1. Henning E. Larsen, INFN h.larsen@risoe.dk 12 July 2006
FP420 Low voltage supply
Henning E. Larsen,
INFN
12 July 2006

2. Functions for a superlayer supply
Superlayer consists of..
Detectors (Atlas pixel detctors),
Front-end pixel-readout (Atlas Pixelchip),
MCC,
Optical transmitters
PT1000 temperature sensor
Each superlayer will be LV (+HV?) supplied by a separate unit=>
Electrical isolation
Redundancy
Modular service
Isolation implies that it will have its…
own ac-ac power transformer secondary winding
own AC-DC rectifier bridge
digital communcation interface isolated
Linear regulators for radiation hardness (LHC4913 STM tol 5kGy) (2.0V and 1.6V, 5W)
Remote controlled fine adjustable offset and shutdown.
Monitoring of the voltage, current and local heatsink temperature
Monitoring of the superlayer frontend temperature (PT1000 sensor)

3. Front-end assembly
Pixels:50x400um and 400x50um
From Ray Thompson

4. Bias supply segmentation
PT1000 Temperature
Make segmentation to account for different radiation damage. Damage depends on distance from the beam
From Ray Thompson

5. LV supply blockdiagram
Candidate for the crate controller is the solution from ALICE DETECTOR CONTROL SYSTEM (DCS) PROJECT
Monitor ADC: Use of the integrated Detector Control Unit, DCUF used in the CMS central tracker for the monitoring of some embedded parameters like supply voltage and currents on the front-end read-out modules. Digital interface is I2C which only reach a few meter so the controller must be close by
Control DAC: Currently no suggestion for a DAC.
Local control: Rad tolerant FPGA from

6. LV+HV Crate for 1 station with 8 superlayers=16 hybrids=48 detectors

7. Open issues Location of support electronics in tunnel
Communication interface module in LV supply?
Communication medium to slow control system: Fiber or cupper?
Supply LV and HV requirements (Volts, Currents, Stabilty, Ripple)

8. Location
If the LV electronics should stay within some 20m from the detectors, there are only two possible locations (ref Daniela Macina):
Below the new cryostat, where the radiation level is estimated at about 700 Gy per year of running at full luminosity;
Below or near the adjacent magnets, where the radiation is much lower and estimated at about 15 Gy per year, but where there are already other things.

9. Critical issues for support electronics (LV and HV)
What is the infrastructure and environement in LHC:
Location of support electronics in the tunnel
Radiation levels at these location
Suitable electronics to survive there
Temperature range
Data communication
To quote Scott: The infrastructure in the tunnel is critical: Where is room for service electronics, and what is the environment there?
We should look to what has LHC machine done to solve these issues for their support electronics. Posibly use their methods if appropriate. For slow control we have no problems with speed.

10. LV Block diagram
This shows a blockdiagram of the LV supply. Modules must be located some meters from the frontend for reasons of space and radiation protection. All electronics should support the environment in the LHCtunnel: Radiation and temperature.
The frontend supply (2.0 and 1.6V, <10W) is regulated with a linear regulator. It has remote sense to compensate for line drop. It is likely to be required to adjust the set-voltage remotely during operation. One candidate for regulator is LHC4913 STMicroelectronics. Used in Alice. For reasons of redundancy and electrical isolation, one module for each detector plane or super-plane.
Monitor ADC: Use of the integrated Detector Control Unit, DCUF used in the CMS central tracker for the monitoring of some embedded parameters like supply voltage and currents on the front-end read-out modules. Digital interface is I2C which only reach a few meter so the controller must be close by
Local processor: Some local intelligence is needed to simplify the communication and perform local house-keeping. Interface to I2C. Suggestions: FPGA’s from Actel, Atmel or micro controller from Aeroflex
Control DAC: Currently no suggestion for a DAC
Optical interface: Gigabit Optical Link (GOL) (Cern EP/MIC) is probably far to complicated to interface to, so something better has to be found
Located in shielded location some meters (3..20m) from frontend
One module per plane for redundancy and noise reasons
Remote monitor of V,I and T
Remote control of deltaV and power ON/Off
Local over-temperture protection

11. Prototype test
Prototype to test the analogue part of the LV and HV supplies for one superlayer or at least a part of a superlayer. This involves design and construction of
Control, montoring and protection circuitry using real cable lengths
Demonstration of its stability and noise properties in a test rig.
No prototype of digital part of interface to slow control : Whats new here is
slightly higher than usual radiation load
longer distance to the counting room
But we expect to be able to reuse interfaces developed by other experiments so prototyping is less relevant for this part
Michele: First you should put up a slide explaining the system in broad terms, of course. I assume you have that by default.
It is important to prototype the analogue part of the design because in that part there are more unknown factors which can influence the system. Factors which can be hard to predict and account for by design. A prototype is a good way to deal with these issues.
The supplies should be segmented in a manner reflecting the segmentation of the frontends. In this way the whole system can be made more fail safe. If one detector, frontdend or supply fails, only the segment where it resides will have to be shut down. In addition this architecture naturally invites for a design with isolation between the differnt parts of the system. This is a critical system design issue to combat noise and crosstalk.
We suggest to prototype at the essential parts of such a segment supply. This will mainly be the analogue part but it will also include digital parts in the form of ADC and DAC’s for control of the.
The interface to the slow control which is purely digital is not considered essential to test as it is expected that elements from other experiments can be adapted to fp420’s requirements
We plan to make a prototype which will demonstrate the design of the supply HV/LV for one segment of the detector.
Ad7294 adc/dac/highside sense chip
LTC6101HV up to 100V high side sesnse

12. Alice DCS
Alice slow control interface board.

13. Resources 6k€ for components
4 man months of technician work for design, purchase, construction and test
1 man month of technician for print circuit board layout
I think thats more or less what we agreed on the phone?

14. Notes:
Current to voltage shunt converter: ina19X from ti.