Some question for the HFT team

1. There is some confusion about how hits on the HFT are used. In previous presentations, a track reconstructed in the TPC+SVT would include the primary vertex and project through the HFT. If a hit was found within a certain radius of the projected hit the track was considered primary; else the track is considered as originating from a secondary. Is this still the philosophy? Or are hits from the HFT to be used in another fashion?

2. Section 3.3. Simulations. Figure 12. What point resolution was used for SVT, and at what radius?. How do curves look as a function of HFT point resolution (6 um is difficult to achieve in practice; what if it is 12 um?).

3. Figure 13. Why not use real TPC+SVT data events in study of ghost tracks?

4. Same section. How do these results scale if readout of HFT is 8ms rather than 4ms?

5. The HFT requires a SVT-like detector (e.g.) a “pointing” detector. As the SVT already has been dismissed as a candidate for a pointing detector, what is a “minimal” pointing detector that is required? How many layers, what resolution?

6. In section 3.3.1, how are HFT hits used in the D0 daughter reconstruction. It seems that pileup events are not included, so how would these extra events affect the analysis (as ghost hits generally lead to secondary like particle reconstruction).

7. Pileup, pages 42-43. The hit density at the inner SVT quoted at 1 hit/cm2 seems (perhaps naively to us) inconsistent with hit densities of 17 and 72/ cm2 for the HFT over 120 events. Reference 66 explaining the calculation is more of a self-note rather than a clear explanation. We would like to see the discrepancy explained (or reference 66 written better). Also, reference 67 which uses SVT data to get 1 hit/cm2 uses a different algorithm than the SVT hit-finder and reproduces less hits. In the note it says the discrepancy needs to be explained. Has it?

8. Extra hits in the HFT from pileup or other background is a source of real concern and needs to be understood, which is why much of the questions above relate to that. Assuming that extra hits from pileup are correctly accounted, the key is background rates from other sources over 4ms in the radii region of 1.5 to 5 cm (SVT for example, live only 5 us, sees tracks traveling through the plane of the wafers). How is this background estimated and how reliable are these estimates. Can this background be measured with existing data or with a simple device during the next run just outside the current beampipe. Why not use a prototype sensor to measure the background?

Sensors

9. It is not clear that sensors have been used in an experiment. Have they? Tests at the Cern-SPS for the Mimosa-5 are quoted but we would like to see the reference. Similarly, we would like to see reference 70 which is quoted as demonstrating that these sensors work for charge particle tracking. Reference 70 is from a proceedings. Can we get this reference?(similarly it would be nice if the sent us all references which are not easily available on the web are supplied to us by the HFT team).

10. Mimostar-5 is supposedly close to what is needed by STAR, only a little slow and has a low yield during thinning. What strategy does Mimostar-1 use to improve on these?

11. What are the lastest Mimostar test results?

12. What are the yield issues associated with thinning?

13. It’s unclear what the geometry and the I/O of a sensor looks like. Also, how it mates on a ladder to a readout chip. Can we get some drawings that show current thinking on what a ladder of these sensors and readout chips looks like and their i/o to the outside world?

Also

14. Can we get a piece of 50 um silicon?

15. Can we get the appendices?

16. What are the relative alignment requirements between the HFT and the pointing device and to the TPC? How is this to be achieved?