Proposal B Meeting: Radiation Damage Studies of OE Components for the SLHC
SLHC Optoelectronics Proposal B Meeting, 30 April 2007<o:p></o:p>
<o:p> </o:p>
Attendances: Tony Weidberg, KK Gan, Jingbo Ye, Cigdem Issever, Todd Huffman, Karl Gill, Pat Skubic, Rusty Boyd, Francois Vasey
<o:p> </o:p>
Duration: 90 min
<o:p> </o:p>
We decided to discuss Karl’s (EEL) and KK’s (VCSELs and Si p-i-n) documents and how we would merge them together. Karl and KK summarized at the beginning of the meeting their documents.
<o:p> </o:p>
The sample size was discussed and Karl pointed out that the sample size depends on how much risk one is willing to take. In case one observes a failure increasing the test sample size could save sometimes a lot of money. We agreed that the sample size for the pre-production tests will be 20 devices and for the market survey phase 4-5 or as many one can get.
<o:p> </o:p>
Jingbo commented that it is important to measure the mean life time, spread and random failure rate of the devices. He also pointed out that it is very important to determine the junction temperature of the devices and not the ambient temperature.
This was also strongly supported by Rusty, Karl and Francois.
This issue was discussed for quite a while and the group agreed that junction temperature (hence the thermal resistance) of the devices must be measured/determined in order to be able to compare results. This should be done not only for the pre-production phase but also for the market survey phase. It is however not necessary to measure the thermal resistance during irradiation.
<o:p> </o:p>
Karl agreed to send the group a publication about how to determine the junction temperature. (will be uploaded to the agenda page.)
<o:p> </o:p>
Francois raised the issue of NIEL. He reminded the group that ATLAS and CMS had in the past not been able to agree on whether to trust NIEL or not. This was the reason that LHC radiation test results could not be easily compared between ATLAS and CMS. CMS decided not to trust NIEL because of the complicated structure of their components and measured the NIEL factors. ATLAS did not measure the factors but did some spot check measurement at different sources and energies and used NIEL factors found in the literature.
<o:p> </o:p>
Karl stressed out that in case of non-linear behaviour one does not have any other choice than to measure the NIEL factors. They observed this for example with p-i-n diodes.
<o:p> </o:p>
On Tony’s suggestion we agreed on that for simple GaAs structures we would use the calculated NIEL scaling factors but make at least one cross check with different beam type/energies. For quaternary structures like InGaAsP for which there are no reliable NIEL calculations, we should do radiation tests at different beam energies and types to determine empirical damage factors.
<o:p> </o:p>
Tony pointed out that the SLHC luminosity will not be 10 times the luminosity of LHC. Here is a direct cut and paste of his email he sent as agreed to Cigdem after the meeting:
<o:p> </o:p>
“Dear Cigdem, all the ATLAS fluence calculations were done assuming 3 years operation at 10^33 and 7 years operation at 10^34. A year was defined to be 10^7 seconds so this corresponds to 730 fb^-1. This is very optimistic for several reasons
(1) the luminosity will not suddenly increase from 10^33 to 10^34 but will take a few years.
(2) the luminosity lifetime will be quite short at LHC, so the effective number of seconds in a year will be less than 10^7,
Then on top of these implicit safety factors all fluences were scaled up by an explicit safety factor of 1.5
Nobody believes that ATLAS will get 730 fb^-1 so the Physics studies have been done assuming 300 fb^-1.
Therefore ATLAS has decided that for the upgrade we should demonstrate radiation hardness for a luminosity of 3000 fb^-1.
I suggest that we go with this and so the fluences from LHC should scale up by a factor of 3000/730.
Regards
Tony”<o:p></o:p>
<o:p> </o:p>
At the end of the meeting Francois wanted to know what the aim of our group for TWEPP in September is, whether we want to come up with a plan or just agree on common radiation protocols. Based on our agreed meeting frequency at our last meeting the only realistic target we could have reached is to come up with common radiation protocols but even this is cutting it tight. He questioned the existence of a SLHC radiation effort if people would not be willing to meet up more frequently. Furthermore he pointed out that it is important to include for each device an estimate of how much resources are needed in order to be able to come up with a plan and also to be able to ask funding agencies for funds.
<o:p> </o:p>
We agreed to meet every two weeks instead of every 4 to 8 weeks and that we will include for each device an estimate on the amount of resources needed in order to perform its full evaluation.
<o:p> </o:p>
Karl and KK will merge their documents and will include the agreements we reached at this meeting into them. The merged document will be sent to the group.
<o:p> </o:p>
Next meeting will be chaired by Todd, because Cigdem can’t attend it. We will discuss the fibre radiation protocol. Todd agreed to include the comments he received from Jan and Karl into his document and to distribute it a week before the next meeting.
<o:p> </o:p>
We agreed that our next meeting will be on the 21.05.07 at 3pm CERN time.