Report in - x36HP SD Card Problems/Corruption

Status
Not open for further replies.

FeedForward

Member
Joined
Jul 7, 2014
Messages
75
rjdj, when your card gets corrupted, is the USB cable hooked up to your computer? It seems like Exivius is onto something when he described how his card corruption is always related to the use of the USB cable. What is confusing to me is why the actual formatting of the card is destroyed. I can understand how some data could be corrupted, but the card formatting? There must be a very poor USB/card interface responsible for all this trouble.

FF
 

rjdj2000

Gone Cuckoo
Feed Provider
Joined
Jan 24, 2011
Messages
347
Location
Central NY
FeedForward,

No it is not hooked up to the computer at all. It just sits on the counter where I have the external antenna wire coming in so that I can hook it to that. The computer is in another room altogether. It is just I went to try a 'Replay' of the audio and when pressing the button there was no reaction from the scanner at all. No beeps or anything. Turn it off and back on I got the SD card error that everyone else has. I was able to recover it once but not the second time. So I now have a 16 Gig, Class 10 card in it and I rolled back to the previous firmware, I think it is 1.02.03? So far it has been running 24/7 since Sunday and I have been able to go back and hit 'Replay' and it functions. Once it did give me nothing to replay and I know there was audio just about 15 seconds before I hit the button.

So far the scanner has been working like it should, but I hate to lose this Class 10 card so I am treading lightly with it. I still have been trying to get the original card back but have not had much luck. Need to get me a card reader for my Mac machines and see if I can do anything then as it didn't work as it did before when it was in the scanner.


rjdj, when your card gets corrupted, is the USB cable hooked up to your computer? It seems like Exivius is onto something when he described how his card corruption is always related to the use of the USB cable. What is confusing to me is why the actual formatting of the card is destroyed. I can understand how some data could be corrupted, but the card formatting? There must be a very poor USB/card interface responsible for all this trouble.

FF
 

SCPD

QRT
Joined
Feb 24, 2001
Messages
0
Location
Virginia
I'm not sure if this is an SD card error or a 536 error

So driving and scanning can be a pain, I have my favorites list scanning with about 35 systems. and sometimes I oops and turn all systems off getting the "Nothing to scan" message. when I go to access a system to scan it refuses to enable that or any system. You have two choices at this point.

A: unplug the radio, re-power it and it brings up last saved memory option.
B: shut it off and drive home and re-write the file to the radio.

Is this an SD card issue or a 536/Sentinel issue?
 

JamesO

Member
Joined
Jan 22, 2003
Messages
1,814
Location
McLean, VA
So driving and scanning can be a pain, I have my favorites list scanning with about 35 systems. and sometimes I oops and turn all systems off getting the "Nothing to scan" message. when I go to access a system to scan it refuses to enable that or any system. You have two choices at this point.

A: unplug the radio, re-power it and it brings up last saved memory option.
B: shut it off and drive home and re-write the file to the radio.

Is this an SD card issue or a 536/Sentinel issue?

If it is an "I opps and turn all systems off" then this is pilot error and you need to understand how to re-enable systems to Scan.

Choose - Press the Menu button "Set Scan Selection", then "Select Lists To Monitor", then make sure everything is set to "On" that you want to scan.

The other issues are Service Types

Choose - Press the Menu button, choose "Select Service Types", the toggle/turn on any Service Types you want to listen to. Make sure ALL Service Types are NOT turned off.

Not sure this is an SD card issue, might be a pilot error issue?

Buy a spare SD card, configure the radio the way you want it. Copy the functional SD card, take the copy and wrap it in a small fold of paper, then tape the spare SD card on the side, bottom or rear of the radio. For the HP1 and 436HP, tape the spare card on the inside of the battery door. Then if the current SD card rolls over on you, you have a spare on hand to get the radio back and functioning again quickly.
 

Anderegg

Enter text in this field
Premium Subscriber
Joined
Mar 7, 2010
Messages
2,662
Location
San Diego
I cannot replace batteries with the SD card read error. Card must be removed and reinstalled before the scanner will power up again normally.

Paul
 

JamesO

Member
Joined
Jan 22, 2003
Messages
1,814
Location
McLean, VA
Occasionally when changing batteries, the SD card sometimes will have errors and require the SD card to be removed and reseated.
 

drdiesel1

Member
Joined
Jul 5, 2007
Messages
145
Location
Terre Haute IN
My factory card died a few days ago. Noticed the scanner was locked up, turned it off, corruption error when back on.

I programmed my 536 when I first got it, one of the first batches out the door, only programmed once. No audio recording and it's been powered up pretty much the entire time.

The card was dead, unable to format/mount via computer as well.
 

dmdx86

Member
Premium Subscriber
Joined
Mar 19, 2014
Messages
51
Happened to me today - about an hour ago. Scanner is hooked up to a power conditioning UPS so it should not be affected by power burps. This was totally random as far as I can tell.

Scanner model - 536HP
Card manufacturer and size - SanDisk 4 GB
Card original or replacement - original
Was card able to be recovered - NO. Unable to mount in Windows. Unable to format (freezes computer when I try). Tried putting it in my Android phone. Phone said card was damaged - allowed me to format it, but after formatting, it still said SD card was damaged.
Was problem caused by power disruptions - NO
Did problem happen during normal use - Scanner was doing normal scanning. Was not recording.
Has replacement card(s) become corrupted - Haven't replaced it yet. Going to take care of that in the morning.
 

dmdx86

Member
Premium Subscriber
Joined
Mar 19, 2014
Messages
51
I have two theories on this:

(1) As mentioned by others, the "replay" feature cannot be disabled. Are we perhaps reaching the limit on the number of write cycles on these cards?

IMHO, Uniden made a poor design choice by buffering the replay data to the card rather than using internal RAM for this.

Uniden could mitigate this problem by issuing a firmware update that allows replay to be disabled, thus drastically reducing the write load on these cards.

(2) The other possibility is that the SD card module inside the scanner is ruining the cards with too much voltage or some other electrical fault. If this isn't a write cycles issue then that's where I would point my fingers.

If this is the problem, it very likely would require expensive board re-work and could explain why Uniden is not acknowledging the problem despite many months of numerous reports.
 

jonwienke

More Info Coming Soon!
Joined
Jul 18, 2014
Messages
13,416
Location
VA
(1) As mentioned by others, the "replay" feature cannot be disabled. Are we perhaps reaching the limit on the number of write cycles on these cards?

Unlikely. A typical 4GB card has over 3GB of free space. If the card controller does any kind of wear leveling (which all newer flash devices do), writes are being spread out among all free sectors on the card. So any given physical sector of the card is only being written to once every 100 hours or so. If the card is rated for 100,000 write cycles, that works out to an expected card life of over 1141 years.
 

JamesO

Member
Joined
Jan 22, 2003
Messages
1,814
Location
McLean, VA
Unlikely. A typical 4GB card has over 3GB of free space. If the card controller does any kind of wear leveling (which all newer flash devices do), writes are being spread out among all free sectors on the card. So any given physical sector of the card is only being written to once every 100 hours or so. If the card is rated for 100,000 write cycles, that works out to an expected card life of over 1141 years.

Spec and "Real World" are 2 very different things. Also do you "know" for sure the way the card is being written to has wear leveling and is it working properly??

The facts are kind of speaking for themselves here.

Maybe a bad batch of original SD cards?

Maybe a batch of counterfeit SD cards?

Maybe something in the scanner is zapping the cards?

Unfortunately these reports keep rolling in and there is nothing yet that appears to be concrete as to what is happening.

I guess we will have to wait and see how many of the replacement SD cards have similar issues.
 

jonwienke

More Info Coming Soon!
Joined
Jul 18, 2014
Messages
13,416
Location
VA
If the card isn't wear leveling, it's defective. I did misread the specs; typical life is more like 5000 erase cycles instead of 100,000, so 100 hours x 5000 write cycles / 24 / 365 = 57.2 years.

Cards with more free space will last longer because writes are spread out among a larger pool of available storage.

The problem very well could be a batch of bad cards. But it's not due to the way Replay works.
 

JamesO

Member
Joined
Jan 22, 2003
Messages
1,814
Location
McLean, VA
If the card isn't wear leveling, it's defective. I did misread the specs; typical life is more like 5000 erase cycles instead of 100,000, so 100 hours x 5000 write cycles / 24 / 365 = 57.2 years..

Tell this to the GRE owners that had the hit counter enabled in the scanners that destroyed the on board RAM chips!! The clearly did not last 57 years.

Also specs are specs, but at the end of the day how did the SD cards get qualified and tested for 57 years? These numbers that vendors put out are theoretical at best because most of the solid state memory was not likely stress tested to the full number of write cycles.

It also comes down to supporting chips that the primary vendor may not even manufacture.

Just ask OZTech about the Hitachi chips in their solid state hard drives!! I had an office that some IT monkeys put all SSD in the computers and I had to take everyone out after less than 18 months due to epic failure of these SSD's.

Uniden should have designed these scanners with volatile replay buffer chips. If all the replay audio is being written to the SD card, there is just too much going on and unnecessarily.

Then if you wanted to actually record the audio it could be stored in the volatile memory and then transferred in batches to the SD card. This would probably reduce the write cycles by a factor of 10 to the SD card. Only issue there is the possibility of loosing recently recordings due to a power loss.

The problem with public safety conversations is they are usually very quick/short. Anything more than 10 seconds is considered long. Just look at the lengths of the recordings stored on the card. So even in a few hour time frame scanning a busy system there may be hundreds or close to a thousand transmissions that would need to be recorded.

I think the scanner is probably the busiest and most stressful thing writing to an SD card. Everything else is peanuts compared to the scanner. Smart phone, cameras and so forth do not have the write duty cycle that a scanner may have recording all the transmissions in a 24 hour period.
 

dmdx86

Member
Premium Subscriber
Joined
Mar 19, 2014
Messages
51
I replaced the broken SanDisk card with a Sony 8GB card last night. So far so good.

The first observation I have with this new card is that I don't get the delay between P25 transmissions that I did before. Previously, I had an issue with a local fire dispatch channel where the tone out would be sent, followed by about a .25 - .5 second delay, and then the dispatch itself. The scanner would not react quickly enough and miss about .5 seconds of the dispatch. Doesn't happen anymore.

I'm not 100% sure that's card related (maybe powering the scanner down for over 24 hours helped somehow) but a theory on this is that old card was not writing the replay data to the card quickly enough and caused the scanner to experience this latency issue (i.e. it had to wait for the data to be written to card before it would continue doing anything).
 

jonwienke

More Info Coming Soon!
Joined
Jul 18, 2014
Messages
13,416
Location
VA
Tell this to the GRE owners that had the hit counter enabled in the scanners that destroyed the on board RAM chips!! The clearly did not last 57 years.

Also specs are specs, but at the end of the day how did the SD cards get qualified and tested for 57 years? These numbers that vendors put out are theoretical at best because most of the solid state memory was not likely stress tested to the full number of write cycles.

Wear leveling needs free space to be effective. If there is only one sector of available storage, then that one sector is going to take all of the writes and wear. If the writes happen every 10 seconds on average, then the 5000-write lifespan of that one available sector would be less than 14 hours. I'm fairly sure the GRE scanners you mention lasted longer than that.

Automatic recording does impose a finite lifespan on the flash storage. You can estimate how long your storage should last with the following formula:

(Average time between writes) * (number of rated write cycles) = (expected operating lifespan of flash storage device)

I recorded some audio with my 436HP, and found that just over 15 hours of recorded audio fit in 1GB of space. So if your card has 3GB of free space, then your worst case average time between writes on each sector would be about 45 hours, and the card should last for just under 26 years of continuous use before write-cycle wear kills it, if it has a 5000-write lifespan.

If you're not getting that kind of life from your card, then there is some defect or other problem preventing you from getting the full rated lifespan of the card.
 

Anderegg

Enter text in this field
Premium Subscriber
Joined
Mar 7, 2010
Messages
2,662
Location
San Diego
Does the SD card speed rating affect the speed of the scanners interface? I just swapped out the factory 4GB with a SanDisk Extreme 16GB version, and now the 0.5+ second delay in response when pressing numbers (turning on and off favorites etc) has vanished. The favorites also load faster when I power up, the bar shoots across the display about twice as fast. Before it was like typing on a very old Windows computer where the text you would type would continue to show up on the screen after you finished typing!

Paul
 

jonwienke

More Info Coming Soon!
Joined
Jul 18, 2014
Messages
13,416
Location
VA
All else being equal, a faster card means the device it's in spends less time sending data to or reading data from the card.
 

JamesO

Member
Joined
Jan 22, 2003
Messages
1,814
Location
McLean, VA
Wear leveling needs free space to be effective. If there is only one sector of available storage, then that one sector is going to take all of the writes and wear. If the writes happen every 10 seconds on average, then the 5000-write lifespan of that one available sector would be less than 14 hours. I'm fairly sure the GRE scanners you mention lasted longer than that.

Automatic recording does impose a finite lifespan on the flash storage. You can estimate how long your storage should last with the following formula:

(Average time between writes) * (number of rated write cycles) = (expected operating lifespan of flash storage device)

I recorded some audio with my 436HP, and found that just over 15 hours of recorded audio fit in 1GB of space. So if your card has 3GB of free space, then your worst case average time between writes on each sector would be about 45 hours, and the card should last for just under 26 years of continuous use before write-cycle wear kills it, if it has a 5000-write lifespan.

If you're not getting that kind of life from your card, then there is some defect or other problem preventing you from getting the full rated lifespan of the card.

I appreciate all your and the manufactures theoretical information, but I call BS on all of it.

There is no way that 26 or 50+ years of service life has been tested on any of these products. If you frequent many of the photography forums there are all sorts of complaints of flaky SD cards, pro photographers are probably running the closest write cycle to these scanners and they are probably just a fraction of what the scanners do. These are all theoretical and calculated life cycles, not actually tested data and cards that are hammered on time and time again.

Lets talk in 26 years to see who's SD cards are still working. There are few electrical things that still work after 25, much less 50 years without issues, degradation or repairs. The only thing I can think off is the lasted over 50 years is 1950 spare refrigerator at my moms house that still works to this day. Has to be manually defrosted and it just keeps chugging away. I will likely find a spot in my garage for it at some point as it seems to have nine lives.

Out of all the USB flash drives I own and have laying around, I would bet I only have about 100 write cycles on them combined, so most flash memory gets very little usage. Even digital video cameras may continually or batch write to the SD cards, but I am not hearing high failure rate with video writing.

People do not use their flash storage like these scanners. I am not sure how Uniden writes to the the SD card, but if they do not batch write to the card and write every 5-10 seconds transmission, this is just a recipe for disaster.

I expect that faster and higher quality, larger SD flash cards may yield better overall performance in these scanners, but I expect to see continued issues on units that are recording on a 24 hour a day basis.
 

jonwienke

More Info Coming Soon!
Joined
Jul 18, 2014
Messages
13,416
Location
VA
I appreciate all your and the manufactures theoretical information, but I call BS on all of it.

There is no way that 26 or 50+ years of service life has been tested on any of these products.

Actually, your assertion is the BS, The number of write cycles a flash storage sector can undergo before failure is easily testable, and manufacturers can and do such testing constantly. 5000 read/write cycles can be tested in a few seconds under lab conditions. You just read and write to a physical sector continuously (with wear leveling turned off), instead of waiting 45 or more hours between each read/write cycle.
 
Status
Not open for further replies.
Top