In order to calibrate and tune the crystal spectrometer before the measurement a source of X-rays of well known energy and sufficient intensity is needed. Presently the best realized one is an ECR source. This source consists of three main parts, a magnet providing a mirror field, a hexapole magnet and high frequency source. As mirror magnet the cyclotron trap II is ideally suited. A high frequency supply existing at PSI at the Phillips injector cyclotron (10.12 GHz, 1.5 kW) could be used as a stand by as long as no other more powerful source is available. The hexapole magnet has to be bought from industry. The technical expertise to set up and to run the ECR source exists in the institute of one of the members of the collaboration (ATOMKI Debrecen).
We propose to observe the X-rays from the ECR source in a first step directly with the CCD detector in order to determine the efficiency of the X-ray production. In this way a minimum temperature of the ECR plasma will be found for a still sufficient yield of the one-electron atoms.
In a second step a diagnosis of the line width as a function of temperature
is necessary. This can be done with the available 
Bragg crystals.
The response function of these crystals is well enough known for this purpose.
In a third step the crystal spectrometer with the two-crystal set-up
of 
will be used.
It is planned to have 2 measurements with the ECR source of one month
each in the NA hall. We assume here a number of 
produced useful
X-rays per second. These measurements could take place at the end of 1998 at
the very earliest but are more realistically planned for spring-summer 1999
after the measurement of the pion mass has been finished [86].