I acquired this in a trade.  It's supposed to work, but haven't had a chance to do anything with it because of the other 11's I've been trying to get working.



I finally powered up the 11/20.  I didn't have a key for it, so I'm borrowing the one from the 11/45.  I have to find a source for these keys.

After powering it up and trying out a few things, it's obvious that there's a problem.  It looks like a core problem but worse than what I had with the 11/10.  This just doesn't seem to want to do anything rational.  Poked around a bit but it doesn't look good -- grrr.


I took the 11/20 over to a friend's (he has one and some spares) to see if we couldn't isolate the problem.  All the boards (there are a lot of flip chip modules in this thing) seemed to work OK.  So we're checking out the core.  We swap core from his 11/20 to mine and that seems to be happy.  We check out the M1091 and are a bit suspicious but we can't find any documentation that seems to indicate how to jumper it properly (unusual for DEC).  We finally get what we think is something rational when we get a stuck bit (bit 15).  Damn.  Fighting too many problems.  He's got a spare front panel but it's not with his other stuff and it's late so we put off trying to see if we've solved the problem with the 11/20.


I get the spare front panel board (it's rough -- missing switch covers, etc) to test to see if the front panel is where the stuck bit is.  It appears to be but toggling in stuff is hard due to the missing switch covers.  But the 11/20 seems to work and not have a stuck bit.

When I examined the failed front panel board, I noticed that one of the 2.2K resistors was cracked.  Upon further examination it was broken in two.  To see if that was the problem, I put the bad front panel board in the 11/20 and jumpered a 2.2K resistor from the other board to it (I didn't have any 2.2K resistors on hand).  With the jumpered resistor in place, everything seemed to work.  So the plan is tomorrow to go out and get some 2.2K resistors and repair the front panel board.


Who would have guessed that it would be so hard to find 2.2K resistors.  No one seemed to have any!  I finally found some at Radio Shack (a desperate last resort).  Removed the remains of the old resistor and soldered in the new one.   And the 11/20 is working!  Now I have to find (or make) a G3000 so that I can put an SLC in it and have it use RS-232.  If I build that adapter for the 11/10, then I could do a variation on that for the 11/20 that just has the level converter and not include the baud rate generator.


Locating memory for the 11/20 and the fact that it requires a bit of work to get RS-232 working in it, I decided that adding a BA11 to the the 11/20 would be the way to go in order to get it into a functional state.  I found another BA11.  I thought it would be best to keep the "flavor" of the '20, so I wanted to put core memory into it.  This involved putting in an MM11-UP.  Since there was already 4K of core in the 11/20 already, I just added an additional 16K bringing the total to 20Kwords.  This should be enough to run RT-11!  I also added to the BA11 an M7800 SLC which would give me a console.  I determined that a 6' BC11 UNIBUS cable would be long enough to go from the 11/20's chassis to the BA11.

All was hooked up and power applied.  After finding a couple of gremlins I was able to put characters out on to a terminal.  I also wrote a little memory test program that allows me to check memory (this was actually necessary to get the memory going).  I was glad I retained the original 4K to contain the test programs.

I had also put in an RX11 controller hooked to a pair of RX01 diskette drives.  I wanted to see if I could boot RT-11.  This proved to be unsuccessful.  I'm still not sure why, but the 11/20 is not terribly happy with the floppy drives.  This will require considerably more debugging.


The following is how everyting is configured now.