Re: Mark IV Checkout Continuing - 1

[Tom Droege <droege@FNAL.GOV>]

Fletcher Henderson and I are moving ahead on the checkout.  One of seven
major boards is now completely working.  Only 10 little errors.  By little
I mean something that can be fixed without moving any parts.  i.e. a line
on the wrong side of the board that shorts another line.  

>From my PC program I can now say "Read the Analog Multiplexer Lines" and 
32 calibrated voltages appear on the screen.  I am using a 16 bit ADC for
this purpose and it seems good to about 1+ bit rms.  Just what one expects.
I have not done any linearity checks, but there should be no problems.  

The analog multiplexed measurement system is completely "unnecessary" for 
the functioning of the system.  It purpose is to allow continuous checks 
that everything is working as expected.  This is not the system used for
pixel read out.  There will be four more ADCs for that.  But I expect the
same performance from them.  i.e. sigma of the electronic measurement good
to 1 part in 65535 or so.  This should be much better than the sky 
background which will be 25 or so ADC counts.

>From the PC program I can now also load any of 16 DACs with 8 bit 
resolution.  These will set things like clock levels and the trim on the 
RA rate.  Many of the DACs can be read back by the analog multiplexer 
system.  This gives a great check that everything is working as expected.
Also from the PC program I can send out 24 control pulses.  Some of these 
are used to load registers, some can control model airplane servos, etc..  

Things are going much faster now that I have both ends of the PC to Stamp
loop working.  There is now sample code that works at each end, and I 
only have to duplicate sections to add additional functions.  i.e.

The goal is to get all the major printed circuit boards tested, and a stack
of production boards on hand for this summer when I will have the power of 
two Watsons (two each Merle Watson Sons) to do production.

The last of the prototype boards (the memory board) arrived today.  I now
have everything (almost) needed for the pre-production camera.  I still
need a machined part before I can build one of everything.  

I worry a lot about the details of how things can go wrong.  Even so, I 
managed to design a system where I could plug a connector one pin off, and
thus put +10 volts on poor Stamp #3.  RIP!  The design is not going to be 
foolproof.  One will have to follow instructions, and make cable changes
carefully.  OK, I will try to put in some protection, but it will still be 
possible to burn up a lot of stuff by being careless.

Tom Droege