I got another question via PM, it goes like this:
whats the bit in the troubleshooting section of your manual about the non powered hub latching the 5V power supply on?
the problem I ran into to was using a "self powered" hub that happened to have a power input port on it. it was not a "powered hub." I supplied power to it using a simple 5v linear regulator and found out that it was causing the 5v regulator to stay enabled.
the computer would shut off as normal, but the 5V system would stay on. the regulator was sending power to the hub, the hub was sending that power into the computers USB ports and back out the USB port the enable signal was using. it had latched itself on but it would turn off when I pulled the USB from the hub to computer.
the fix was cutting the USB cable open and soldering in a diode to the 5v wire. that stopped the power going into the computer. when it shut down there was no way for the power at the hub to make it around to the enable signal.
The computer took no damage and still functions to this day. Also I only experienced this with one of the two hubs in my system. the other had a diode built into it, it was labled a "powered hub" the offender was labled a "USB or self powered hub".
I suppose the problem could also have been fixed by simply cutting the 5v line in the USB cable (usually red) and insulating it so it couldn't short out. I am working on making a small adapter that plugs inline with the usb port adding the diode, solving the issue.
the problem is the 5000qty order they want for male type A connectors (like those on USB flash drives), the female type A (computer side ports) are easy to come by. I just need to find a low quantity distributer.
Maybe I can make a version of the ACPPS V1 that routes the USB data signals from the computer to the hub but stopping the 5V from back feeding, making thia a non issue. but making the adapter would be way easier and universal.
what would be awesome is if you could also convert this into a USB hub, so that the PC resuming from standbuy would turn the psu on, and it would reconnect all usb devices. wouldn't be hard, I don't think?
Definitely want one but not right now (paying for some deductables due to accidents, and need a new engine in the 300m...)
Integrate a 7\8port usb hub for a "deluxe" model and this would be insane.
The only way I see that happening is finding a store bought hub and mounting it inside my chassis. A USB hub has so many parts that, with my volume, if i built a 4 port into my design it would add $40+. I think I might be able to find one that could fit inside without too much problem and add $20. maybe?
Originally Posted by pwntbypat
You could probably get away with a 7 port hub for around $12-$13
7port USB HUB: http://focus.ti.com/lit/ds/symlink/tusb2077a.pdf
You dont need a EEPROM, it is optional. So a couple resistors, a 6Mhz crystal, and a new area on your PCB bing bang boom, 7 USB ports.
$5 chip in low (meaning even 1) quantities, and then maybe $1 each connector (high price, I am sure you can get them cheaper like for the Fusion Brain the USB-B Female is $0.30)
I think if you add a hub it will distinguish yourself from an Opus or other power supply like this.
That is a USB1.1 device. and I'd really rather just canabalize another product.
Originally Posted by 2k1Toaster
The USB power supply has been installed in my car since Saturday night and has been functioning properly
Installation was a breeze with the exception of the following:
The back of the G4 has only two USB ports so I wanted to trip the ACPPS from one of the hubs and so I had to splice the usb cable and cut the power wire that supplied +5v to the hub from the computer
That would of worked good except for the fact that the freaken hub WOULDNT TURN ON ON EXTERNAL POWER UNLESS IT WAS ALSO POWERED BY THE COMPUTER
This was a belkin 4 port hub
I tried two other ones using the same (modified usb cable) and those ones worked great
Sooo..... in other words DONT USE A BELKIN HUB!
I m powering two hubs from the ACPPS for a total of 10 devices (I cut computer power to the other hub as well so that I can lighten the load on the G4 PSU and it works great)
I think the fan on the G4 comes on less often
No noise was introduced to the system with the addition of the ACPPS
It runs pretty cool and up to now it has been very stable
Very prompt shipping and tech support!
I definitely recommend it
I will report back if there is any new developments
Sorry to hear about your problems, Spiro.
Originally Posted by spiro
I have since developed a USB isolator/adapter that could have solved your problems. It places a diode between the computer and hub, and has wires coming off the computer side 5V, hub/device side 5V, and ground. I have attached a rough schematic and pictures of the device.
This product serves 4 purposes.
1. Stop power from back feeding after the computer is turned off.
2. Allow tapping the USB’s 5V signal for enabling the ACPPS without cutting a USB cable open.
3. Allow the USB port to be used to enable the ACPPS and still be a functioning USB port.
4. Power a hub using only the USB cable.
1. In a normal scenario, the computer ‘feeds’ power to the hub. ‘Back feeding’ is the opposite, the hub ‘feeds’ the PC, which it shouldn’t do. If power ‘back feeds’ the ACPPS will stay on after the computer turns off. In a round about way, the ACPPS 5V output gets tied to the enable signal and it gets stuck in a loop. The ACPPS stays enabled until you disconnect a wire that is part of the loop. That’s where the isolator description comes in. The USBi stops power ‘back feeding’ into the computer with the use of a diode. (Orange arrow)
2. In many Car-PC applications the computer has a limited number of USB ports. Sacrificing a USB port solely to trigger the ACPPS doesn’t make sense. One option is to cut the USB cable open and extract the 5V signal to enable the ACPPS. Using the USBi will allow tapping into the 5V signal without cutting the cable open. (Blue arrow)
3. Some powered USB hubs won't turn on if the 5V signal is missing from the USB cable. The isolator allows power on the 5V wire to continue (Green arrow) so that the hub turns on. The diode used in the USBi is a Schottky diode, known for low voltage drop and quick response time. I have run several USB speed tests using flash drives connected through the USBi and saw no impact on performance or operation.
4. The USBI is capable of providing power from the ACPPS to the HUB without using a dedicated power cable. A buss powered hub is limited by the amount of current the computer supplies to it. USB ports have current limiting circuitry set at 0.5A. When more than 0.5A is used, the port will shut off and all attached devices will disconnect. In buss powered mode, a hub is limited to 0.5A to power its self and all attached devices. Using the USBi, the 5V output of the ACCPS can be connected to the device side of the diode, allowing the USB cable to carry more than the 0.5A. (Red arrow) A hub could act like an externally powered hub, wile in buss powered mode, because it is receiving power from the ACPPS instead of the current limited USB port of the computer. This use of this capability is AT YOUR OWN RISK. USB cables and connectors are only designed for 1A. Exceeding this may cause the cable to burn up, possibly damaging your equiptment or turning your ride into a car-b-que!
These are built and ready to ship! I have performed performance tests using USB2.0 flash drives and saw no impact on performance.
The USBi wire connections are as follows:Enable connection is the White/Grey wire, the 5V Input is Red, the Ground connection is Black.
I think the same idea can be applied inside a hub to all ports so that they will allow devices connected to the hub to draw more than .5Amp. It would a nice experiment.
I have been researching hubs that could fit inside my design, and modify the hub to provide higher current to all outputs. Then only 1 USB cable is run to subsequent hubs and it would be able to power any attached 4 port hubs at close to full output. the big risk would be damaging cheap/small USB cables. those wires aren't very big and not very good at carring high current 2A would be pushing it.
Originally Posted by 0-0-7