Continuing PDP11 saga

From: Pete Turnbull <pete_at_dunnington.u-net.com>
Date: Wed Jan 9 13:30:54 2002

On Jan 9, 9:12, Tom Leffingwell wrote:
>
> I tried taking an M8027 and setting its address to 170400, to make it
look
> like the ADV11-C. However, when typing that address before and after I
> installed the board, at the _at_ prompt, it returned 177777.

Whatever the address you set, it wasn't 170400...

> Also, the
> DRV11, at 167770, does the same thing. The DRV11-B gives 000000 (at
> 772410), although at the CSR (772414) it gives 006800.

Well, if you're using 18-bit addressing -- which you must be if the DRV11-B
responds to 772410 rather than 17772410 -- then all the I/O addresses begin
77.... rather than 17.... 17.... in ODT would address the memory. I/O
addresses are often given as 17.... for older QBus devices, because the
original LSI/11 used 16-bit addressing. You have to mentally add two or
six more '1's on the front for an 11/23 or later processor.

> My printer card that I tried to use as a fake ADV11 has jumpers on pin 12
> and pin 8, with the rest removed. Is this correct for setting it to
> 170400?

No, 'fraid not. This is an LPV11, M8027? The jumpers that go between the
two pins in each of the pairs labelled A3 to A12 set the address, a jumper
inserted makes it a '0' and "no jumper" sets a '1' in that bit position.
 You can't control address bits 13 or higher, nor can you control A0...A2.
 A0 doesn't matter, A1 distinguishs the CSR address from the buffer address
(one word higher), and A2 is fixed at binary '1' -- so this card can never
be set to anything that doesn't end in octal '4'.

If all the jumpers are out except A8 and A12, that would set binary
1110111011111100 as the base address, which is octal 167374. Not quite
what you wanted :-) The closest you can get is binary 1111000100000100
(the leftmost bits are already set, A12 should be out, A11...A9 in, A8 out,
A7..A3 in, A2 is set to a '1' and can't be changed, and the other two are
"don't care") or octal 170404.

> I noticed on the DRV11 card that it says jumper in = logic 0. That one
> has one jumper on pin 12, which I assume gives it the proper address of
> 167770. This one appears unmodified from the factory.

That's right. The jumper sets address bit 12 to 0, so you get binary
1110111111111000.

> Are my jumpers on the printer card the opposite of what they should
> be? Is 177777 what is returned for a missing board? If I have the
> jumpers reversed on the printer card, that would explain that situation,
> but it doesn't explain the DRV11, unless its bad.

Yes, you have the jumper sense wrong on the printer card. What you get
back from a non-existent address depends on the bus termination, but
all-ones is plausible. Maybe you have a bad DRV11, or maybe you've
overlooked a jumper.
At least some of the CSR bits should read as zeros.

-- 
Pete						Peter Turnbull
						Network Manager
						University of York
Received on Wed Jan 09 2002 - 13:30:54 GMT

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:34:53 BST