SY-EPC-LPC Cobalt Design

Design Notes

Design / Project Notes

(2014-05) Rack DC Earthing:
- After a discussion with Jean Paul, it appears that implementing a DC disconnection system (multicontact pins) was too expansive and difficult to locate in the rack. Moreover, Jérémie B. from magnet side confirmed that no real need for disconnection the DC circuit exists.
- Then, simple forks for earthing DC Busbars, located at the rear bottom of the rack will be designed and implemented.

(2014-05) DCCT calibration:
- After a discussion with Michele, it appears that only the I2V card need to be recalibrated, not the DCCT head.
- Then, no need to foresee a system to disconnect easily DC cables (to inject a known DC current to calibrate the DCCTs).

(2014-01) FGC3 & DCCT common PSU, leading to a optimized FGC3 chassis. Reasons are:
- After a discussion with Miguel, it appears that only the FG3 analog part will take power from the +/15V lines also used by the I2V card.
- It is then possible to use the same FGC3 PSU, powering both FGC3 and I2V card without any risk of common mode noise between both systems.

(2013-11) Blocking signal not implemented in SKCMD.
- It was decided not to implement this signal since this signal is not present on commercial power sources also, and since without it, the strategy of developping a FGC3 capable of controlling a LHC-type covnerter remains valid (2013-11). Moreover, it appears that on a standard 1-quadrant power converter, the blocking signal which ensures the power converter doesn't try to react to induce voltage in its magnet (load) can be obtained simply requiring a null voltage to the power source. Blocking signal is indeed used to let the converter in a "idle" mode, where it is running to be ready to run on demand, while in blocking mode, voltage and current in the load can happen (magnetically induced by operation conditions, and local magnets being, them, operating) without leading to any counter effects from the converter which simply doesn't react. Magnet of a blocking mode operating converter can then be assumed to a secondary of a transformer with less active load (never forget that a 1-Quadrant power converter will show diode in one direction, and high value capacitors at the output, reacting like a short circuit for given frequency signals being developped.

(2013-xx) Polarity Switch Unit will not integrate the Power Module specification, since
- Only some are needed for hie-isolde project.
- It add more complexity to the specification.
- This unit could be developped by CERN referred to low quantities required, and why not to standardize this unit.

(2013-xx) Number of DCCT was set to 2, mainly
- To reproduce a LHC environement
- To profit from the same approach and advantages which led to this decision in the LHC

(2013-xx) FGC3 chassis should ideally contain 3 FGC3s.
- I2V card will adapt its size from 6TE to 10TE depending on the required space needed.

(2013-xx) FGC3 chassis will be based on a dedicated multi-layer backplane PCB collecting the FGC3 + I2V connectors on one side, and providing Burndy connectors on the other side.
- Location of these Burndy connectors can be or in the 3U height, or in the bottom 1..2U "free space PCB" of a potential 4U..5U electronic backplane (then 3U for FGC3 signals + 1..2U bottom for Burndy collecting signals), taking in account the 1..2U required for fan cooling of the FGC3.

(2012-10) Place of the DCCT was chosen not to be between Power Module and Polarity Switch Unit, mainly
- To ease their mechanical integration.
- By principle, and then to measure the direct magnet current.
- A top rear location allows to re-direct air flow not to heat electronic chassis, integrating the DCCT also out of hot air flow.
- Connection cables lengths from electronic chassis to DCCT are minimized.

(2012-08) Power Module will be as closest as possible from LHC power Module,
- Effort required to make FGC3 matching a LHC type converter can be further re-used in the case of a FGC3 is required to control a real and "old" LHC Power Converter.
- One should never forget that from the R2E project, we should get back many FGC2, and Power Converters, sometimes with a need to mix FGC3 + LHC Power Module for internal use, or new project.

(2012-07) Choice of DIM_LHC type was dictated by
- A complete package exists: spares in sufficient quantities, specification available.
- If low speed on analog values is critical, RAD-DIM can replace it in some years, with higher speed.
- DIM solution remains a very good solution providing a simple interface for manufacturer, with the 1st fault function implemented.
- The solution is very well known (LHC) and very well supported in the group.

(2012-07) Commercial devices were considered once, then forgotten, leading to a specification converter to be written for our specific purposes. Reasons are:
- This range of voltage current can be strategic for EPC group.
- Commercial solutions were too far from our current needs.