12/15/97
                                                    B.Baldin
                                                    D.Denisov
                                                    K.Johns
                                                    G.Alexeev
                                                    V.Tokmenin

        MEMO on AMPL-8 and DISC-8 INTEGRATED CHIPS DEVELOPED
        FOR D0 MINI-DRIFT TUBE AMPLIFIER-DISCRIMINATOR BOARD

1. Introduction. 

    A set of AMPL-8(2) and DISC-8(1)  chips designed for muon mini-drfit
chambers at Minsk semiconductor factory has been tested by JINR group in
Dubna and at  Fermilab in  November-December 1997.  The results of these
tests have been discussed at muon  upgrade meeting and several decisions
and requirements have been  considered and recommended for this project.
The following is a description of the proposed steps.

2. Amplifier AMPL-8.

    The second iteration of the AMPL-8 design has the following
features:
            Amplifier gain          - 120 mV/uA
            Output noise, RMS       - 12 mV  (0.1 uA @ input)
            Rise time               - 33 ns  (Note 1)
            Fall time               - 160 ns (Note 1)

    Note 1. Measured at standard 10-90% levels with the square pulse
            applied at the input.

    AMPL-8 also  required an  external operational  amplifier to balance
the output  DC level.  Limited  bandwith and  missing DC  balance ot the
AMPL-8 do not allow us to consider this chip as a possible candidate for
mass production. It is obvious that another iteration of the chip design
is necessary.

    A decison about mass production  of this chip can only be made after
the third iteration of it would have demonstrated the following features
(assuming that the other parameters remain within existing spec):

            Amplifier rise and fall times       20..25 ns (Note 1)
            DC balanced output without 
            external components
            Amplifier gain                      100 mV/uA
            Noise level                         < 10 mV (0.1 uA @ input)

    Note 1. Measured at standard 10-90% levels with the square pulse
            applied at the input.

    It is expected that JINR group  will provide necessary specification
to the  Minsk factory  and will  test the new  version of  the AMPL-8 in
three month time period from now. In order to back up the whole project,
another  version  of the  amplifier  based  on  discrete  transistors or
transitor arrays  has to be  developed and tested by  JINR group by that
time as well.  The decison  on final  version of the  amplifier for this
project will be made as  soon as the  information on the test results is
available, but no later then April 30, 1998.  Muon electronics group may
provide assistance in developing a  schematic design for back up version
of the   amplifier. JINR  group has  to  inform muon  group  about their
preference on this matter no later then January 15, 1998.

3. Discriminator DISC-8.

    DISC-8 chip has satisfactory parameters and can be mass producted.

4. Packaging.

    The decison was made to request  smaller 48 pin plastic package with
0.635 mm (0.025") pitch for both AMPL-8 and DISC-8 chips to reduce their
size and  allow better  PCB layout.  JINR group  will  provide necessary
specification  to the  Minsk  factory which  will guarantee  use of high
quality export plastic material for packaging.

5. PCB layout.

    The  layout of  the PCB  (though  demonstrated  stable  operation in
typical  experimental environment)  has to be changed  as following. The
location of the chips should follow  the signal flow which will minimize
amplifier input and output traces. The size of the SMD components has to
be reduced to 805  instead of 1206  used in the  present PCB layout. The
new  layout has to  be done  in a way to  minimize  input  traces to the
amplifier  inputs. The  discriminator  chips have to be  placed near the
output connector with short traces  connecting to amplifier outputs. Any
trimming elements have to be  eliminated, if possible, by replacing them
with low  tolerance resistors. A  two side component  placement (top and
bottom) is preferable to reduce the board size.

6. Reliability.

    The ADB components have to be chosen very carefully to guarantee low
maintenance operations during  several years. Some of the ADBs will work
in no access zones of the D0 detector and may not be replaced during six
months period. Special attention  has to be paid to the selection of the
materials for  mass  production of the  AMPL-8 and  DISC-8 chips. Remote
testing  and  monitoring  has to be   implemented  into  mini-drift tube
octants as a part of the ADB design.