Main Site Documentation

Win7 laptop and the narcoleptic Panda II


#1

I did some more digging on the deployment issues between win7 on 64 bit laptops and this is what I have found to date.

Microsoft is aware of some devices not waking up after hibernating and I tried every suggestion they had and none worked except on a temporary basis. Some would work until a full reboot of the laptop.

Before I installed service pack one the port that the Panda was hooked to would not redeploy a Panda, or work with anything else after the first deployment. If you move to a second (and third) port they would work once and then become essentially dead and not recognize anything, not even a mouse or thumbdrive.

After installing service pack one the port would work with anything except the panda until the panda was rebooted.

If the code on the panda has a single Debug.Print, when the laptop goes into hibernation the Panda will freeze. This timing became apparent when I added a serial lcd to the panda.

If the panda has no Debug.Print lines it will happily go on working but if you try to redeploy after restarting the laptop it will not be recognized.

This happens regardless of what is running on the laptop. If there are no programs running and the panda is plugged in it will freeze if it has a Debug line.

The workaround I have been using is to unplug the panda before the laptop hibernates, and plug it back in after a reboot and view the debug info through MfDeploy, or just shutdown the hibernate function which is really not a good option in the field running on batteries.

I put a scope on the usb port to see if I could see what it was saying to drive the Panda into a catatonic state, but my coal fired dual trace has no record facility. The DSO Nano is only single trace which is not useful on D+ D- lines.

While on the Scope subject, has anyone actually received a DSO Quad, and are they worth the postage?


#2

Thanks for the feedback and details. Although his is not top priority and the fix is not on the GHI side I believe, we will look into this further and try to forward as much info to Microsoft as we can. Scope will not tell you anything about USB but a $5000 USB analyzer will :slight_smile:


#3

Thanks Gus. I have no use for a usb analyzer except for what it is saying to scare the Panda to death. And I have workarounds.

While I have your ear… When are you going to get some Fez Touch to Robotshop.ca. Our postal service is on strike and the ups customs charges are horrendous or I would buy them direct.

Also, I note that you are supplying Ribbon cables from the shop. Any chance you will offer the cables with the unused pins broken out on a separate header? I can always use more.


#4

[quote]Thanks Gus. I have no use for a usb analyzer except for what it is saying to scare the Panda to death. And I have workarounds.
[/quote]

What I meant is that we have an analyzer so it will be easier for us to investigate.

[quote]While I have your ear… When are you going to get some Fez Touch to Robotshop.ca. Our postal service is on strike and the ups customs charges are horrendous or I would buy them direct.
[/quote]

I am not the right person to answer these questions but I suggest you contact RobtoShop directly for lead time or special requests.

[quote]Also, I note that you are supplying Ribbon cables from the shop. Any chance you will offer the cables with the unused pins broken out on a separate header? I can always use more.
[/quote]
The unused pins are already exposed on FEZ Touch…plenty of them :slight_smile: Dare you to use them all :wink:


#5

I have a feeling with the debug line freezing the panda it could be more a case of it’s waiting for something back from the PC which it’s never going to get since it’s in hibernate, therefore it would be blocking until it gets the response it’s looking for.

This is where you should be using the Watchdog timer to make sure that nothing sits there too long, which will cause the device to reset if it does.


#6

Thanks Mark. The oddity is that it does this only with the win7 laptop, not the vista laptop or a win7 desktop. If the cable is unplugged it happily sends the debug info into the ether without getting a response back, and it does it when not connected to VS or MfDeploy.

It is not a problem for me now that I know what does it. I have a serial lcd on the datalogger and all the major error possibilities hooked to led warning lights.


#7

Perhaps windows 7 sends some sort of signal down the USB cable when hibernating that confuses the heck out of the debugger, and therefore crashes it?

Have you lodged this as a bug on Codeplex?


#8

Gus is all over this with his inside track to Microsoft. ;D

I guess the worst part of this is for new fez users. I was fortunate because when my first batch of Panda ii’s arrived I used an older Vista desktop and laptop as their launch field until they proved themselves. I had no issues at all with any of them. It was only when I moved the next batch of them to the newer, faster win7 rigs that problems started popping up.


#9

Hi randomr,
I think I came across the same issue after I keep the panda plugged in my win 7 laptop for a while. PC cannot recognize panda.
if the panda is already freeze, what can I do?


#10

it would be better if you start a separate thread rather then adding to a year old thread. there a many possible causes for your problem.