Minutes of the Tracker Electronics Meeting

April 30, 2001

 

Present: Robert, David, Masa, Mutsumi, Wilko, Hartmut, and Dieter

 

 

1.    GTFE: Ned submitted chip this morning.

2.    MOSIS Submissions: Robert; Robert stated that MOSIS has an intermediate prototype run of about 100 to 1000 chips. This could cost substantially less than a wafer run, but one has to submit a complete RFQ to find out the cost.  A wafer run in the Agilent 0.5um process costs $107,200 for 15 wafers.  An additional 15 wafers costs $48,200.  These are 6-inch wafers.

3.      Comparator: Wilko; Wilko stated he observed an offset of 500 mV on the test chip, but his results were not stable with time. Wilko tried to use a meg-ohm resistor instead of the bias current source. Dieter believes now that the input polarity of the comparator (clock receiver) may be miss-labeled.

4.    Shaper peaking time: Masa; Masa stated he measured the shaper output with and without the Pico probe. He plotted the delay difference versus the discriminator threshold setting. David then simulated the test chip using the same thresholds and added capacitance to match Masa's delays. The simulation needed an additional 150ff to match Masa's measurements. This is in reasonable agreement with the advertised capacitance of the Pico probe that is 100ff.  We still don't understand the delay.  See Masa’s note below and the attached PDF file.

5.    SEU: Masa; Masa stated he and Mutsumi found that they can cause a latch-up in the TOP5 test chip. They use a camera flashlamp to cause it to latch-up. 3 of the registers are SEU hardened, and one is the Tanner standard cell. Robert stated it would be good to try looking at upsets and verify that the SEU-hardened cells upset less easily.

6.    Test system: Wilko reported that his write-up is nearly done. There will be some effort to speed up the testing. Wilko stated that there could be some additional testing needed. Robert stated he wants to purchase any hardware needed this year. We need a signal generator and probe cards. Robert stated that there might be a conflict with ATLAS with the probe station. SLAC may buy another one, however. Hartmut and Masa will look into the probe station issues.

7.    Chip qualification: Hartmut; Hartmut stated that Nick Virmani needs to sign off on all parts. Hartmut also stated that the SSD qualification document is converging nicely. Nick is working with Hartmut. Robert stated the LET is more important than total dose for the ASICs.

8.    GTRC: David; Plans to submit it to MOSIS on May 7th.

9.    Test boards; Robert; Robert stated that we need to make a 6 GTFE version of the MCM for testing. David stated that this would be done at SLAC. Robert stated that UCSC would make a single GTFE test board. David will make a test board for the GTRC.

10.   MCM: David: David stated there is still an outstanding issue of what kind of gold plating is needed for the surface mount components as compared to the soft gold needs of the wire bonding.

11.   Flex cable: David; The cable 0 flex cable is being layed out by Serge. Cable 7 is an issue because of the narrower than expected routing area at the top connector. We will look into this.

 

 

Masa’s note on his measurements of peaking time:

 

Suggested by Dave, Wilko and I did a couple of measurements on shaper peaking time.  We measured time of rising edges of comparator outputs since the charge injection, with and without a pico-probe on the shaper output pad.  Plots are in a PDF file attached to this message.

Here are the parameters:

 

      AVDD = DVDD = 2.5 V, AVDD2 = 1.5 V, Vref = 0.5 V

      Bias current: 37.7 uA into PVI1 and 2.5 uA into p500nA

      Charge injection: 2.2 fC (200 mV at the pulser, 1/200

      Attenuation, and a 2.2 pF cap)

      Capacitive load: 2.2 pF (grounded through the 50 ohm register)

 

From the plots, I estimated the difference between rising times with/without a probe by comparing times at which comparator output passes 1.25 V (50% of the DVDD voltage).  One set of the cables and the conversion chips was used throughout the measurements, so that cable/chip delays do not affect the time difference between with and without a probe.  Here is the result.

 

=======================================

threshold      14 mV      30 mV      49 mV

---------------------------------------

ch1         12.7      20.3      39.5

ch2         9.7   24.9      38.8

ch3         10.3      22.5      35.9

---------------------------------------

average           10.9      22.6      38.1