0. Turning on discriminators made noise go away.  Only 2 channels
   remain noisy: Module 17 channel 14 and 15. But they correspond to
   beam strobe and detector strobe.  So actually there are no noisy
   hits.

   from rd_map:
   module  channel (0..31)
     17      14           0    1   22         Special       Beam strobe 
     17      15           0    2   22         Special       Detector strobe

1. events 1-100 module 17 channel 16-31. looks OK

2. Let's check a bad module.  Slot number 33 (0x21) - presumably has a
   wrong TDC ID bit and report itself as module number 35 (0x23).
   (rstdc_33a_bad.hbk). That module has little yellow tag on it with
   ID=544192

   Header is indeed wrong.  Took off the module and plugged in a new
   one.  New module produce header 43 instead of 31.  Same
   story.... (rstdc_33a_badagain.hbk) I believe it is not the module,
   but CAMAC crate.

3. Trying channels 0-15 on module 24. They had double peaks, separated
   by 256 counts (rstdc_24a_bad.hbk).  Took that module out.  It has a
   yellow tag with ID=540372.  Took module with "bad TDC header"
   (ID=544192) and plug it to be module 24. (rstdc_24a.hbk &
   rstdc_24b.hbk)

   New module report itself as a module 24.  So TDC header is actually
   OK.  That makes me more confident that error is in the CAMAC crate,
   not in the module.  

   Channel 0 and Channel 16 has very different stop time compared to
   other channels (again 256 counts).  Start time for all channels is
   OK.  Let it be for a moment.

4. Testing module 26.  It has 256 TDC counts double peak problem.
   (rstdc_26a.hbk & rstdc26b.hbk). Channel 0,1,15,16 has 256 shifted
   start/stop position compared to other channels.

5. IMPORTANT: I noticed that cables in modules 27 were plugged in
   WRONG They were shifted by 1 pin up.  So channels 19M10 and channel
   19M18 would get ground instead of pin 15/31. Plugged them in right
   position.  I though it could help module 26 in it's strange
   start-stop pattern.  It did not.

   Found similar problem in upper connector on module number 00.
   Chennel 15 (1M10) was getting ground instead of the signal. Someone
   tried hard that bend the pin and almost broke the connector. 
   
   I looked on all channels.  Overall 3 wrong plugged connectors were
   found.

6. I'm sick of bad modules.  Let's check a good module for a change.
   I picked module 27.  Well this module indeed is good.  No funny
   start/stop patterns.  Just good and reasonable.  For the sake of
   speed I plugged 2 16-bit cables and tested all 32 channels
   (rstdc_27.hbk)

7. Check bad module number 28. Again it produced 256 TDC counds double
   peaks.  Again chanel 0,1 have strange start/stop pattern.
   
8. Testing ALL modules 500 pulser events each.  

   WARNING: module 12 has pin for channel 13 BENT, so that it missed
   the connector.  BAD.... Fixed.

   Module 12: 2 peacks in start/stop separated by 20 nsec. 
        channel 0 (9P3) is missing 
        channel 1 (9M3) has bad stop signal
   Strange, they looked good on my map.

   Module 24:
        chanel 0 nas bad stop signal
    
   Module 26:
        channel 0,1 has bad start/stop signals

   Module 28:
        channel 0,1 has bad start/stop signals
        Very similar to Module 26

   Module 31
        looks like some lower bits are missing

   Module 32 
        looks like all channels are shifted

   Module 33
        Wrong header (43 instead of 33).  Looks like some bits are missing

   Module 34 
        looks like channels are shifte

   Module 35 
        tdc header is wrong (module=41, instead of 35)

12 001100
13 001101
14 001110
15 001111
30 011110
31 011111
32 100000
33 100001 read as 43 101011
34 100010
35 100011 read as 43 101011

   Moved all the files with the original scan to the direcotry orig_scan
   Took module 12 out it's ID=540641. Put in instead spare module