View previous topic :: View next topic |
Author |
Message |
newguy
Joined: 24 Jun 2004 Posts: 1913
|
|
Posted: Thu Aug 14, 2008 3:19 pm |
|
|
Hook the probe to MCLR or hold it there, then do the step over trick that resets the processor. Set the triggering mode to normal, not auto, and set the triggering threshold to about 4V (judging by your traces), falling edge. Set the time/div to maybe 1ms/div. Verify that there is no trace on the scope before you proceed. Then do the step over thing. If a trace appears, it's because the MCLR voltage fell below 4V. On the relatively fast trace time (1ms/div) you should be able to see how long it fell below 4V, unless it's so long that it goes offscreen.
On the 500ms/div auto mode that you're using, a quick transition of a ms or so won't be captured/discernable. |
|
 |
iso9001
Joined: 02 Dec 2003 Posts: 262
|
|
Posted: Thu Aug 14, 2008 5:12 pm |
|
|
Well, Newguy, you were right.
There is a drop in MCLR but I was too far out to see it. Before I tested at 1ms but I didn't have enough hands to pay attention to everything.
Here is the captured waveform... Its saved (grey) because of my aforementioned lack of hands. Image4
Here is an image showing the consistency of the problem. The process is Reset chip, Go, break at temp=5, step over, step over, chip resets, re-breaks at temp=5. Image5
SO.... this actually doesn't tell me much I suppose :\ |
|
 |
newguy
Joined: 24 Jun 2004 Posts: 1913
|
|
Posted: Thu Aug 14, 2008 5:23 pm |
|
|
No, it tells you a lot! Namely that 1) you're not crazy, and 2) it's not your software that's the cause. It's the ICD unit.....you're using the CCS ICD U40, right? [Sorry, wasn't paying attention earlier]
Now throw the ball into CCS's court. |
|
 |
iso9001
Joined: 02 Dec 2003 Posts: 262
|
|
Posted: Thu Aug 14, 2008 6:43 pm |
|
|
Oh I already did
I have a new unit coming and waiting on an RMA number to said the old one back.
For as much as I've been cursing them out the last couple days, they so far have been very receptive to my problem and I appreciate the level of customer service.
HA! I knew I wasn't crazy! ...Although I guess I wasn't 100% sure it wasn't a design flaw, but now that I think about it, the only way for the MCLR to go to 0V is via the ICD.
Thanks for the help on this, I'll have to follow up when I get the new unit in. |
|
 |
iso9001
Joined: 02 Dec 2003 Posts: 262
|
|
Posted: Mon Aug 18, 2008 4:26 pm |
|
|
Well.... F**K.
The new unit got here but does the exact same thing. I tried both units with new and old prototypes, same thing. Tried my workstation.... WORKS FINE.
Crap Crap Crap.... That means my laptop is the problem.
I'll need to figure out whats going on, I don't think it's a USB voltage issue because the VDD is not noisy or dropping on the scope.
I've tried installing the latest ICD software, latest drivers, etc. Nothing.... I'll try another laptop of the same model (toughbook) tomorrow. If that one works (can't imagine it wouldn't) then I'll have to wipe windows off my laptop and reinstall.
What ever the problem is, its STRANGE. |
|
 |
PCM programmer
Joined: 06 Sep 2003 Posts: 21708
|
|
Posted: Mon Aug 18, 2008 4:51 pm |
|
|
Are you using the laptop to supply power to your board through the ICD ? |
|
 |
iso9001
Joined: 02 Dec 2003 Posts: 262
|
|
Posted: Mon Aug 18, 2008 5:50 pm |
|
|
Yes,
Laptop with no battery plugged into the wall, laptop with battery and plugged, just battery, all the same result.
I doubt this is the issue as this setup has been working for 3 years with no issues. But I welcome the suggestions, |
|
 |
PCM programmer
Joined: 06 Sep 2003 Posts: 21708
|
|
Posted: Tue Aug 19, 2008 12:48 pm |
|
|
Laptops have limited power for USB ports. Laptops have all kinds of
power saving schemes. One of these may have been enabled by a
recent software installation or O/S update.
Change to a desktop PC. See if the problem goes away.
Or change your test setup so the target board is self-powered.
All of my test setups are done this way. |
|
 |
iso9001
Joined: 02 Dec 2003 Posts: 262
|
|
Posted: Tue Aug 19, 2008 2:22 pm |
|
|
It's been working for years. Unless someone did so without my knowledge, there has been no updates to the system or it's power management profiles.
I did try a desktop pc and it works fine.
I tried under laptop + self-powered and the error persisted.
I'll try wiping CCS and the drivers on the laptop and re-installing everything |
|
 |
PCM programmer
Joined: 06 Sep 2003 Posts: 21708
|
|
Posted: Tue Aug 19, 2008 2:27 pm |
|
|
There is nothing more I can do. |
|
 |
iso9001
Joined: 02 Dec 2003 Posts: 262
|
|
Posted: Tue Aug 19, 2008 4:46 pm |
|
|
PCM programmer wrote: | There is nothing more I can do. |
Thats OK. I fixed it,
Well, that is to say I got it working again, but there was not specific 'fixing'. Wiped all files and references to ICD and CCS off the drive and out of the registry, reinstalled and ta-da.
Which is just as bad as still having the problem because now I will never know when it might come back.... at least I can keep working now.
Thanks all, |
|
 |
|