Is it possible to simulate a USB drive with software only? I mean, for example, to expose a local memory space to the OS so the device manager will recognize it as a USB device.
I'm not familiar with hardware implementation, but I'm sure that it is possible somehow to emulate the USB protocol.
I began with this book - USB Design By Example.
Can anyone please give me additional references to begin with?
Thank in advance!
You need to create a set of kernel-mode drivers that will do the job (block device driver and some supplementary code). Alternatively you can check our virtual storage products that let you create a virtual disk and expose it to the system.
Related
I recently acquired a Microchip EVB USB7252. I can use it as a USB hub. There are some pins on the board, and I have been looking for ways to use these as an interface between my computer and other devices (preferably my LCD screen).
So I got the IDE + compiler that is supposed to work for this product (MPLAB), but I could find no way to choose my model and get the correct target board.
I was hoping anybody (with likely more experience than me) could steer me in the right direction here. My only previous experience with dev boards is Arduino, where boards show up automatically as COM ports in the IDE. My ultimate goal here is to connect my peripherals to USB without using virtual COM-ports.
https://www.microchip.com/en-us/development-tool/EV33X03A
You can Try The IDE call "MPLAB" you said
and i think arduino CANT use "Microchip EVB USB7252"
So Try it And Good luck
It's a long way because USB is not easy protocol. You need to read USB specs first. Take a look at http://www.usb.org. Generally USB communicate between host and devices. There is special protocol used in this communication.
Perhaps the easiest way to do this will be using Linux and libusb or similar library. When you achieve some experience with this protocol you may be trying to write your own USB driver. I could also recommend to analyse some USB stacks like in Linux or in U-Boot one first.
If you have possibility then I'd also try to use some hardware/software USB protocol analyzers for some overview how does the communication look like.
I need to write a number of drivers for both HID USB devices as well as some old serial devices. The drivers are to pull data off the device and then send the data over to an application that then consumes it. Since the Apple Docs mention that a lot of USB and HID communication can be done from the user space I had assumed that I would not need to write a kernel extension, at least not for the HID devices. Could some one tell me a more solid way to determine this?
Thanks!
If you're writing a single application that must talk to one or more USB HID devices you may well find you can just access the devices straight from the application using the application-level USB APIs.
A kernel driver would be more for something like a networking or mass storage device that needed to integrate with the kernel to be be available to multiple applications.
This Apple document Common QA and Roadmap for USB Software Development on Mac OS X goes into some detail on the matter and links to example code too.
we are developing a sendor device, with a arm7(current: LPC2368) .
this device samples a mv signal,A/D, and need to send this signal data to the PC.(continusly)
at the same time, PC need send command to arm7 (like get temperature, control status, etc..)
rs232 is too slow, so we choose USB.(20K/s - 200K/s)
but the question is, we donnot known how to do usb programming(both pc and arm..)
any direction? any portal? any tutorial?
currently we only sim the device as a HID....
For the ARM side you need a USB Stack. For the PC side you need to implement an USB driver and an application interfacing the driver. It is therefore easier to stick to one of the common profiles (HID, Mass Storage, Virtual COM). For all these you will be able to find USB stacks and not to have to implenent your own. Also you won't need to implement a USB driver for the PC.
I think that the easiest thing to do is to use a Virtual COM approach. From the PC side it would like you are accessing a Serial Port. The speed however can be higher than standard RS232 ports. I have found this USB Stack targetting an earlier processor. You could adapt it for your needs or use it as reference. Generally a Virtual COM driver for the PC will be provided along with the ARM USB stack.
Another approach is to use libusb. This will allow you to interact with USB without writing a kernel driver.
For application notes and commercial USB stacks look here. If you are determined to write your own stack and driver, Jungo is the industry leader for embedded USB stacks and drivers.
I have a device that came with an AC power adapter where the connector is a mini USB plug. The device however doesn't seem to power itself from a computer's USB port (using a standard USB-mini USB cable) unless a specific driver is installed. The driver is only available for Windows. I would like to charge the device from USB plugs on different platforms.
My question is: why isn't power getting to the device without the driver? Is a driver always required for a USB port to start giving power? Or is it this device that's specifically made not to take a charge unless some software routine triggers it to do so?
I guess my question can be summarized as: Is power not present on the USB cable or is it present but the device ignoring it. If the answer is the former, I'll be trying to figure out how to write software that will enable the voltage to always be present.
Thanks
Why isn't power getting to the device without the driver?
USB ports are always powered when the computer is on and the USB control software hasn't detected current overdraw.
Is a driver always required for a USB port to start giving power?
No, the USB port is always required to start off providing power to the device, otherwise the device could never initiate a connection.
Or is it this device that's specifically made not to take a charge unless some software routine triggers it to do so?
This can be complex. To meet the USB spec a device cannot pull more than a few mA until it's registered with the computer.
However, nearly every computer allows the USB port to pull the full 500mA (and more) before it'll shut the power off.
The device you're charging is being nice by not pulling any significant power until the computer gives permission.
Writing software won't help, the device has to register with the USB bus, which will best be done with the driver.
However, the plug in charger doesn't do that. It likely has shorted the two data lines of the USB plug together, which signals the USB device that it's not connected to a computer and can pull the full 500mA without waiting.
Take a USB extension cable, cut off the jacket, and short the data lines (green and yellow, sometimes) together on the end going to the USB device, and leave them cut without touching anything on the end going to the PC, and leave the read and black power wires connected through.
It might work. If not, take the wall charger apart and find out what it's doing with each of the four USB wires, and see if you can duplicate that.
This might be helpful if you are targeting a linux system.
This seems to be platform-specific. In Linux, USB ports are always energized, while on Windows they don't. Thumbdrives with LEDs turn off when unmounted in windows, but in Linux they stay lit. My cellphone's manual says that it can't be charged by a PC, but I regularly do on my linux machine, I guess that's because they don't have a driver and windows won't power up without one.
Have you tried plugging it into a 'dumb' USB port - like the one on a car charger? Those ports are pure power and don't create a USB network. I think.
Unless you have the hardware specs from the manufacturer, I think you are out of luck. You could try reverse engineering the driver to see what it does, but I'd expect it would be cheaper and easier just to buy one with cross platform drivers or charges without the driver.
I have a piece of hardware that sends USB data over ethernet (only the data stored in the package will be send). On a remote PC the data is recieved via ethernet. How can I send this data to the USB driver so it translates the data into commands applications can use?
You're better off getting hardware that does the reciprocal, sends the IP-based USB information to the USB subsystem, rather than try and hack the software driver itself. I can't imagine your hardware vendor doesn't have a device that does this.
You need a server listening on whatever port/socket that you are trying to connect to. Twisted Matrix makes decent Python libraries for network communications.
I think this is going to be troublesome.
USB is generally set up to associate a driver with a connected device, based on the device's various ID numbers, as discovered during bus traversal.
Your data comes in over Ethernet, so the platform's USB driver stack won't know anything about the device in question. This means you somehow need to directly talk to the proper driver, and also get it prepared to handle events from a (from its point of view) non-connected device.
I can think of several reasons why even a well-designed USB stack won't handle this happily.
On Linux, you might be able to "cheat" by interpreting the data yourself and sending it on, using the same API:s the actual driver would have used. That won't work for any USB device of course, it requires you know what the device is.
It's doable on windows as well, but you need a lot of kernel/usb knowledge to make it work i don't think i will be wrong by estimating this task as few man years (you can reduce this estimation dramatically if you have a limited selection of devices/types of device to support.
You will need to develop a bus driver that will simulate the host controller driver to the native usb host, unfortunately this interface is not public and we did not managed to get MS cooperation on that.
There is additional option to work on hub level, instead on controller level, this interface is available, but i did not managed to find my notes on that.
You can download the evaluation version and investigate the driver stack it might give you a clue where to start.