Help collecting gear from west of Chicago

Tony Duell ard at p850ug1.demon.co.uk
Mon Jun 6 17:51:06 CDT 2005


> 
>  > > Teradyne-built items; I don't recognize the bus -- two edge connectors,
>  > > each wider than qbus.  It doesn't look like the unibus cards I've seen.
> 
>  > Unibus and Qbus use the same double-sided 36 pin (18 pin per side)
>  > 0.125" pitch connectors. Q bus boards are normally dual or quad,
>  > Unibus are normally quad or hrx.
> 
> These in the Teradyne chassis have 60-pin edge connectors.  The full
> width boards are about the same size as quad-width UNIBUS boards.
> The end of the chassis with the memory cards has two such connectors.
> The other end has only one connector, and an assortment of I/O cards --
> variously full of UARTs, driver transistors, ADC, etc and connected
> to about 5 substantial cables.  Haven't followed the cabling to see
> what goes into the main CPU yet.  Both front and back of this cab
> are chock full of stuff so it's hard to see inside.

The 11/44 has a 22 bit address bus, which means you don't normally put 
memory on the Unibus (you put it on special slots in the CPU backplane 
that carry the 22 address lines, etc). The unibus is uses only for I/O 
devices, only accesses to the I/O space go to the Unibus. I think it's 
possible to configure the machine so that the top 128Kwords of address 
space is mapped to the Unibus (so you can put Unibus memory on the 
machine for special appliactiosn). I would be suprised if that's what's 
been done, though

> 
> Given that two of the boards in the teradyne chassis are memory,
> I'd expect some kind of CPU, but the other cards on that end of
> the bus don't seem to have appropriate parts.  Perhaps some kind of
> marble machine?  I'll have to post photos.

Remember you can make a CPU from simple TTL chips, and it doesn't take 
that many of them either. I would guess 150 chips would be ample....

> Unfortunately, didn't have my hands on the machine until it arrived
> in my driveway.  I suspect the machine was moved before the auction
> started anyway.  Yesterday I finally got inside the two drives
> which are not labeled bad.  One has no hda at all; the other is not
> locked down.  Sigh.  Is this a guaranteed bad, or just highly likely?

Very highly likely, I'm afraid :-(

> Any way to test or estimate non-destructively?  I imagine working
> RA81 HDAs are unobtainium?

I think uou just have to spin the drive up and see what happens. There 
should be a DB25 socket inside the drive chassis, this is a serial port 
that uou can connect a terminal to to run diagnostics, etc. Somebody must 
have the manual.. I am pretty sure you can check the heads will load and 
the servo will lock with a terminal here (and I don't think you need the 
drive to be linked to a controller). If that seems to work the HDA may be 
OK.

> 
> Still haven't opened the RL02s.  Seems like there's something
> (a powered solenoid) locking the top latch?  Any way to defeat it?

Old RL's have a plate on the RHS about halfway along. Remove it (2 
screws) and you can see the solenoid armature. Pull it down and you can 
operate the door catch. Don't forget to put the plate and gasket back on, 
IIRC that's part of the clean air ducting on this drive (!).

On all RL's uou can get inside by getting the drive as far out of the 
rack as possible (that's to the normal load position, then fiddle with 
catches on the rack rails, then pull it out a bit more). Undo the 4 
(captive) screws on the top of the drive that hold the logic cover in 
place. Lift this up, starting at the back and free it from under the edge 
of the drive door. Hook the logic cover (which contains the main logic 
PCB, and is linked to the rest of the drive by numerous cables) onto the 
back edge of the drive.

Now lift the rear edge of the door carefully, operate the catch to free 
the front edge. Once you've done that, put he rear edge down again (so 
that the hinges fit into their notches) and swing the door up. It's 
loose, and there are a couple of spring-loaded arms that run on rollers 
on the chassis (this will make sense when you do it), so you need to take 
a little care.


> Started through the manuals from Bitsavers this morning, but haven't
> found the golden page yet.
> 
>  > The 11/44 PSU is the second most complicated I've ever worked on
>  > (it was the most complicated until I worked on an HP9845..) It's
>  > also got some lethal features, like 400V DC, straight from the mains,
>  > on barrier strip terminals inside. Please ask here (or me directly)
>  > before diving inside the PSU.
> 
> Thanks for the warning.  For curiosity, would the wimpy US power
> version have 400V, or more like 200?

No, it's still 400V. 230V models full-wave rectify the mains and apply it 
to a pair of coke-can size capacitors at the left side of the chassis. 
115V versions use those 2 caps and 2 of the diodes of the bridge to make 
a voltage doubler. You get 400V applied to the chopper in either case. 
And it's easily enough to kill you.

> 
> Supposedly I have a set of 11/44 prints on the way from .nl which
> may help.

I have the printset (including the PSU) but no easy way to get bits of it 
to you.

-tony



More information about the cctech mailing list