The Naked Scientists

The Naked Scientists Forum

Author Topic: How have engineers rectified the problem with weather satellite NOAA17?  (Read 7857 times)

Online syhprum

  • Neilep Level Member
  • ******
  • Posts: 3819
  • Thanked: 19 times
    • View Profile
For at least three months there have been problems with the scan motor on NOAA17 resulting in useless pictures now at last the engineers seem to have fixed it, would really like to know how.

http://www.oso.noaa.gov/poesstatus/ gives no clues as yet.
« Last Edit: 29/05/2010 10:18:59 by chris »


 

Offline LeeE

  • Neilep Level Member
  • ******
  • Posts: 3382
    • View Profile
    • Spatial
Hmm... yes, the relevant status page, via that link, still indicates that the AVHRR subsystem data is now unusable.  Are you retrieving the data via your own dish?  I guess you must be, to know that it now seems to be working.

I think we can only guess what the ground engineers have done but I think that there may actually be some clues on that status page.

Note that they've detected an "Elevated spin motor current" i.e. an electric motor is drawing more current than it should do, which indicates that the motor is not spinning as freely as it should be.  There could be a number of reasons for this, ranging from wear to the bearings, to something loose floating around inside the motor casing and jamming the motor.

It's possible that wear to one of the motor bearings resulted in a burr, but the engineers may have been able to break the burr off by continued operation of the motor (but note that if this is the case then there will now be a bit of unwanted bearing material floating about inside the motor casing, which could lead to subsequent problems).

Loose material inside the motor casing could have come from any non-essential bit of the equipment detaching and getting lodged in the motor: as well as a detached burr, this could be something as simple a a flake of radiation degraded paint from the inside of the motor casing, and once again, continued operation of the motor might have cleared it, or worn it away.

The only way to be sure what happened and how it was fixed will be to wait until they update the status page (they may be holding off doing this until they're sure that the fix will be effective in the long-term), or of course, you could try asking them.
 

Online syhprum

  • Neilep Level Member
  • ******
  • Posts: 3819
  • Thanked: 19 times
    • View Profile
NOAA15 had similar problems about 7 years but it seems have cured itsself and run quite happily ever since.
I wonder if later satellites will abandon mechanical scanning now that pixel type imagers are now highly developed.
I notice tape recorders have been replaced by solid state drives.
Early weather satelites used image orthicon TV tubes but they had a very short life.
Yes I have my own receiving system, its been going since 1965.

As you say it might be a foreign body in the motor, I used to work with so called printed circuit motors that had ceramic magnets and it used to happen to them.

PS I have tried asking but get no replies !
« Last Edit: 26/05/2010 13:39:03 by syhprum »
 

Offline LeeE

  • Neilep Level Member
  • ******
  • Posts: 3382
    • View Profile
    • Spatial
Yes I have my own receiving system, its been going since 1965.

Nice, and respect  :)

Quote
I have tried asking but get no replies !

That's a shame, but worth a try.
 

Online syhprum

  • Neilep Level Member
  • ******
  • Posts: 3819
  • Thanked: 19 times
    • View Profile
It seems that the engineers allow it to run for a few orbits with the excessive spin motor current when it gives excellent pictures then restore the current limitation when the pictures become only partly usable.
Not doubt they are hoping something will get run in.
Here is a small section of the IR picture with the misplaced VIS picture superimposed
« Last Edit: 27/05/2010 21:16:01 by syhprum »
 

Offline LeeE

  • Neilep Level Member
  • ******
  • Posts: 3382
    • View Profile
    • Spatial
Hmm... the left 5/6 of that image looks like what you'd expect if the scan motor wasn't working but the right 1/6 looks like calibration data.  It's interesting to note that the horizontal noise seems to have similar characteristics in both images though - do you know where that's coming from?
 

Online syhprum

  • Neilep Level Member
  • ******
  • Posts: 3819
  • Thanked: 19 times
    • View Profile
The noise is due to the low altitude and range of the satellite and the damage to to the antenna in recent storms (I hope to collect a new one on my trip to Indianapolis this weekend).
I think the patterning is due to failure of the data read in pixel clock (presumably generated by a disk on the scan motor) so that one line of data is continually read out.
What you are seeing on the picture is a combination of the IR pic and the Visible pic,(which is done to produce a false colour picture) when the two componets are viewed separately the IR picture is OK.
« Last Edit: 28/05/2010 08:42:19 by syhprum »
 

Offline LeeE

  • Neilep Level Member
  • ******
  • Posts: 3382
    • View Profile
    • Spatial
Hmm... how low is the orbit of the satellite, and how frequently does it produce images?

The most common sort of scanning scheme for this sort of thing is known as 'line-scan' and it would fit with the image we're seeing if a single scan motor was malfunctioning.

In this scheme the sensor doesn't use a 2D sensor array, which could capture the entire 2D image in a single 'shot', but uses a single point detector.  In some ways it is similar to the way that we read a page of text i.e. word by word, from left to right, then moving down to start the next line once we've reached the end of the current line.  Once we've read the last word in the last line we then turn the page to start a new one, starting at the top left again.

If the satellite is using such a line scan scheme, the purpose of the scan motor is to move along each line, scanning each word in succession, whilst the forward motion of the satellite is what advances us down to the next line.  Once the entire 'page' has been scanned the data is combined into the single 2D image and transmitted (alternatively, each line could be transmitted as it is acquired but this means an added comms overhead, as each line of data will need to include metadata, instead of just a single block of metadata for the entire image).

If the scan motor is not working though, then we can't advance from one word to the next along each line and only see, for example, the first word in each line (note though, that it doesn't have to be the first word in each line; the actual word that's being scanned depends upon where the scan motor has got stuck - it could mean that we only see the third, or last etc. word in each line but the point is that while the scan motor isn't working we're only getting a single word from each line instead of the complete line).  However, the satellite keeps moving so once the satellite thinks it's completed each line it still starts another one.  The result is that each line of the page is different but each line just consists of the same word repeated over and over again.

This would explain why we're seeing a series of vertical lines in that image, with each pixel in any particular line being the same but with each line being different from each other.

If the IR data is being acquired by the same scan motor it could explain the horizontal lines on the right of the image too, but it would suggest that the data orientation for this block has been swapped prior to transmission, which is why we're seeing horizontal lines instead of vertical lines.  However, I'm not sure why this should be done, other than to ensure that the two blocks of data can't be confused with each other, and bearing in mind that the metadata should be able to identify where the two blocks of data start and end I can't say that it's a very convincing arguement.
 

Online syhprum

  • Neilep Level Member
  • ******
  • Posts: 3819
  • Thanked: 19 times
    • View Profile
The height is about 800 Km and the orbit time 101.2 minutes.
All these problems are pretty familiar from the Hell and Muirhead scanners I have worked with, I can only think they consider the pictures from some orbits more important than others and are prepared to risk a higher motor current.
Here is a section of a normal composite picture from early this morning.
The pictures are transmitted as a continuous data stream with each line lasting 0.5 sec and the Vis and IR side by side, the overhead time is about 15 minutes the the passage of the satellite overhead providing the vertical component of the scan.
I have sent a further email to the engineers requesting information but I don't expect they will reply to a private individual, maybe if I was representing a government dept or university.
« Last Edit: 28/05/2010 23:02:17 by syhprum »
 

Online syhprum

  • Neilep Level Member
  • ******
  • Posts: 3819
  • Thanked: 19 times
    • View Profile
The latest bulletin

Spacecraft Mission Data 
Spacecraft Letter: M  International Designation: 2002 032A  Catalog Number: 27453 
Launch Date: 06/24/2002  Operational Date: 10/15/2002  Operational Status: AM Backup

 
Notes: AVHRR degraded performance continues due to the elevated scan motor current. MIRP Re-phase Eable/Disable commands are being applied via SCT once a day at the souther day terminator.

This appears to take place over the Azores
« Last Edit: 20/07/2010 16:54:25 by syhprum »
 

Online syhprum

  • Neilep Level Member
  • ******
  • Posts: 3819
  • Thanked: 19 times
    • View Profile
They seem to have it well under control now giving good pictures most of the time.
 

The Naked Scientists Forum


 

SMF 2.0.10 | SMF © 2015, Simple Machines
SMFAds for Free Forums