return to Top specifications

Camac 460 scaling for Tevatron DFGs

A "normal" dipole is defined as a dipole with a +50 amp power supply and a transfer constant of 3.4 kG-in/amp. (Design transfer constant of the Tevatron correction dipoles is 3.62 kG-in/amp, but we have always used 3.4 for our scaling in the past.) This allows for a bend angle of .1265 mrad at 1 TEV at full current. There are two different types of dipoles that will be adressed as well. At the 11 and 49 locations around the low beta inserts, there are 75 amp regulators attached to the correction elements. These will be referred to as "high current" dipoles. Also, at the 12 and 48 locations around the inserts, there are stonger dipoles (larger transfer constant). These will be refered to as "strong" dipoles.

The output at a given time from a 460 card is a combination of 3 terms of the form V = sf * M1 * f(M2). In this equation, V is the input of the dac which has a range of -32768 to +32767. sf is a scale factor that can take on discrete values of n/256 where n is an integer between -32768 and +32767. M1 can be either a constant supplied by the user, or an MDAT value read from the link. f(M2) is a term calculated by interpolating a table supplied by the user. The table can be a function of time, or a function of an MDAT frame that is read from the link. At a given time, the DAC output is calculated from 3 such terms where one term is a time table, and 2 terms are MDAT tables.

The MDAT frame marked by M1 in the equation is almost always frame 10 for the Tevatron which is the programmed current for the main bus. Full scale for this frame is 5000 amps which is equal to 65536 in raw data. MDAT receiver hardware at Fermilab traditionally divides all unipolar MDAT frames by 2, effectively making 32767 the full scale value corresponding to 5000 amps. The term M1 used in the card is this 16 bit number divided by 256 (or the most significant 8 bits). The valid range for a table entry is -32768 to +32767. Knowledge of the valid data ranges in the 460 card and the operating conditions foreseen in the Tevatron is used to establish the optimum scaling for the 460 card devices.

A dipole power supply should be able to output +50 amps at all energies that the collider will actually run. Since the largest table value allowed is 32767, a scale factor of 15/256 would create a 50 amp output at the injection energy of 150 GEV. This implies that a table value of 4920 would create a 50 amp output at 1 TEV.

The PDB for the settings on the main device will be x' = 0.1265 * x / 4920 to give the proper conversion from raw data to mrad.

At the present time, we are administratively limited to 50 amps on all correction elements. We choose a scale factor of 10/256 for these supplies and all other scaling remains the same as for a regular dipole. A table entry of 32767 at 150 GEV will produce a DAC output of 21845 which gives 50 amps from the power supply. With this choice of scaling, the supply behaves to the users as if it were a 50 amp supply. In the future if we are allowed to operate these supplies at 75 amps, we will change the scale factor back to 15/256 and adjust the scaling PDBs for the settings of all the ramp devices so that the maximum table value at injection energies would be 1.265 mrad rather than .8433 mrad.

The strong dipoles should have a maximum output of 50 amps with the largest allowable table entry at 150 GEV. We set the scale factor to 15/256 to accomplish this. With these dipoles though, 50 amps at 150 GEV corresponds to an angle of 1.38 mrad rather than .843 mrad with the regular dipoles. The PDBs for the settings of the table devices will be changed to reflect this difference in strength. A scaling of x' = 0.207 * x / 4920 will give the correct conversion for these devices.

Notes about using mrad as fundamental engineering units:

Since 4920 is the table value that corresponds to 50 amps at flattop, the resolution in dipole current will be a factor of 6.5 times worse at that energy than if we did not use MDAT scaling. The lower resolution is still OK with the people planning the position feedback control at the IPs.

If we want to load a card so that MDAT scaling is not used, the engineering units will not make sense. If the user wants 20 amps in a certain slot, they will have to enter the value that corresponds to 20 amps at 1 TEV (provided the scale factor gets set to 1).

Between breakpoints the output of the card can have unexpected results. If two energy slots contain angles that produce the same current, the output of the card does not go linearly between them. This has always been the case with the 160 cards as well. The diagram below shows a normal 160 that has a two slot energy ramp. The 150 GEV slot is .421 mrad and the 800 GEV slot is .0791 mrad. Both of these slots correspond to 25 amps, but it can be seen that the output of the supply goes close to 50 amps between these break points. This is not usually too much of a worry since ramps are not normally built this way, but it does add some complications in figuring out how the software overcurrent protection should be implemented.

Updated: 8-Jan-1998