🚀 go-pugleaf

i have a problem with the general pipe:  {n,}eqn|tbl|{p,n,t}roff.

input to the sequence is file.n where file.n contains a table inside
a memorandum macros display (with vertical bars in the tbl format
section) and equations within data elements of the table.  i've
tried beginning the processing with tbl before {n,}eqn and results
are worse than the sequence shown.

the vertical bars indicated in the tbl format fail to come through
correctly, if at all.  (when using the nroff sequence, i pipe through
col before viewing results.)  is this kind of request a basic conflict
between eqn and tbl?

below is an example file.n (hard tabs are shown by \t):
=======================================================================
.nr O 1.0i                              \" SET PAGE OFFSET AT 1.0 INCH
.so /usr/lib/tmac/tmac.m                \" CALL THE MEMORANDUM MACROS
.EQ                                     \" SETUP FOR EQN
delim ##
.EN
Table:
.DS
.TS
box;
l | l | l
.
=
head1\thead2\thead3
=
.sp
#dW over dt#\ttwo\tthree
.sp
#psi#\tfour\t#delta#
.sp
=
.TE
.DE
=======================================================================

thanks for any help,
sid
=======================================================================
s cowles                         att:  +1 415 423 0929
lawrence livermore national lab  uucp: {backbone}!lll-lcc!scowles
environmental sciences division  internet:  scowles@lll-lcc.llnl.gov
p. o. box 5507; L-524
livermore, california  94550
=======================================================================