Pages:
Author

Topic: Raspberry pi usb device limitation? (Read 15120 times)

hero member
Activity: 784
Merit: 500
December 02, 2012, 03:38:01 PM
#81
Bridging polyfuses may not be necessary with later revisions, unless input amperage needs to be raised.

http://elinux.org/Polyfuses_explained
Quote
As of the end of August 2012 some users were reporting that newly arrived PI's were fitted with 0 Ohm resistors for the USB polyfuses F1 & F2, and Liz of the RPF foundation confirmed that it was official RPF policy to remove the offending fuses. However F3 still limits the total input current to about 1A.

No F1 or F2 on my RPi only F3. I'm hoping I can get it to work with my incoming asics. Certified powered hub is next on my list.
It arrived this last week.
Do you have anything else to stress test the USB stack with?

Is anyone in general aware if a powered hub fixes the rest of the problems?

As for my tests a powered hub does not fix the issues with usb..... It could cause problems too. Remember that thing hates feedback power over the USB cable. Only cutting the  power lanes of the USB cable helped a little. Still the device is not an option for me with my 18 Ztex boards...

Are your boards bus powered?

No Ztex Boards geht their power through a barrel plug. I dont know why the PI wont like my boards..... not a single one without a hub is a lasting solution.
sr. member
Activity: 406
Merit: 250
LTC
November 26, 2012, 08:37:20 PM
#80
i should be finally getting mines in the mail tomorrow. i probably shouldn't use a microusb + usb to wall adapter  as my power source huh?
If the usb to wall adapter supplies 700mAh at least, it should be ok (according to the producer).
full member
Activity: 210
Merit: 100
November 26, 2012, 08:24:05 PM
#79
i should be finally getting mines in the mail tomorrow. i probably shouldn't use a microusb + usb to wall adapter  as my power source huh?
sr. member
Activity: 406
Merit: 250
LTC
November 26, 2012, 08:22:43 PM
#78
Bridging polyfuses may not be necessary with later revisions, unless input amperage needs to be raised.

http://elinux.org/Polyfuses_explained
Quote
As of the end of August 2012 some users were reporting that newly arrived PI's were fitted with 0 Ohm resistors for the USB polyfuses F1 & F2, and Liz of the RPF foundation confirmed that it was official RPF policy to remove the offending fuses. However F3 still limits the total input current to about 1A.

No F1 or F2 on my RPi only F3. I'm hoping I can get it to work with my incoming asics. Certified powered hub is next on my list.
It arrived this last week.
Do you have anything else to stress test the USB stack with?

Is anyone in general aware if a powered hub fixes the rest of the problems?

As for my tests a powered hub does not fix the issues with usb..... It could cause problems too. Remember that thing hates feedback power over the USB cable. Only cutting the  power lanes of the USB cable helped a little. Still the device is not an option for me with my 18 Ztex boards...

Are your boards bus powered?
sr. member
Activity: 330
Merit: 250
November 26, 2012, 04:51:06 PM
#77
The power for the Rpi will be from my cell charger 5v 1A.
The powered hub will not power the Rpi with my set up. As stated that can cause problems.
The other recommendation is to put both the Hub and Rpi on same power bar. IF I need to turn the RPi off the USB should also be powered down at the same time to avoid random writes.
I don't have a link but all that information is on the RPi forums/FAQs.

I don't have any FPGAs to test though... Sad

I'll try install the BFGminer today or tomorrow.

Should have bought a case with it...
member
Activity: 75
Merit: 10
November 26, 2012, 02:25:55 PM
#76
I've had no problems with 2 BFL singles running through BFGMiner on my RPi, though be sure to read and install EVERYTHING for bfgminer. Its a neat little device and should handle a few more without a problem. Just make sure you have a powered USB Hub and a good power supply for your RPi.
hero member
Activity: 784
Merit: 500
November 26, 2012, 10:54:07 AM
#75
Bridging polyfuses may not be necessary with later revisions, unless input amperage needs to be raised.

http://elinux.org/Polyfuses_explained
Quote
As of the end of August 2012 some users were reporting that newly arrived PI's were fitted with 0 Ohm resistors for the USB polyfuses F1 & F2, and Liz of the RPF foundation confirmed that it was official RPF policy to remove the offending fuses. However F3 still limits the total input current to about 1A.

No F1 or F2 on my RPi only F3. I'm hoping I can get it to work with my incoming asics. Certified powered hub is next on my list.
It arrived this last week.
Do you have anything else to stress test the USB stack with?

Is anyone in general aware if a powered hub fixes the rest of the problems?

As for my tests a powered hub does not fix the issues with usb..... It could cause problems too. Remember that thing hates feedback power over the USB cable. Only cutting the  power lanes of the USB cable helped a little. Still the device is not an option for me with my 18 Ztex boards...
sr. member
Activity: 389
Merit: 250
November 26, 2012, 12:57:32 AM
#74
Bridging polyfuses may not be necessary with later revisions, unless input amperage needs to be raised.

http://elinux.org/Polyfuses_explained
Quote
As of the end of August 2012 some users were reporting that newly arrived PI's were fitted with 0 Ohm resistors for the USB polyfuses F1 & F2, and Liz of the RPF foundation confirmed that it was official RPF policy to remove the offending fuses. However F3 still limits the total input current to about 1A.

No F1 or F2 on my RPi only F3. I'm hoping I can get it to work with my incoming asics. Certified powered hub is next on my list.
It arrived this last week.
Do you have anything else to stress test the USB stack with?

Is anyone in general aware if a powered hub fixes the rest of the problems?
sr. member
Activity: 330
Merit: 250
November 25, 2012, 09:45:48 PM
#73
Bridging polyfuses may not be necessary with later revisions, unless input amperage needs to be raised.

http://elinux.org/Polyfuses_explained
Quote
As of the end of August 2012 some users were reporting that newly arrived PI's were fitted with 0 Ohm resistors for the USB polyfuses F1 & F2, and Liz of the RPF foundation confirmed that it was official RPF policy to remove the offending fuses. However F3 still limits the total input current to about 1A.

No F1 or F2 on my RPi only F3. I'm hoping I can get it to work with my incoming asics. Certified powered hub is next on my list.
It arrived this last week.
legendary
Activity: 1973
Merit: 1007
November 25, 2012, 04:14:53 PM
#72
Bridging polyfuses may not be necessary with later revisions, unless input amperage needs to be raised.

http://elinux.org/Polyfuses_explained
Quote
As of the end of August 2012 some users were reporting that newly arrived PI's were fitted with 0 Ohm resistors for the USB polyfuses F1 & F2, and Liz of the RPF foundation confirmed that it was official RPF policy to remove the offending fuses. However F3 still limits the total input current to about 1A.
sr. member
Activity: 406
Merit: 250
LTC
November 24, 2012, 08:23:44 PM
#71
I thought this post from another thread was pretty interesting:

I drive around 80 FPGAs from a Rasp Pi, you need to do this (bridge polyfuses and add low ESR[ http://en.wikipedia.org/wiki/Equivalent_series_resistance] cap to /Vdd - Vss/ output USB port [nichicon/os-con]:
http://theiopage.blogspot.nl/)
Since I did that hack everything went smooth and cool, before it was a nightmare.
Man, it is tested and verified, just one example, 70 fpgas, one 8 bees controller is out, e.g.:
Quote
...@raspberrypi ~/m5 $ ./as
start_time (unix time)=1353803823.686, elapsed_time=2521.778 secs
accepted_works=3604, rejected_works = 108, result_upload_errors=0, total=3712
active_ktrls=9, inactive_ktrls=0
total_bees=69
KTRL_LIST
ktrl=166, nr_bees=8, bee_mask=ff, stat=00, systat1=07, systat2=87, init_counts=0
ktrl=134, nr_bees=8, bee_mask=ff, stat=00, systat1=87, systat2=97, init_counts=0
ktrl=72, nr_bees=6, bee_mask=cf, stat=00, systat1=07, systat2=97, init_counts=0
ktrl=210, nr_bees=8, bee_mask=ff, stat=00, systat1=47, systat2=97, init_counts=0
ktrl=76, nr_bees=7, bee_mask=ef, stat=00, systat1=87, systat2=87, init_counts=0
ktrl=114, nr_bees=8, bee_mask=ff, stat=00, systat1=87, systat2=97, init_counts=0
ktrl=82, nr_bees=8, bee_mask=ff, stat=00, systat1=87, systat2=87, init_counts=0
ktrl=118, nr_bees=8, bee_mask=ff, stat=00, systat1=87, systat2=97, init_counts=0
ktrl=92, nr_bees=8, bee_mask=ff, stat=00, systat1=87, systat2=97, init_counts=0
legendary
Activity: 952
Merit: 1000
November 24, 2012, 08:17:12 PM
#70
I thought this post from another thread was pretty interesting:

I drive around 80 FPGAs from a Rasp Pi, you need to do this (bridge polyfuses and add low ESR[ http://en.wikipedia.org/wiki/Equivalent_series_resistance] cap to /Vdd - Vss/ output USB port [nichicon/os-con]:
http://theiopage.blogspot.nl/)
Since I did that hack everything went smooth and cool, before it was a nightmare.
sr. member
Activity: 364
Merit: 250
November 21, 2012, 03:10:18 PM
#69
An HTC Sensation charger.
legendary
Activity: 1973
Merit: 1007
November 21, 2012, 03:09:35 PM
#68
Which power supply are you using?
sr. member
Activity: 364
Merit: 250
November 21, 2012, 02:20:32 PM
#67
http://www.raspberrypi.org/downloads
Raspbian “wheezy”

Again, all my issues solved with a proper power supply source for the Pi.
full member
Activity: 210
Merit: 100
November 21, 2012, 01:36:16 PM
#66
i received notice on my 512 RPI shipment yesterday Tongue

i'm curious to test my MMQ with the RPI

I have my BFL single flying on my Pi. Just one though.
Didn't work so well unless I got a better USB charger to supply power.
When you say it didn't work so well with a bad power supply what happened? Did it work for a while then power off like mine? I'm not opposed to trying again with a better PS, but I've got more than one and I don't want to waste my time if there are still Com issues.

I would say the USB/Network was failing.
Using wireless USB or network port it would become unpingable. I would have to power cycle. The Pi would still be on.

which OS were you using?
full member
Activity: 196
Merit: 100
November 19, 2012, 09:14:47 PM
#65
I have my BFL single flying on my Pi. Just one though.
Didn't work so well unless I got a better USB charger to supply power.
When you say it didn't work so well with a bad power supply what happened? Did it work for a while then power off like mine? I'm not opposed to trying again with a better PS, but I've got more than one and I don't want to waste my time if there are still Com issues.

I would say the USB/Network was failing.
Using wireless USB or network port it would become unpingable. I would have to power cycle. The Pi would still be on.

sounds like the hangup issue of the USB... Since the Networking infrastructure WIFI is routed via the same protocol.

It is something that pisses me off, if you look at MANY of the WIFI/Network chips, they state they are 10/100 'compatible'
Yes they ARE.... BUT what they DON'T tell you is that they then interface it to a USB V1.0 core(cheaper IP), before bringing it to the outside of the Network/USB interface chip, which means you can get about 5mbs out of a 100Base before you bottle neck the chips.

(this is NOT the RPi issue, its something related to the shitty little cheap USB pluggable devices for $5.00)

So you have an ARM/cortex device capable of shit hot processing, BUT then the network is tied to 5mbs with some shittly little Asian chip.
Really you are fucking yourself before you start.., ESP. if you are on a hub.

sr. member
Activity: 364
Merit: 250
November 19, 2012, 02:02:55 PM
#64
I have my BFL single flying on my Pi. Just one though.
Didn't work so well unless I got a better USB charger to supply power.
When you say it didn't work so well with a bad power supply what happened? Did it work for a while then power off like mine? I'm not opposed to trying again with a better PS, but I've got more than one and I don't want to waste my time if there are still Com issues.

I would say the USB/Network was failing.
Using wireless USB or network port it would become unpingable. I would have to power cycle. The Pi would still be on.
legendary
Activity: 1973
Merit: 1007
November 19, 2012, 01:38:25 PM
#63
I have my BFL single flying on my Pi. Just one though.
Didn't work so well unless I got a better USB charger to supply power.
When you say it didn't work so well with a bad power supply what happened? Did it work for a while then power off like mine? I'm not opposed to trying again with a better PS, but I've got more than one and I don't want to waste my time if there are still Com issues.
sr. member
Activity: 364
Merit: 250
November 19, 2012, 12:05:43 AM
#62
I have my BFL single flying on my Pi. Just one though.
Didn't work so well unless I got a better USB charger to supply power.
Pages:
Jump to: