Advice wanted on Kennedy 9600 and PDP11/23

From: CLASSICCMP_at_trailing-edge.com <(CLASSICCMP_at_trailing-edge.com)>
Date: Tue Jul 6 10:07:44 1999

>>What is the third party card?

>It's an Emulex. The card has silkscreened Assy TU0210401 REV C. One one of
>the chips, it says Top Assy TC0210201-FSH, Sub Assy C6716 C.

OK, I use many TC02's here, and have never had a problem at 800 BPI.
(The only density problems I've had have been on Dilogs, having to do with
expecting PE bursts at the beginning of the tape.)

>>Well, the tape does need to be re-inited at the new density. Are you
>>doing the INIT before the BACKUP, or doing a BACKUP/INIT? Does the INIT
>>operation work OK? Can you do a DIR MS: after the INIT at 800 BPI?

>In my version of RT11 (5.04), /init isn't a valid option on the backup
>command.

Historically, the BUP.SAV options have been out of sync with the CCL BACKUP
(command line) options for many versions of RT-11. This was always
a damn shame, as it discouraged folks from using BACKUP. Things get fixed
in 5.5 and later.

> But, when I do a backup/dev du1: ms0: it does ask if I want to
>initialize, that seems to complete, then the backup goes on and finally an
>output error. Once I get the tape controller reinstalled, I'll tell you for
>sure.

What happens if you just INIT MS0: and COPY a couple of files to it?
Can you read them back off at 800 BPI?

BACKUP uses 4096-byte blocks, while the COPY (PIP.SAV) operations use
512-byte blocks. It's conceivable that you'll only have troubles with
the long blocks.

>>I'll also point out that NRZI requires a lot tighter physical tolerances
>>on the alignment of the tape head (the reason why many drives don't support
>>800 BPI NRZI at all) than 1600 and 6250 BPI (which allow substantial skew
>>between the tape channels as part ofthe spec.) If at some point the head
>>in your transport had been replaced or knocked around without properly
>>being re-aligned you might see something like what you're seeing.

>Really! I didn't know that.

Yep. NRZI is a pretty low-tech interface, and it relies entirely on
good alignment of the heads for timing. PE and GCR allow skews of up
to several bits that are disentangled by the electronics in the formatter.

> When I got the tape drive, it had been in
>storage and was immaculately clean. Other than usual wear on the door, it
>looked in fantastic condition. What's involved in aligning the heads? I take
>it I would need an alignment tape? 1600bpi seems to work great though, both
>reading and writing.

It takes a master alignment tape, a scope, and a resistor network that sums
the outputs from the tape channels. The master alignment tape has transitions
that are supposedly well-aligned between all channels, and you tweak the
head until you get a nice pattern on the scope. I can fax you a few pages
from drive maintenance manuals on the procedure if you're really interested.

Many drives do deskewing of NRZI in electronics (usually with shift registers
running at several time the data rate) in addition to having the head being
aligned correctly overall.

-- 
 Tim Shoppa                        Email: shoppa_at_trailing-edge.com
 Trailing Edge Technology          WWW:   http://www.trailing-edge.com/
 7328 Bradley Blvd		   Voice: 301-767-5917
 Bethesda, MD, USA 20817           Fax:   301-767-5927
Received on Tue Jul 06 1999 - 10:07:44 BST

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:32:11 BST