Source for DEC TC01 (and similar) bulbs?
Josh Dersch
derschjo at gmail.com
Mon Dec 6 23:37:38 CST 2021
On Mon, Dec 6, 2021 at 7:39 AM Michael Thompson <
michael.99.thompson at gmail.com> wrote:
>
>
> On Sun, Dec 5, 2021 at 11:46 PM Josh Dersch <derschjo at gmail.com> wrote:
>
>> I swapped boards around in the datapath for these bits on the TC01 side
>> last night and the problem doesn't move. On a whim I put my thumb on the
>> tape as it passes over the head and it seems to affect the glitchiness of
>> the data, at times making it appear to go away entirely. Doesn't appear to
>> be a tape tension problem -- it seems adequate and putting my thumb on the
>> supply reel to increase tension a bit does not have the same effect. This
>> would seem to point to the TU55 being the issue, and quite possibly the
>> heads. I haven't had time today to look at the signals coming off the
>> heads but I plan to soon.
>>
>> Thanks for the help!
>> - Josh
>>
>
> I would measure the resistance of the head coils at the relay board
> connector. Hopefully they all measure about the same. Figure 2-2 in the
> maintenance manual shows how the MARK and DATA tracks are wired, and the
> extra connection available for the TIMING track. We found a head on the
> PDP-9 where the +D0 connection was open.
>
Looks like I lucked out. Testing the head signals at the backplane on the
TU55 revealed that pin BD1 was a flat line at about -0.5V. Turned out to
be a bad relay on the G851 relay board. (There was also some gnarly black
gunk on the pins on the head connector which somehow wasn't the
problem...) I don't have any spare relays, but I swapped a G851 in from my
TU56 and the glitches have gone away. If anyone has any spare relays for
these, or a G851 lying around, let me know.
The Basic Search diagnostic still isn't finding any blocks on the tape, and
I can tell by the lights on the panel that the LPB register isn't doing
much of anything, so I have some more debugging to do. But I'm really glad
that the TU55 is OK. Thanks for the help!
- Josh
>
> --
> Michael Thompson
>
More information about the cctalk
mailing list