enough to produce 690 Accepted shares then went Zombie.

Either brake cleaner was too strong a chemical for the plastic screws or I had over-torqued.  I cleaned the CP2102 again with the brake cleaner, blew it off/dry, replaced the screws with I think 3mm steel screws, plugged it in and it's hashing again.
AMU 7:| 333.9M/318.8Mh/s | DA: 24 DR:0 HW: 0 WU: 6.75/m  now back to cutting insulation for my well house so my pump doesn't freeze this winter...


 

[soy]

Had a BE go bad recently.  Troubleshooting: Not communicating; looked at voltages on the edge connector and found 7 to 6 on a good unit showing a little over a volt but on the faulty unit over 3 volts; cleaned a spot between a couple of pins on the inside side of the BE chip and now have the same voltage between edge pins 7 and 6 but not communicating.  The LED was flashing the correct sequence and duration so it wasn't the CP2102->crystal but found both the SUSPEND and SUSPEND-NOT pins high.  They should be complementary.  Not shorted so internal or elsewhere on the board.  Tried a manual reset of the CP2102 bringing the reset pin to ground but that was no help either.  The CP2102 seems to have the proper voltages aside from the SUSPEND and SUSPEND-NOT pins.  Since they're the states of an inverter input and output it's disallowed logic.  Anyone seen this and know of a fix aside from replacing the CP2102?  I note I had been getting strange high frequency noise on my network, it stopped around the time this went belly up.

first:
did you check the usb connector solder pins?
i had many of them with bad soldered usb connector pins...
try to resolder this 4 pins first.
or:
desolder the asic and check if it is detected.
is cp2102 getting hot?
if you desolder asic and it is detected again, you know your asic is demaged.
the cp2102 works without any external components.
i mean, if you desolder every component but keep only the cp2102 on board, it must be detected by your computer.
if not, maybe its bad soldered but mostly death cp2102.
i have some new cp2102 if you need.

Thanks for the reply.  I've inspected the board closely with a loupe  The USB pins seem quite well soldered visually inspecting.  The ASIC seems to be just a little askew from dead center but I'm betting on the CP2102 as dead - ordered one from Digikey earlier today.  I replaced one once before and it took me two days to get it right so I know they can tolerate soldering and desoldering, not sure about my tolerance for difficult work without the best (expensive) tools however. 
I've measured continuity between the data pins and the corresponding rear surface mount pins on the USB, that earlier today, but perhaps the signals just aren't reaching the fingers in the USB plug.  Think I'll grab a USB extension, plug it in and measure....yes, continuity.  It seems to be in a suspend state according to the suspend pin out (12) although the reverse is indicated by the suspend-not pin out (12) as they are both high.  I believe it wouldn't be capable of identifying itself if in a suspended state.  A good unit immediately gets a device reset notation in linux to std out when plugged in.  I would expect to see a similar notation when I bring the reset pin to ground but nothing. 

from where are you?
i can solder it for you in 1 minute.
get a cheap hotairgun, with it its easy to desolder and solder.
i would not know how to solder the cp2102 without hotair or irda.
and resolder the usb pins first, sometimes you cant see visualy bad contacts.
i had about 10 erupters with bad solderes usb pins.
same with asic blade..
i have ordered 2 of them, one had a problem with lan led, this was also a bad
soldered contact.
it seems to be a problem with bigger components, on these often 1 pin loose contact.
@all
what oszillator pinout is used in the erupter blade?
did someone overclocked the new blades which have only 1 oszillator soldered on board?

I live in the US state of Georgia.  Did smt board repair for Yokogawa for a couple of years recently.  Now retired.  Have just gotten it working this morning: AMU 5:|334.4M/325.2Mh/s | DA: 32 DR: 0 HW: 0 WU: 4.72/m.  And yourself?

Ah well.  Ran long enough to produce 690 Accepted shares then went Zombie.

Either brake cleaner was too strong a chemical for the plastic screws or I had over-torqued.  I cleaned the CP2102 again with the brake cleaner, blew it off/dry, replaced the screws with I think 3mm steel screws, plugged it in and it's hashing again.
AMU 7:| 333.9M/318.8Mh/s | DA: 24 DR:0 HW: 0 WU: 6.75/m  now back to cutting insulation for my well house so my pump doesn't freeze this winter...

Zombie again.

[qtronix]

anyone no the thread size for the screws i need to order a tap.
thanks

[soy]

anyone no the thread size for the screws i need to order a tap.
thanks

M3-.50

[soy]

anyone no the thread size for the screws i need to order a tap.
thanks

M3-.50

I've use M3-.50x6 but ran into a problem with one that seemed to clear when I went back to plastic.  I can't imagine a multi-layer layout that would threaten shorting anything but ground planes at screw thru holes but who knows.

[qtronix]

thanks guys saved me a lot of messing about.

[soy]

Had a BE go bad recently.  Troubleshooting: Not communicating; looked at voltages on the edge connector and found 7 to 6 on a good unit showing a little over a volt but on the faulty unit over 3 volts; cleaned a spot between a couple of pins on the inside side of the BE chip and now have the same voltage between edge pins 7 and 6 but not communicating.  The LED was flashing the correct sequence and duration so it wasn't the CP2102->crystal but found both the SUSPEND and SUSPEND-NOT pins high.  They should be complementary.  Not shorted so internal or elsewhere on the board.  Tried a manual reset of the CP2102 bringing the reset pin to ground but that was no help either.  The CP2102 seems to have the proper voltages aside from the SUSPEND and SUSPEND-NOT pins.  Since they're the states of an inverter input and output it's disallowed logic.  Anyone seen this and know of a fix aside from replacing the CP2102?  I note I had been getting strange high frequency noise on my network, it stopped around the time this went belly up.

first:
did you check the usb connector solder pins?
i had many of them with bad soldered usb connector pins...
try to resolder this 4 pins first.
or:
desolder the asic and check if it is detected.
is cp2102 getting hot?
if you desolder asic and it is detected again, you know your asic is demaged.
the cp2102 works without any external components.
i mean, if you desolder every component but keep only the cp2102 on board, it must be detected by your computer.
if not, maybe its bad soldered but mostly death cp2102.
i have some new cp2102 if you need.

Thanks for the reply.  I've inspected the board closely with a loupe  The USB pins seem quite well soldered visually inspecting.  The ASIC seems to be just a little askew from dead center but I'm betting on the CP2102 as dead - ordered one from Digikey earlier today.  I replaced one once before and it took me two days to get it right so I know they can tolerate soldering and desoldering, not sure about my tolerance for difficult work without the best (expensive) tools however.  
I've measured continuity between the data pins and the corresponding rear surface mount pins on the USB, that earlier today, but perhaps the signals just aren't reaching the fingers in the USB plug.  Think I'll grab a USB extension, plug it in and measure....yes, continuity.  It seems to be in a suspend state according to the suspend pin out (12) although the reverse is indicated by the suspend-not pin out (12) as they are both high.  I believe it wouldn't be capable of identifying itself if in a suspended state.  A good unit immediately gets a device reset notation in linux to std out when plugged in.  I would expect to see a similar notation when I bring the reset pin to ground but nothing.  

from where are you?
i can solder it for you in 1 minute.
get a cheap hotairgun, with it its easy to desolder and solder.
i would not know how to solder the cp2102 without hotair or irda.
and resolder the usb pins first, sometimes you cant see visualy bad contacts.
i had about 10 erupters with bad solderes usb pins.
same with asic blade..
i have ordered 2 of them, one had a problem with lan led, this was also a bad
soldered contact.
it seems to be a problem with bigger components, on these often 1 pin loose contact.
@all
what oszillator pinout is used in the erupter blade?
did someone overclocked the new blades which have only 1 oszillator soldered on board?

I live in the US state of Georgia.  Did smt board repair for Yokogawa for a couple of years recently.  Now retired.  Have just gotten it working this morning: AMU 5:|334.4M/325.2Mh/s | DA: 32 DR: 0 HW: 0 WU: 4.72/m.  And yourself?

Ah well.  Ran long enough to produce 690 Accepted shares then went Zombie.

Either brake cleaner was too strong a chemical for the plastic screws or I had over-torqued.  I cleaned the CP2102 again with the brake cleaner, blew it off/dry, replaced the screws with I think 3mm steel screws, plugged it in and it's hashing again.
AMU 7:| 333.9M/318.8Mh/s | DA: 24 DR:0 HW: 0 WU: 6.75/m  now back to cutting insulation for my well house so my pump doesn't freeze this winter...

Zombie again.

Took the CP2102 off once again.  When I last removed and replaced, some of the board fingers along the lower edge (15-21), the unused string of pins, had come free and one had become embedded in the heatsinking solder on the main pad causing it not to sit fully flat.  Cleaned it off and it's back in service once again.  Wonder if it will fail or not.  (....up a couple of hours and okay)

[Zeek_W]

An update - my UV LED swap was a success! I'd post a pic but damn I can't seem to get a good picture. I'm happy though.

[ooxtcoo]

Had a BE go bad recently.  Troubleshooting: Not communicating; looked at voltages on the edge connector and found 7 to 6 on a good unit showing a little over a volt but on the faulty unit over 3 volts; cleaned a spot between a couple of pins on the inside side of the BE chip and now have the same voltage between edge pins 7 and 6 but not communicating.  The LED was flashing the correct sequence and duration so it wasn't the CP2102->crystal but found both the SUSPEND and SUSPEND-NOT pins high.  They should be complementary.  Not shorted so internal or elsewhere on the board.  Tried a manual reset of the CP2102 bringing the reset pin to ground but that was no help either.  The CP2102 seems to have the proper voltages aside from the SUSPEND and SUSPEND-NOT pins.  Since they're the states of an inverter input and output it's disallowed logic.  Anyone seen this and know of a fix aside from replacing the CP2102?  I note I had been getting strange high frequency noise on my network, it stopped around the time this went belly up.

first:
did you check the usb connector solder pins?
i had many of them with bad soldered usb connector pins...
try to resolder this 4 pins first.

or:
desolder the asic and check if it is detected.
is cp2102 getting hot?
if you desolder asic and it is detected again, you know your asic is demaged.

the cp2102 works without any external components.
i mean, if you desolder every component but keep only the cp2102 on board, it must be detected by your computer.

if not, maybe its bad soldered but mostly death cp2102.

i have some new cp2102 if you need.

Thanks for the reply.  I've inspected the board closely with a loupe  The USB pins seem quite well soldered visually inspecting.  The ASIC seems to be just a little askew from dead center but I'm betting on the CP2102 as dead - ordered one from Digikey earlier today.  I replaced one once before and it took me two days to get it right so I know they can tolerate soldering and desoldering, not sure about my tolerance for difficult work without the best (expensive) tools however.  

I've measured continuity between the data pins and the corresponding rear surface mount pins on the USB, that earlier today, but perhaps the signals just aren't reaching the fingers in the USB plug.  Think I'll grab a USB extension, plug it in and measure....yes, continuity.  It seems to be in a suspend state according to the suspend pin out (12) although the reverse is indicated by the suspend-not pin out (12) as they are both high.  I believe it wouldn't be capable of identifying itself if in a suspended state.  A good unit immediately gets a device reset notation in linux to std out when plugged in.  I would expect to see a similar notation when I bring the reset pin to ground but nothing.  

from where are you?
i can solder it for you in 1 minute.

get a cheap hotairgun, with it its easy to desolder and solder.
i would not know how to solder the cp2102 without hotair or irda.

and resolder the usb pins first, sometimes you cant see visualy bad contacts.
i had about 10 erupters with bad solderes usb pins.

same with asic blade..
i have ordered 2 of them, one had a problem with lan led, this was also a bad
soldered contact.
it seems to be a problem with bigger components, on these often 1 pin loose contact.

@all
what oszillator pinout is used in the erupter blade?
did someone overclocked the new blades which have only 1 oszillator soldered on board?

I live in the US state of Georgia.  Did smt board repair for Yokogawa for a couple of years recently.  Now retired.  Have just gotten it working this morning: AMU 5:|334.4M/325.2Mh/s | DA: 32 DR: 0 HW: 0 WU: 4.72/m.  And yourself?

from austria..

but good, you solved it.
hope it not brake again.


for any reason i dont know, on 2 of my block erupter blades, 16 asics died on same day!

i dont know why, but i soldered the asics from by block erupter usb into the blades, now my blades working again, but i lost 16 usb erupters now.

so i have many spare parts for usb erupters...

cp2102, dc dc converter, attiny, mainboards, pcb itself, flip flops, and so on...


so if someone have demaged usb erupters or asic blades where some asics are ok, please let me know...
i buy it...

why friedcat dont sell asics alone too?



btw. now i made some test with my block erupter blades...
i soldered to 16 mhz oszillator (used in usb erupter overclocking) into the asic blades (was a bit tricky, cause the oszillator is smaller) and i soldered the 8 resistors in place, and it seems to work, but only short time.

it starts hashin, go up to 8 gh/s but then it go slow down again...
i think i need a better cooling now....
as i dont have any free fans anymore, i soldered all back for now and run it on stock...

if now some asics die, i dont have mutch reserve anymore...

now i only have about 8 usb erupters working... rest is without working asics...



here are photos of the asic blade with 16mhz usb erupter oszillator:





here is my modded usb hub (since this mode, i can connect more usb erupter to my hub):

[sidehack]

Yeah up around 16MHz blades get really particular about staying cool, chipside heatsinks and good CFM recommended. They'll run past 14 on stock cooling though.

Nuts on all the defunct BEs. There's some folks around here that are using the stock heatsinks for various projects though, if you want to make a few bucks off of 'em still.

[qtronix]

Had a BE go bad recently.  Troubleshooting: Not communicating; looked at voltages on the edge connector and found 7 to 6 on a good unit showing a little over a volt but on the faulty unit over 3 volts; cleaned a spot between a couple of pins on the inside side of the BE chip and now have the same voltage between edge pins 7 and 6 but not communicating.  The LED was flashing the correct sequence and duration so it wasn't the CP2102->crystal but found both the SUSPEND and SUSPEND-NOT pins high.  They should be complementary.  Not shorted so internal or elsewhere on the board.  Tried a manual reset of the CP2102 bringing the reset pin to ground but that was no help either.  The CP2102 seems to have the proper voltages aside from the SUSPEND and SUSPEND-NOT pins.  Since they're the states of an inverter input and output it's disallowed logic.  Anyone seen this and know of a fix aside from replacing the CP2102?  I note I had been getting strange high frequency noise on my network, it stopped around the time this went belly up.

first:
did you check the usb connector solder pins?
i had many of them with bad soldered usb connector pins...
try to resolder this 4 pins first.

or:
desolder the asic and check if it is detected.
is cp2102 getting hot?
if you desolder asic and it is detected again, you know your asic is demaged.

the cp2102 works without any external components.
i mean, if you desolder every component but keep only the cp2102 on board, it must be detected by your computer.

if not, maybe its bad soldered but mostly death cp2102.

i have some new cp2102 if you need.

Thanks for the reply.  I've inspected the board closely with a loupe  The USB pins seem quite well soldered visually inspecting.  The ASIC seems to be just a little askew from dead center but I'm betting on the CP2102 as dead - ordered one from Digikey earlier today.  I replaced one once before and it took me two days to get it right so I know they can tolerate soldering and desoldering, not sure about my tolerance for difficult work without the best (expensive) tools however.  

I've measured continuity between the data pins and the corresponding rear surface mount pins on the USB, that earlier today, but perhaps the signals just aren't reaching the fingers in the USB plug.  Think I'll grab a USB extension, plug it in and measure....yes, continuity.  It seems to be in a suspend state according to the suspend pin out (12) although the reverse is indicated by the suspend-not pin out (12) as they are both high.  I believe it wouldn't be capable of identifying itself if in a suspended state.  A good unit immediately gets a device reset notation in linux to std out when plugged in.  I would expect to see a similar notation when I bring the reset pin to ground but nothing.  

from where are you?
i can solder it for you in 1 minute.

get a cheap hotairgun, with it its easy to desolder and solder.
i would not know how to solder the cp2102 without hotair or irda.

and resolder the usb pins first, sometimes you cant see visualy bad contacts.
i had about 10 erupters with bad solderes usb pins.

same with asic blade..
i have ordered 2 of them, one had a problem with lan led, this was also a bad
soldered contact.
it seems to be a problem with bigger components, on these often 1 pin loose contact.

@all
what oszillator pinout is used in the erupter blade?
did someone overclocked the new blades which have only 1 oszillator soldered on board?

I live in the US state of Georgia.  Did smt board repair for Yokogawa for a couple of years recently.  Now retired.  Have just gotten it working this morning: AMU 5:|334.4M/325.2Mh/s | DA: 32 DR: 0 HW: 0 WU: 4.72/m.  And yourself?

from austria..

but good, you solved it.
hope it not brake again.


for any reason i dont know, on 2 of my block erupter blades, 16 asics died on same day!

i dont know why, but i soldered the asics from by block erupter usb into the blades, now my blades working again, but i lost 16 usb erupters now.

so i have many spare parts for usb erupters...

cp2102, dc dc converter, attiny, mainboards, pcb itself, flip flops, and so on...


so if someone have demaged usb erupters or asic blades where some asics are ok, please let me know...
i buy it...

why friedcat dont sell asics alone too?



btw. now i made some test with my block erupter blades...
i soldered to 16 mhz oszillator (used in usb erupter overclocking) into the asic blades (was a bit tricky, cause the oszillator is smaller) and i soldered the 8 resistors in place, and it seems to work, but only short time.

it starts hashin, go up to 8 gh/s but then it go slow down again...
i think i need a better cooling now....
as i dont have any free fans anymore, i soldered all back for now and run it on stock...

if now some asics die, i dont have mutch reserve anymore...

now i only have about 8 usb erupters working... rest is without working asics...



here are photos of the asic blade with 16mhz usb erupter oszillator:





here is my modded usb hub (since this mode, i can connect more usb erupter to my hub):

i have a blade with 10 avlons on if its any good to you, the power input on the blade has faults but the asic works, i am in the uk.

[ooxtcoo]

yes, iam interessted in your blade, how mutch you like for it incl. ship to austria?
should i send you pm?

but you mean BE100 asics not avalon or?
you have asic miner blade or avalone blade?

[luketanti]

Hi again. I swapped the crystal to 14.75MHz and added --icarus-timing short to the command but the BE still mines at an average of 365MH/s It is supposed to reach about 410MH/s.
This is my command in Linux
screen sudo ./cgminer-3.4.2/cgminer --icarus-timing short --config /home/luke/cgminer.conf
Is it OK or should I right it after the config?

Can any one help me please?

[kr105]

Hi again. I swapped the crystal to 14.75MHz and added --icarus-timing short to the command but the BE still mines at an average of 365MH/s It is supposed to reach about 410MH/s.
This is my command in Linux
screen sudo ./cgminer-3.4.2/cgminer --icarus-timing short --config /home/luke/cgminer.conf
Is it OK or should I right it after the config?

Can any one help me please?

Use --icarus-timing short=90 instead of --icarus-timing short.