USB Breakout

Interrupting a USB cable, and inserting USB devices like you suggested is pretty much impossible.

What's the purpose of this exercise ?

I have a machine that can accept data and perform as follows, if char 5 == char 6 then do something. The new machine cannot do this, but can do, if char == 1 then do something.

It was a f**k up between 2 departments in my work that will cost in excess of £15,000 to put right by getting custom software and additional hardware(which would double the machines size), or hopefully £10 for a couple of PICs.

Regards

Yes, a couple of pics will not do as you ask.

Do you understand that the PC is the master and the USB device is the slave ?

To get inline on a USB cable is to have a slave for the PC and a master for the device.

Please read up on how USB works, this can not be done as you ask (with a couple of pics).

Good Luck

donald

PS: too bad about the £15,000.

Yes

I was hoping that one of the micros could be master and one slave

I have found a couple of interesting sites, but I think there would be too much to learn in the timescale I have

Thanks for your replies, you have saved me a lot of time on what would have turned out to be a fruitless exercise.

Regards.

Sorry about that, Microchip does not make _any_ master devices in a PIC chip.

I don't even think Microchip make any master parts at all.

There are master chips out there, however the cheapest way to get this done, the with £15,000.

( I hate seeing projects like this go so far before someone finally catches on )

donald

Not with USB. However, if the speed is suitable, you could change the information display to use a RS232 link, probably using ASCII. Then a PIC could make the appropriate changes on the fly.

Wait a second. Depends what the device is. Microchip make parts with USB OTG ports and they provide the software stack to talk to storage devices. It would probably be feasible to add serial class support.

Really ??

I'll have to look for those.

I would like to use a USB stick on a project.

Thanks

donald

Atmel has a branch of their AVR processors that implement "reduced host" functionality. From the app note:

"AT90USBxxx parts can operate both as USB device or USB host, accurately in host mode AT90USBxxx operates as reduced host controller. A reduced host controller has a unique USB port and does not handle full USB tree with hub. It means that a reduced host controller is designed to handle a unique point to point connection with a unique USB device. A reduced host application supports a known targeted device list (VID/PID list). Only the devices listed within this list are supported by the application. In addition the AT90USBxxx USB software library is able to support a targeted list of CLASS/SUBCLASS/PROTOCOL."

So the OP's solution might be achievable. Mouser, Newark, and Digkey have the chips and dev kits.

I haven't worked with any of these (I mostly use the plain AVRs and their CAN relatives), so this is just a pointer, not an endorsement.

Amen!

ld=20

-=20

=20

=20

ing=20

to=20

s.

is=20

=3D=3D=20

=20

s=20

=20

If this is a one-of-a-kind fix, perhaps you can intercept the data=20 before it gets to the USB device and modify the data there. There's a reason you need an XActo knife and wire-wrap wire in your toolkit! Insert the PIC chip of your choice and modify the data stream before it gets to the USB chip.

If you have to fix a production run of 1000+ units, you're probably=20 best off redoing the software.

Mark Borgerson

Yes, it's doable. But it's a good size project by itself.

You can use an At90USB162 ($3 USB device) talking to an At90USB646 ($8 USB master).

That's why we get work fixing other people's problems.

If your company is big enough to have 2 departments that can't talk to each other properly then it's big enough to lose the £15,000 and learn a salient lesson from the ordeal, surely?

Nobby

Just curious - what kind of machines are these? Seems odd to have such specific behaviour in what sounds like relatively expensive equipment.

I agree with the other poster that modifying the data stream with a micro would best be done at some other spot than USB, if possible.

Andrew

That application is specifically why they implemented it.

I'll do a me-2 on that one, however here are some ideas for USB hosts:

Don...

Including, maybe, to use something less inheritly master slave. Maybe even a protocol readily available like say ethernet.

Robert

"}{" skrev i meddelandet news:BOFsj.95097$ snipped-for-privacy@fe2.news.blueyonder.co.uk...

You could conceivably put an MCU with a USB host and a USB device in between, and let the MCU process the packet before sent out/in to the downstream device.

There will be some latency involved, but that may or may not be a problem. If it is a one off project, you could get an AT91SAM9260 board with a device and a host somewhere, maybe at

SAM9-L9260 DEVELOPMENT BOARD FOR AT91SAM9260 MICROCONTROLLER Olimex hasn't got them yet Ulf. :-(

Nice looking board. I tried to order a bunch recently but no luck. Give him a couple of months I think.

Cheers Don...

As they said you can not just splice in. The slave side is easy Microchip ( and many others ) has chips for that. The Host side you could use a FTDI Vinculum. But you will have to deal with the driver code if it is not a standard device.