SSD Software Error/Problem List


The page is used to keep track of problems found with the SSD software.
Light green entries are problems that have been solved or for which an explaination/solution has been found.
WhoWhenSummaryCodeDescription/Solution
Lilian11/17/04Clusters with the same starting stripStSsdPointMakerThe cluster splitting in StSsdwafer.cc seems to gives wrong results in the case of large clusters. In the following example, the cluster 0 is splitted in 3 clusters having the same first strip (433).
------------------------------>List of strips
 id,sig,noise : 433 9 3.5
 id,sig,noise : 434 7 2.9375
 id,sig,noise : 435 7 2.875
 id,sig,noise : 436 9 2.9375
 id,sig,noise : 437 7 2.9375
 id,sig,noise : 438 9 2.875
 id,sig,noise : 439 7 2.625
 id,sig,noise : 440 7 2.6875
 id,sig,noise : 441 9 3.0625
 id,sig,noise : 442 6 2.75
 id,sig,noise : 443 7 2.9375
------------------------------>Clusterisation done
N,Size,FirstStrip,StripMean,TotAdc,FirstAdc,LastAdc,TotNoise : 9 11 433 437.857 84 9 7 24
------------------------------>Splitting done
N,Size,FirstStrip,StripMean,TotAdc,FirstAdc,LastAdc,TotNoise : 28 2 433 433.438 16 9 7 5
N,Size,FirstStrip,StripMean,TotAdc,FirstAdc,LastAdc,TotNoise : 29 5 433 434.714 35 9 3 10
N,Size,FirstStrip,StripMean,TotAdc,FirstAdc,LastAdc,TotNoise : 30 7 433 435.964 55 9 7 15
N,Size,FirstStrip,StripMean,TotAdc,FirstAdc,LastAdc,TotNoise : 31 11 433 437.857 84 9 7 24
Due to a bug in the StSsdClusterList::splitCluster method. Within the loop over the maxima found in the cluster, the following line
//      localFirstStrip = CurrentCluster->getFirstStrip();
was reseting the current FirstStrip to the initial first strip of the big cluster. When commented, the result look like this :
------------------------------>Splitting done
N,Size,FirstStrip,StripMean,TotAdc,FirstAdc,LastAdc,TotNoise : 28 2 433 433.438 16 9 7 5
N,Size,FirstStrip,StripMean,TotAdc,FirstAdc,LastAdc,TotNoise : 29 3 435 435.789 19 7 3 5
N,Size,FirstStrip,StripMean,TotAdc,FirstAdc,LastAdc,TotNoise : 30 3 437 438 23 7 7 6
N,Size,FirstStrip,StripMean,TotAdc,FirstAdc,LastAdc,TotNoise : 31 4 440 441.448 29 7 7 9
<------------------------------
There is still something to look at since the signal (7 adc counts) on strip 437 is shared by the clusters 29 (3 adc counts) and the cluster 30 (7 adc counts) which gives a total signal bigger that the original.
Lilian11/16/04Underestimated number of photoconversionssisdgeo.gA.Wetzler and Max are reporting a number of photoconversion from simulations smaller compared to real data. With on old geometry, the number was 50% lower. With the updated geometry, the ratio is about 70% more or less like the SVT. See this message
To be investigated
Lilian11/16/04Volume positioned twiceStiSsd Claude Pruneau is reporting that some SSD volumes (ladder and hybrid) are positioned at the same location. See this message and replies
To be investigated
Lilian11/16/04Wafer orderStSsdDaqMaker, StSsdPointMaker At some point the order of the wafer has to be flipped in the SsdDaqMaker to get the hits back in the StSsdPointMaker :
if (id_side==0)
	 id_wafer=7000+100*(16-(my_channel/768))+ladder;
else
	 id_wafer=7000+100*((my_channel/768)+1)+ladder;
The data reading sequence in the StSsdDaqMaker and the standalone SSD reconstruction code look different. After a careful comparison, the results are virtualy the same. The StSsdDaqMaker code can be modified to simpify the data reading. Would be done once the code is officialy committed in CVS
Lilian11/16/04This is the summaryStSsdXXX This is the description of the problem
This is how to fix it...

Lilian MARTIN
Last modified: Fri Feb 11 18:49:42 CET 2005