Author Topic: Firmware version during winter 2008/2009? (Read 17164 times)

Rickard · « **on:** February 04, 2012, 05:01:15 PM »

Hi,
After many hours and many testings with various settings and memory cards I still haven't solved the issue with memory loss/gaps in the recordings with my 981. But I have looked back and inspected old recordings and I think those made in the winter 2008/2009 are the best and most gap free. The SD card I used in those days is gone, but I think I can figure out which type that was. However, I think the firmware version is the most important factor. I would like to reinstall the firmware that was used that winter, but I don't know which firmware that was. As I recall it I used the latest firmware. Does anyone know the version number in 2008/2009?

There are dates in the Download section, but I don't think they refer to the release dates.

Kind regards,
Rickard

sonar2000 · « **Reply #1 on:** February 04, 2012, 06:24:02 PM »

Rickard, I feel your pain.....It would be nice if there was a matrix to crossreference date with code levels.
Maybe that can be an addition to the forum if Robert can get some time..........

Rhis could be a simple excel file...

However if someone puts a matrix together I will see what I can do to get it in the download section....
chuck

Rickard · « **Reply #2 on:** February 05, 2012, 08:55:00 AM »

Thanks for your compassion Chuck...

I must tell about one of my testing methods, the 'bath tub' test. I fill my traditional enameled steel bath tub with water until it covers the transducer. When I start the unit I can hear the pings clearly. Although the frequencies are far above what's audible the pings seem to produce audible overtones. If I turn the 50 kHz downward signal on and off the difference is easy to hear. If the SI range is extended the ping rate goes down, as it should. In normal operation, with no recording on, the pings come fast at a steady rate. When recording the ping rate goes down and there is a stumbling and stuttering sound. At the same time there is a similar discontinuity in display scroll performance. These things sound and look worse with the faster card. I have found the unit doesn't transmit at all during short moments while recording, so there is actually no data loss since there have been no echoes to receive and save during the silent moments. It seems as if the unit can't communicate with the card and transmit at the same time. For some incomprehensible reason things seem to run smoother with slow cards with more uniform images as a result at the price of slower overall system performance.

I think I know now which firmware I used during the good days. There are rather few to shoose between anyway.

Rickard

sonar2000 · « **Reply #3 on:** February 05, 2012, 10:05:00 AM »

Rickard,
Great testing setup...Hearing some of the noise gererated by pinging is good for helping in the problem determination for several issues.
We have tried a 55 gallon plastic drum which does about the same but because of the small diameter we dont often see any sonar. But what you are doing helps in making sure the transducers is wet and is sending and receiving something....Good work.

Since we dont have access to the operating code of the units I would guess that the processor is a single thread processor and cant poing, receive and transfer to a storage card at the same time..

Dont know if this is a programming arrangement or is due to unit paging memory size, or simply the clock speed.

Do you have the downlevel code you need?

Would be interesting to see if that changes any thing..

Chuck

Rickard · « **Reply #4 on:** February 05, 2012, 11:09:14 AM »

Recording is not the only thing that stops pinging, so does also pushing of buttons and changes of menu settings. The pauses are very brief, but they show the system can't control the sonar part of the system while doing something else. It would be interesting to learn if there has been any fundamental change of system structure in the later models that allows for parallel control of sonar and things like communication with memory. If the newer systems have this capability I will consider upgrading my gear...

I actually tested with old firmware this summer and found that performance with the fast card was the same independent of firmware version. The slow card seemed to perform equally well with old and new firmware. But I didn't check too carefully and I didn't bother about checking which old firmware that worked best.

Rickard

sonar2000 · « **Reply #5 on:** February 05, 2012, 11:35:21 AM »

Good points. I had not thought about other functions going on like the key buttons.
I totally agree with you that there is not a good reason to upgrade.
HB was quick to move the 1197 to a legacy area in order to promote the 1198.

We are still fine with the 1197.

Glad you are taking the time to try some of this out and find the problem. Now if HB would just give you the code and let you fix it...

Chuck

Bob B · « **Reply #6 on:** February 05, 2012, 07:56:26 PM »

Great troubleshooting Rickard. You have found a way to prove that the unit is missing pings rather that missing data in the recording.

Sure sounds like the CPU is having trouble keeping up with everything. With any CPU operating system/firmware, there is a prioritization scheme. In this case, it sounds like the pinging is given a lower priority than other functions. Using the faster SD makes the problem worse because it gives the CPU more to do....and it sounds like the recording function, button response, etc have a higher priority than the ping.

To get a clean recording....give the CPU as few other things to do as possible.

Just my best guess.

sonar2000 · « **Reply #7 on:** February 10, 2012, 03:03:53 PM »

I got to thinking

about this and was wondering if you are using a sd card map and doing a sonar recording could some of the hesitation be also to the unit having to page from the sd map to update the display on the map part of the unit.
Even if we are just on a sonar screen, the unit has to be updating position from the card.
could slow down the display.
chuck

Rickard · « **Reply #8 on:** February 10, 2012, 03:44:37 PM »

No, that shouldn't be a problem because the 981 doesn't have mapping capability and there is only one card slot in the unit. But it can save tracks and waypoints from an external GPS. If the gap issue persists I will test with erasing the waypoint list. Communication with the GPS has no noticable effect on memory performance.
Rickard

Rickard · « **Reply #9 on:** February 11, 2012, 04:45:08 PM »

By accident I found an easy method for detection of memory gaps, or missing pings as it should be called. The old, 4.260 and earlier, firmware for the 981 (and possibly some other models too) had a peculiar bug that caused slanted lines and fields in the SI view when display brightness was set to anything below 10 (10 was default). This bug is gone in the later firmware versions (4.750 and 4.950). However, this interference issue presents an excellent means for detection of the exact moments when pings are lost because the slanted lines are broken at those occasions. The snapshot from Humviewer (thanks again Martin!) show several instances with lost pings. The recording was made with brightness set to 8 at 455 kHz, 20 m range and it was saved with a "fast" SDHC Extreme (30 MB/s) card. Left SI view is almost empty because that side was covered with cork since the transducer was arranged for ice scanning.
Added: 50 kHz was turned off! With the 50 or 200/50 kHz 2D setting the pattern gets too dense for easy inspection. I suppose the same goes for 83 kHz with other transducers.

Rickard

ITGEEK · « **Reply #10 on:** February 15, 2012, 02:04:36 PM »

Wow Rickard,
You know enough about this stuff to build your own sonar unit.
And it might be better than anything that's out there now.

sonar2000 · « **Reply #11 on:** February 15, 2012, 03:02:30 PM »

ITG..indeed a good observation.
There are about 10 good guys here that could put almost any sonar company out of business if they had the financial backing.
The reason for the success of the fourm are the experts that take time to join and contribute.
There are guys that have taken the transducers apart and made their own multi beam large cone.
Others have made towfish.
There are those that understand the operation and keep us straight on how things work.

And last but not least are the IT guys that keep all this incompatibility software working...

Thanks to all...

chuck

Rickard · « **Reply #12 on:** March 29, 2012, 01:58:54 PM »

Through trial-and-error I suddenly found an unexpected condition when the issue with missing pings/lost data when recording seems to be at least alleviated.

When I record with the 200 kHz downward setting only (not 50 or 50/200 kHz) the gaps appear frequently and they are quite disturbing. But, for some very strange reason, the gaps disappear with the 50 or 50/200 kHz downward setting. The gaps may still be there, but if they are, they are probably very frequent and brief.

This observation makes no sense to me. The 200 kHz setting means less data need to be transferred to the memory card, thus things should run easier with this setting. But it causes more data loss!

This bug-like phenomenon may be unique to my model, the 981, but it would be interesting to learn if anybody else have made the same observation?

Regards,
Rickard

sonar2000 · « **Reply #13 on:** March 29, 2012, 04:46:21 PM »

Rickard
Is this a loss of data or just a gap in the screen display during the transfer information to the sd card.

I ask this because if we have gaps in the recorded data file it could mean missing a target on a review with an external viewer.

Not something good for the world of SAR if we are looking for a victim..

Should I be concerned?

chuck

Rickard · « **Reply #14 on:** March 30, 2012, 06:38:44 AM »

Chuck,

I don't say you should be concerned, not yet. I have to make more tests before I dare go out with a general warning. My system is modified almost to death, perhaps, with 50 m with cable and the Doubleducer. I have looked back on some old recordings and looked for gaps at different frequency settings and found no consistent pattern. There are recordings with the 200 kHz only setting which look almost gap-free and there are 50/200 kHz recordings with some gaps. But I have to do this more systematically before any conclusions. All I know for sure is that with my present setup the 50/200 kHz downward setting is far better than the 200 kHz setting.

The gaps appear on the display and in the saved file and seem to be caused by transmit pauses. Some gaps are large enought to hide a drowning victim so a SAR system should be checked for this gap issue I think.

Rickard

sonar2000 · « **Reply #15 on:** March 30, 2012, 10:21:46 AM »

Thanks...
We will be look at recordings closer for gaps and not just a target.

Thanks for all of this and as always let us know of future findings from you point of view and test.

Chuck

Rickard · « **Reply #16 on:** March 30, 2012, 04:16:54 PM »

I actually declined to be voluntary when a driver disappeared with his plow through the ice earlier this winter. I didn't trust my system when I found all these gaps and I didn't want to induce false expectations on the part of the relatives. The police, the Coast Guard, the firebrigade and the military had failed so there was really not much hope for success anyway. But I know my images have better resolution than theirs (between the gaps) and I think I have spent much more time with ice scanning than they have. But my job takes most of my time and my energy and suppose I had been successful, what should I say next time they call and there is no time for me to come? I feel I'm at risk for dilemmas...

The police doesn't know about me and the guy is still missing... The ice won't disappear until the end of april in the area. Volunteering for SAR suits the retired, not me, I'm afraid.

Rickard

sonar2000 · « **Reply #17 on:** March 30, 2012, 04:37:03 PM »

Rickard, you are exactly right.
SAR can be very demanding on the searcher.

As I posted before:
The end goal of a search is to say; The target is right here....Or to say with 99% assurance that the target of interest does not exist in the search area.
To be able to make that statement can require a lot of time but that is what has to be done.
Volunteering is demanding on the person and can be a long term thing.
Once a resource shows their capabilities that name will be spread around.
If you are good then it will be a good spread. If you are not, likewise, the spread will be "dont use that person".
It can be expected that you will volunteer again and again. If you have time fine, but if not then dont do it..
Agencies and especially the family want results so dont do anything that is not productive..
We as providers of sonar have to be good or we waste every ones time...
It also goes to say, that you have to know your equipment and its limits. Dont continue beyond this....call for the next level of support that can produce..

Chuck

Rickard · « **Reply #18 on:** March 31, 2012, 01:23:20 PM »

At last,

I think I have isolated the most important factor that causes data loss when recording. I went out today and made some scans under ice at the 200 kHz and the 50/200 kHz settings with a SanDisk Ultra SD card (class 4/15MB/s) and a SanDisk Extreme Pro SDHC UHS-1 card (45MB/s).

The results show the downward frequency setting is much more important than the memory card R/W speed. At the 50/200 kHz setting the image is almost gap-free with both cards. At the 200 kHz only setting both cards show frequent gaps, but the faster card shows less frequent but larger gaps than the slow card shows.

I have looked back on old recordings and as expected recordings with gaps have been obtained at the 200 kHz only setting. I thought for a long time that the R/W speed in the memory card caused data loss, but I wasn't aware I changed frequency setting at the same time as when the cards were changed. Thus, I was fooled by a classic confounding flaw.

Conclusion: any SD, SDHC or UHS-1 memory card can be used. Use the 50/200 or 87/200 kHz setting.

Note: data loss occur only when recording, in normal scanning there is no data loss and snapshots will be free from gaps.

/Rickard

sonar2000 · « **Reply #19 on:** March 31, 2012, 05:27:04 PM »

Wow. Such detective work.

Good information...

Does this also exist on the SI recordings?
455 or 800...
Chuck

Rickard · « **Reply #20 on:** March 31, 2012, 05:43:06 PM »

Chuck, thanks..

Yes, the gaps appear at the same time in all recordings, the SI channels too. That's why this have been so frustrating.

/Rickard

sonar2000 · « **Reply #21 on:** March 31, 2012, 05:51:08 PM »

This does put a new look at recordings used in searches.
We are going to have to be very careful now.

I would guess we will have to double our scans if the first does not produce.

We cant afford to miss the target..

Thanks for all your work with this. If you see any thing else be sure to let us know..

Hope other SAR guys get to read your post..

Chuck

News:

Author Topic: Firmware version during winter 2008/2009? (Read 17164 times)

Share me

Related Topics