digital_addict
Posts: 66
Joined: Fri Sep 16, 2011 10:40 pm

Re: USB Port Current Boost

Thu Jun 21, 2012 10:19 pm

Ed Raket wrote:
selsinork wrote: Now if you've got a 500 watt 'phone charger' it's probably not going to be that :)
:mrgreen: :mrgreen: :mrgreen:
Wow a 100amp micro usb plug and socket. :lol:

Just out of curiosity, what is the limit on a micro usb plug and socket? 2amps?

User avatar
AndrewS
Posts: 3625
Joined: Sun Apr 22, 2012 4:50 pm
Location: Cambridge, UK
Contact: Website

Re: USB Port Current Boost

Fri Jun 22, 2012 12:51 am

Did a quick google - most links say 1A but http://japantechniche.com/2011/09/28/re ... y-mitsumi/ says 1.8A

User avatar
jbeale
Posts: 3254
Joined: Tue Nov 22, 2011 11:51 pm
Contact: Website

Re: USB Port Current Boost

Fri Jun 22, 2012 12:59 am

The microUSB socket data sheets I've seen say 1.0 amp per pin if you use all four pins, but you're allowed 1.8 amps if you only use the 2 outside pins (power and ground) with the data pins unused, which is the case in the R-Pi power connector.

gritz
Posts: 449
Joined: Sat Jan 28, 2012 2:33 am

Re: USB Port Current Boost

Fri Jun 22, 2012 1:38 am

Assuming of course that the manufacturer of your cable / connector etc. cares a jot about specifications and standards. Assume nothing. Even if a device is in spec. then "riding the edge of the tolerance" is a dangerous game because your mating device may also be taking liberties! The road to Hell (and landfill) is paved with assertions of "It should work..."

User avatar
Lob0426
Posts: 2156
Joined: Fri Aug 05, 2011 4:30 pm
Location: Susanville CA.
Contact: Website

Re: USB Port Current Boost

Fri Jun 22, 2012 2:14 am

Larry_Adlard wrote:Well here it is - The Pi-pass. It simply bypasses the R-Pi and all it's fuses.

The R-Pi is full of surprises. It appears that the USB isn't hot plugable. You plug in a small device and the machine does a hard reset.
Nice work Your board is still protected on the other side of the 700ma polyfuse. Then you have more power to the USB. Looks like a nice easy mod. One caveat, make sure your hub does not back feed power into this setup. You will then have two PSU's hooked directly together. There is a thread about a RasPI running from the hub power. Maybe Gert can come up with a fused connection like yours to protect against this. Some type of inliune fuse?
512MB version 2.0 as WordPress Server
Motorola Lapdock with Pi2B
Modded Rev 1.0 with pin headers at USB

http://rich1.dyndns.tv/
(RS)Allied ships old stock to reward its Customers for long wait!

g4eml
Posts: 56
Joined: Wed Oct 05, 2011 1:50 pm

Re: USB Port Current Boost

Fri Jun 22, 2012 9:12 am

I also wanted to be able to run a small 4 port unpowered hub from to Pi. No high powered devices, just four low powered ones. I first tried the shorted polyfuse approach. This seemed to work provided all the devices were connected before booting. However if the hub was hot plugged it caused a power brownout of the board which crashed it. This was the case even if the input polyfuse was also shorted out. Looking with a scope showed that the power supply was holding up OK but the voltage at the USB connector was dipping. It looks like the internal tracking of the board had too high an impedance.

I then did a similar mod to the previous poster by adding a wire from the input directly to the USB sockets. Actually I added a thin twisted pair of wires and also bypassed the ground tracking. Having done this I can now hot plug the hub with no problem. I left all the polyfuses in position and not shorted.

I now have a good stable 5V on the USB ports which has resolved a lot of problems. However I can still get kernel panics with some combinations of devices /hubs /ports. This seems to confirm the opinion that the USB software still has a few bugs.


Colin.

User avatar
AndrewS
Posts: 3625
Joined: Sun Apr 22, 2012 4:50 pm
Location: Cambridge, UK
Contact: Website

Re: USB Port Current Boost

Fri Jun 22, 2012 10:38 am

Colin: I've also seen other posts saying that the capacitors on the USB ports ought to be bigger, to better cope with plugging in devices.

EDIT: Doh, earlier in this very thread in fact... :oops:

jbuehl
Posts: 29
Joined: Fri Jun 08, 2012 2:49 pm
Location: Los Angeles

Re: USB Port Current Boost

Fri Jun 22, 2012 2:22 pm

I posted this http://www.raspberrypi.org/phpBB3/viewt ... =45&t=9036 before I knew about this thread. I was thinking something like the Pi-pass might work. I'm going to try that and get rid of the hub.

User avatar
Lob0426
Posts: 2156
Joined: Fri Aug 05, 2011 4:30 pm
Location: Susanville CA.
Contact: Website

Re: USB Port Current Boost

Fri Jun 22, 2012 5:12 pm

It looks like the capacitor is between the USB polyfuses. Replace that with a 150? Or put one in the line from the micro USB to the power pins? If you use a wire.
512MB version 2.0 as WordPress Server
Motorola Lapdock with Pi2B
Modded Rev 1.0 with pin headers at USB

http://rich1.dyndns.tv/
(RS)Allied ships old stock to reward its Customers for long wait!

g4eml
Posts: 56
Joined: Wed Oct 05, 2011 1:50 pm

Re: USB Port Current Boost

Fri Jun 22, 2012 5:47 pm

AndrewS...

I did consider adding an additional capacitor but wiring directly from the input power to the USB sockets cured the brownout problem so I didn't bother to go any further. I now get a steady 5.01V on the USB which has got to be better than before.

Having now ruled out power problems it has confirmed that there is definitely a software problem if you try to run two low speed devices on an external hub. As soon as you plug two low speed devices into an external hub the network and USB start cycling on and off with lots of errors reported with dmesg. Plugging one device into the hub and one into the Pis second USB port works fine. I have now seen this with any combination of three different hubs, two keyboards and two mice. High speed devices dont seem to suffer from this.

Not too much of a problem once you are aware of it but very confusing if your not.

Colin

User avatar
gordon@drogon.net
Posts: 1953
Joined: Tue Feb 07, 2012 2:14 pm
Location: Devon, UK
Contact: Website

Re: USB Port Current Boost

Fri Jun 22, 2012 8:05 pm

Larry_Adlard wrote:Well here it is - The Pi-pass. It simply bypasses the R-Pi and all it's fuses.

The R-Pi is full of surprises. It appears that the USB isn't hot plugable. You plug in a small device and the machine does a hard reset.
I'd suggest that there's something wrong with your device or the Pi. I've been plugging stuff in & out regularly while it's live and running and not seen this...

-Gordon
--
Gordons projects: https://projects.drogon.net/

User avatar
abishur
Posts: 4478
Joined: Thu Jul 28, 2011 4:10 am
Location: USA
Contact: Website

Re: USB Port Current Boost

Fri Jun 22, 2012 8:14 pm

One question, the fuses provide a small modicum of black flow prevention on the USB ports right? If we're jumping things out would it be wise to put a diode in the mix to make sure a surge down the usb ports doesn't burn our pi?
Dear forum: Play nice ;-)

selsinork
Posts: 151
Joined: Mon Apr 16, 2012 8:31 am

Re: USB Port Current Boost

Fri Jun 22, 2012 8:47 pm

digital_addict wrote:Just out of curiosity, what is the limit on a micro usb plug and socket? 2amps?
before, or after you vaporise the connector ?
:twisted:

selsinork
Posts: 151
Joined: Mon Apr 16, 2012 8:31 am

Re: USB Port Current Boost

Fri Jun 22, 2012 9:04 pm

abishur wrote:One question, the fuses provide a small modicum of black flow prevention on the USB ports right?
Not really, a fuse will pass current in either direction.
abishur wrote: If we're jumping things out would it be wise to put a diode in the mix to make sure a surge down the usb ports doesn't burn our pi?
The thing about a diode is that you'll drop about 0.7 of a volt across it, so you end up with 4.3v at the usb side and have possibly compounded a problem.

A powered hub isn't supposed to feed power to the upstream port, only the downstream ones. In most cases you'll be talking about the upstream being a PC with a 600W+ PSU or a laptop with 100-200W either of which likely has active protection for it's usb ports power. But even if not, 600W vs maybe 10W psu for the hub, who's going to win that argument.

As with a lot of these problems, the RPi is a victim of it's own success. Designing down to a cost inevitably means compromise. On the other hand it also means people buying 99p chinese hub off ebay that doesn't follow the standards because the can't bring themselves to spend a significant part of the Pi's cost on a hub. The poor Pi ends up the victim either way.

gritz
Posts: 449
Joined: Sat Jan 28, 2012 2:33 am

Re: USB Port Current Boost

Fri Jun 22, 2012 9:13 pm

abishur wrote:One question, the fuses provide a small modicum of black flow prevention on the USB ports right? If we're jumping things out would it be wise to put a diode in the mix to make sure a surge down the usb ports doesn't burn our pi?
I know what you mean, but you lose voltage across a diode - about .7V for a silicon power diode and ~.4V for a Schottky. It's difficult - no solution is particularly pretty. Sticking everything on a hub and using the hub to power up / down rather than unplugging the Pi is not great, but it avoids getting the soldering iron out. Similarly, if you only need two USB ports then bridging the polyfuses will allow you to use both onboard USB ports with higher current devices, but it requires a decent PSU, asks quite a lot of the micro USB connector (although you can pipe in power through GPIO) and in the event of a USB short could pull a fair bit of current through PCB traces. Hmmm.

I see selsinork has beaten me to it. :)

g4eml
Posts: 56
Joined: Wed Oct 05, 2011 1:50 pm

Re: USB Port Current Boost

Fri Jun 22, 2012 9:18 pm

As mentioned above, including a diode in the supply would re-introduce exactly the same problem that we are trying to avoid. Voltage drop on the 5V line. Even a Schottky diode will introduce a drop of several hundred millivolts.

By doing what seems to have become called the 'Pi-Pass' mod the Pi itself is still protected by the original fuses as the USB power is routed around the outside of the original circuitry.

On the subject of protection, I have had to repair several PCs in the past, including expensive ones, where the user has burnt out tracks on the motherboard by shorting out a USB port. I don't think there is as much protection circuitry about as you might think.

User avatar
abishur
Posts: 4478
Joined: Thu Jul 28, 2011 4:10 am
Location: USA
Contact: Website

Re: USB Port Current Boost

Sat Jun 23, 2012 3:18 am

Thanks for the info guys! (Cause knowledge is power!)
Dear forum: Play nice ;-)

jbuehl
Posts: 29
Joined: Fri Jun 08, 2012 2:49 pm
Location: Los Angeles

Re: USB Port Current Boost

Sat Jun 23, 2012 3:43 am

I did the pi-pass using a 30ga solid wire. I am currently powering an Asus USB-N10 wifi adapter and a Logitech C250 camera using a 1A iPhone power adapter via the micro USB connector. As an added bonus I am cooking my dinner on the wifi adapter :)

BTW I can hot plug a keyboard and the camera with no problems, but plugging in the wifi adapter reliably resets the RPi.

User avatar
jbeale
Posts: 3254
Joined: Tue Nov 22, 2011 11:51 pm
Contact: Website

Re: USB Port Current Boost

Sat Jun 23, 2012 5:35 am

If that WiFi adaptor takes a current surge of, say 300 mA when it is connected (charging up a power bypass cap, for example), that might cause enough voltage drop across the power cable + connector + input fuse F3 to give you a brief voltage glitch large enough to cause trouble (eg. a reset). The 220 uF input filter cap C6 ought to help there, but maybe not enough.

manis404
Posts: 14
Joined: Mon Jan 02, 2012 4:00 pm
Contact: Website

Re: USB Port Current Boost

Sun Jun 24, 2012 8:05 pm

I fixed the hotswap kernel panics by placing a 120uF capacitor between the usb power planes. That worked for me, but it should ideally be at least 220uF.

I updated the post with a graph which could be helpful to understand the concept of "voltage droop". (NOT voltage drop - voltage droop is momentary)

http://theiopage.blogspot.com/2012/06/i ... -host.html

User avatar
Lob0426
Posts: 2156
Joined: Fri Aug 05, 2011 4:30 pm
Location: Susanville CA.
Contact: Website

Re: USB Port Current Boost

Mon Jun 25, 2012 5:15 am

manis404 wrote:I fixed the hotswap kernel panics by placing a 120uF capacitor between the usb power planes. That worked for me, but it should ideally be at least 220uF.

I updated the post with a graph which could be helpful to understand the concept of "voltage droop". (NOT voltage drop - voltage droop is momentary)

http://theiopage.blogspot.com/2012/06/i ... -host.html
What no photo of the Capacitor in place. Oh for shame. :lol: We Want a Picture, it helps those of us that could not get through the cat in the hat! :oops:
512MB version 2.0 as WordPress Server
Motorola Lapdock with Pi2B
Modded Rev 1.0 with pin headers at USB

http://rich1.dyndns.tv/
(RS)Allied ships old stock to reward its Customers for long wait!

User avatar
Ed Raket
Posts: 164
Joined: Mon Apr 30, 2012 4:28 pm

Re: USB Port Current Boost

Mon Jun 25, 2012 6:51 am

Lob0426 wrote:
manis404 wrote:I fixed the hotswap kernel panics by placing a 120uF capacitor between the usb power planes. That worked for me, but it should ideally be at least 220uF.

I updated the post with a graph which could be helpful to understand the concept of "voltage droop". (NOT voltage drop - voltage droop is momentary)

http://theiopage.blogspot.com/2012/06/i ... -host.html
What no photo of the Capacitor in place. Oh for shame. :lol: We Want a Picture, it helps those of us that could not get through the cat in the hat! :oops:
Yes, give us a picture to aid our "more adventurous" tinkerings on the Pi ;) "to boldly go where no one has gone before" :mrgreen:

kta

Re: USB Port Current Boost

Mon Jun 25, 2012 3:56 pm

Let's say I short all three fuses and use 4 pins of the micro-usb connector for power input. Now the question is, how much current can I supply to the usb devices without burning the pcb traces?

User avatar
AndrewS
Posts: 3625
Joined: Sun Apr 22, 2012 4:50 pm
Location: Cambridge, UK
Contact: Website

Re: USB Port Current Boost

Mon Jun 25, 2012 5:18 pm

kta wrote:Let's say I short all three fuses and use 4 pins of the micro-usb connector for power input. Now the question is, how much current can I supply to the usb devices without burning the pcb traces?
It's not quite what you're asking for, but I think the nearest thing to an answer is http://elinux.org/RPi_Low-level_peripherals#Power_pins

You can "supply" as much current as you want, but I guess the Pi would only draw 1.7A maximum - 700mA for the board and 500mA for each of the USB ports.

manis404
Posts: 14
Joined: Mon Jan 02, 2012 4:00 pm
Contact: Website

Re: USB Port Current Boost

Mon Jun 25, 2012 5:34 pm

Ed_Racket

Posted updated, now with PICTURES. I show an external hdd and wlan working simultaneously. I hope the pcb traces will handle the current.(though I'd be surprised if they don't)

http://theiopage.blogspot.com/2012/06/i ... -host.html

Return to “General discussion”

Who is online

Users browsing this forum: Heater, NotRequired, Starmage and 52 guests