This happens very rarely, I've had about 1 or 2 instances over a listening time of about 3 to 4 hours. The symptom is that while running SmartSDR the SDR waveform and waterfall displays freeze for about a second, then the whole Windows display blinks black momentarily, the SDR displays continue running again and then Windows displays an error message - "Display driver amdkmdap stopped responding and has successfully recovered.", usually say's this twice in quick succession.
I was have tried the latest AMD crimson graphics driver and have downgraded to an older catalyst driver - but this gives the same error.
I have uninstalled and re-installed Smart SDR 1.5.1 twice with no change.
Is SmartSDR being actively tested against Windows 10 now ?
Whilst in power management turn off PCI Express "link state power management".
It is possible your video card is going capute BTW. uninstalling / reinstalling SSDR is irrelevant and may make matters worse. Personally I would go into the Device Manager and look for hidden devices and remove ones that do not belong such as dups.
I chased a stupid Nvidia BSOD problem for a year. It seems above fixed 'er up.
Good Luck with the bit bashing.
I had to reinstall smartSDR because of the current fragility it had/has with the recent windows 10 updates, which just complicates things when chasing silly stability issues. Its all conveniently happened about the same time the graphics card errors started. But all points noted thanks for the words of wisdom. I'll update when I've got to the bottom of the root cause. Wanted to test the water before I started blaming components by getting a sense of how others were getting on with Windows 10 etc.
Reverting the drivers might have been the cause of blue screens upon boot I suffered all day. Removing all the AMD drivers and then installing "one" of the previous drivers might have been a better approach.
After failing to boot even into safe mode and then failing to restore from a restore point, twice, I feel lucky that whatever was going on auto-magically fixed itself somehow I had reached the point of attempting a "Clean" reinstall of win10, when I plugged in the wrong USBstick and Windows booted normally instead of loading the Win10 installer.
Another good reason to do the "Clean Win10 Install"
I'm up and running again on the Crimson driver dated 11/17/15 which came out the day before the current 11/18/15 Crimson version.
I figure my milage did vary this time.
73, Jay - NO5J
I've been running system stress tools as recommended by Neal plus some others which so far have show no repeat of the problem .... Yet and SmartSDR is stable so far.
I agree with Jay, likewise I'm keeping clear of Amd Crimson drivers for at least a few updates before attempting again.
Rather have stable SmartSDR than support latest greatest time wasting games.
73, Ray G6UJB
Other than this particular issue, I have not had any major issues with Windows 10. For the most part, I believe it is a huge improvement of Windows 7. I'm not thrilled with the hoops you have to jump through to keep Microsoft out of what I consider private information on my machine, but once hardened, it seems to be ok.
Not a huge issue but if someone has a fix, I'm interested.
I've been having the same issue as others using an AMD Radeon HD 7870 GPU
card. I'm currently running AMD version 15.300.1025.1001,
windows 10 version is 1511 the recent major release of windows 10. Every couple hours I get the error
"Display driver stopped responding and has successfully recovered",
twice in quick succession. I only get
this error when SSDR is running.
This morning it happened again, but this time monitor two went solidly white. It was accompanied with two error messages, bottom right of monitor 1 that the driver failed and had restarted. SmartSDR was running on 2nd monitor and the program was no responding. I had to reboot the system to recover.
I then decided to stress test the GPU by running GpuTest (FurMark) at 1024x640 for about 8 hours. I had absolutely no errors of any kind during this test.
After the test, I fired up SmartSDR and within an hour, "Display driver stopped responding and has successfully recovered" occurred again.
I use this machine for all sorts of pretty intense projects including photo processing, video processing, editing and conversion, video security surveillance processing and capturing, gaming, etc. All without any problems at all. The other day my grandson was here for several hours playing some pretty GPU intense games on this computer without any issues what so ever.
There definitely seems to be a link of some kind between the
AMD driver issues and SmartSDR. It failed again while I was typing this comment. SmartSDR happened to be running on Monitor 1 this time.
Sometimes a blank and recover, but now some system killing dead driver.
My game machine savvy sons suggested the easiest sure thing is to dump the AMD Radeon the PC I bought from Neal came with and set it aside for now.
Before I do that I wanted to ask the group for any other suggestions?
Another point my sons made, knowing that I tend towards a "Two is One, One is None" ideal in tools and resources, that I really should consider having viable second SmartSDR-capable PC in the Shack.
Thoughts on the AMD problem?
Sorry not a fix, just an update.
I've tried following up some fixes found on Windows and games forums that are related to similar problem that appears to be known as a TDR event https://msdn.microsoft.com/en-us/library/windows/hardware/ff570087%28v=vs.85%29.aspx.
So far I've still only been experiencing the event while running SmartSDR 1.5.1. On a few occasions the event has resulted in SmartSDR crashing out, most times SmartSDR just continues running after the TDR event.
I've Tried older Catalyst drivers and the lasted version of the Crimson drivers 15.30.1025.1001 as others here have, but still the same problem, the frequency seems to be less on older drivers, but this may be due to performance tweaks that are in the Crimson drivers.
I had been running system stress testing on the video card using the Pass Mark BurnInTest v8.1 for several days now and I have seen that that the event is never being triggered by stressing the system with GPU 2D and 3D tests. I've even overclocked the GPU to force it to become less stable but this just resulted in a few graphics memory errors but never a TDR event.
I've stress tested with Smart SDR 1.5.1 running and experienced TDR events but the PassMark software has not detect the TDR event as an error and just continued running and completed with 100% pass! (Weird?)
From what I've seen, TDR discussions on forums seems to be more often related to the HD 78xx family of Radeon cards. I wonder if anybody is running anything other than an AMD Radeon HD78xx family card that is experiencing this TDR issue?
Incidentally, my son's are running Radeon R9 family cards, one has windows 10, the other windows 7, neither have complained of this TDR issue while running graphic intense games. I must see if I can sneak a copy of SmartSDR onto one of their machines...wish me luck
I recently came across a post where someone suggested that going into the AMD Radeon Settings and changing the Graphics Profile to Optimize Performance. I made this change on my system about 14 hours ago and so far it seems to be working. I hope I haven't spoken too soon. Time will tell.
If anyone else has the same results or have found a better approach to the problem, please post your results.