Sam440ep_setup breaks USB with Update 3

[xenic]

Interesting read. Since installing update 3 on my Sam440-flex (733MHz) my keyboard/mouse have locked up several times, usually keyboard only or first with mouse locking up shortly after that. Only solution has been to power off and back on. Not a problem I had experienced before update 3, I had assumed (until reading this) it was related to some bug with MUI-OWB. Will try taking the Sam setup tool out of WBStartup and see if the problem persists.

If by locked up you mean that the keyboard stops working (and all keyboard lights go dark) then a "USBCtrl Restart" command might restore your keyboard without a reboot. My SAM keyboard stops working occasionally and I have added an icon to AmiDock that calls a short script that simply consists of the USBCtrl Restart command. Once I click that icon and select OK in the resulting requester, my keyboard is restored and everything works normally. The only problem is that the system will freeze if I have any USB media device attached when I click the icon.

[Slayer]

Sorry to hear about all these usb quirks. On the 733MHz SAM I run as my everyday system and using OS4.1.3 I've never had the keyboard or mouse once freeze or stop working (yes, I run that ACube patch too)

And I hot swap my dick smith USB 7 in 1 media card reader and my Nokia 63 phone quite often over the course of the week

My keyboard is a simple DSE USB one and a standard USB Microsoft mouse with scroll wheel

The X1000 has a simple DSE USB one and a standard USB Logitech mouse and no problems whatsoever there either

DSE stands for "Dick Smith Electronics" I think, a foreign brand for some of you I'm sure

[chris]

Sam440EP_Setup is irrelevant for these issues. I have tried with and without, and still the keyboard randomly stops working even though the mouse connected to the keyboard's built-in hub continues to be ok.

I find a disconnect/reconnect brings it back to life.

[mechanic]

even though the mouse connected to the keyboard's built-in hub continues to be ok.

I find a disconnect/reconnect brings it back to life.

I wonder if it has something to do with Active and Passive devices, and how they are sensed by the USB chip?

Can you try plugging in another keyboard to the 'on keyboard' hub? Just for grins.

[tonyw]

I've had a similar problem with my Sam 440ep since I first built it into its plastic case. It's in a two-piece plastic case with a 50mm fan that blows air in the back and the air passes out through holes in the front panel.

My experience is that any sort of electrical interference nearby is enough to stop the USB stack with no reported errors or serial output. Just switching on a lamp beside the computer is enough. The lights in the mouse and on the keyboard will go out and stay out until I unplug and replug both cables. Usually when I unplug the first one (say the mouse) and replug it, it will work again, but then when I unplug and replug the other (the keyboard), the first one (the mouse) will go out again. The next unplug/replug will restore normal operation.

As I said, in my case (sorry) it's due to external interference. I find it very difficult using the Sam out in the shed, where the switching on/off of a Weller iron is enough to kill the USB stack.

I don't use the speedup hack and the fan keeps the CPU cool. Leaving the lid off the case doesn't help. It also has always been this way, through all Updates.

[mechanic]

I don't think it has much to do with the stack.

I have a 4 port Belkin hub and the only way it will work after the OS is
up and running is to have something like a mem stick plugged into one of
the 4 ports before inserting into the 440ep. That is true with Linux as well.

I once worked on a communications system where the new modems we got would
work just fine for 3 to 10 days then just lock up. Sometimes even power cycling
would not unlock them, but dragging a long screwdriver or steel wire across one
of the four input bayonet connectors would solve the problem. Not always the same
connector. Here's the kicker, the connectors were insulated from any type of
ground. They did not need ground, and grounding them did nothing. They were
fiber optic modems. Drove me nuts, drove the manufacture nuts, and finally got
replaced for free,,, they tossed them out and ate the loss.

Point is I think the problem is in the chip. Perhaps it is just not for a desktop
system, but works just fine for embedded operations. Or, maybe it's another one
of those things like the VIA chip on the A1 boards and needs a special stack or
an additional signal. ????

[chris]

even though the mouse connected to the keyboard's built-in hub continues to be ok.

I find a disconnect/reconnect brings it back to life.

I wonder if it has something to do with Active and Passive devices, and how they are sensed by the USB chip?

Can you try plugging in another keyboard to the 'on keyboard' hub? Just for grins.

I have no idea what it proves, but I connected a second keyboard to the back of the first one. When the first keyboard froze, the second one carried on working.

[mechanic]

I have no idea what it proves, but I connected a second keyboard to the back of the first one. When the first keyboard froze, the second one carried on working.

Of course I don't know how the first keyboard is wired into USB. perhaps the
plugged in mouse keeps the hub inside the assembly active and thus the second
kbd works.

My plug in hub works best if I always keep a mem stick in it.....?

When I started using a wireless M & K combo the dropping off problems went away.
Maybe that has something to do with the plugged in receiver keeps the port active ?

I just think it's hardware, not stack. Not that the hardware is bad, just not being
properly addressed (?) for a desktop application.

According to the vendor ID the company that supplied the USB is Animation Technologies Inc.. I doubt they made the device, but perhaps had it custom made
per their specs and..... it probably works fine when in the right ecosystem.

Guessing,,,,,,M'eh! what do I know...

[mechanic]

(Talking to myself)

When I started using a wireless M & K combo the dropping off problems went away.
Maybe that has something to do with the plugged in receiver keeping the port active ?

Did a little experiment with a logitech USB kybd that always gave me dropping dead
problems. (Which is why I tried wireless.)

First I disconnected my wireless kit, then.....

I took some paper and folded it, then jammed the pause/break key down then plugged
it into the rear port closest to the MB. The keyboard has been working at least 10
hours now without a problem. It has never gone more than 3 hours before.

I'll leave it on overnight.

P.S. No mouse used at this time.

[astrofra]

Hi,
I'm not sure this is relevant to this thread, but my brand new Sam460cr shipped with OS4.1FE still shows this kind of issue.
Every once in a while, right after workbench shows up, both the keyboard & mouse (an Apple keyboard & a classic wire mouse), or the mouse alone, are not working.
I simply have to unplug/replug each device for it to work properly.

As far as I've seen on the forums, this issue stills occurs in 2015 on the Sam range of computers.
Any pointers ?

(Updated on 2015/03)
Someone uploaded a workaround, here : http://www.os4depot.net/index.php?funct ... ecycle.lha