Philipp Hachtmann wrote:
Ah, really? Of course NOT! The original driver
was DEC8881 which is a
relabelled 7439.
I think it must have been more than just relablled; the original 8881 was from a
non-7400 Signetics family whose name escapes me at the moment, and may have been
binned.
Might be. As I remember the Signetics TTLs had different pinouts compared
to the
TI 74xx series. And now compare the 7438 and the 7439's pinouts: The 7439
doesn't only have a better drive performance (60mA) than the 7438 - its gates
are also in the "Signetics order". I have seen 7439 on a DEC board. Soldered
where other boards have 8881.
The characteristics were somewhat similar to the 7439
but not identical.
The edge rate and maximum leakage current are different. They may have binned
7439 parts that met the Omnibus specs, but there's no expectation that an
ordinary 7439 (or 7438) would.
The 7438 I used has very similar specifications.
As the DEC8881 and the 7438 are not available, I have chosen the 7438.
And it does not make a big difference if the machine is more or less loaded.
The load on the drivers is always the same. It's an open collector bus.
The load isn't the same. Unlike modern CMOS parts, the receiver inputs are not
especially high-impedance. The receivers do present both DC and AC loading on
the bus lines. More receivers presents more loading. The AC loading is often
more of an issue than the DC.
That discussion is a completely theoretical thing. Let's fix some facts:
- I have built working hardware
- I use the "wrong" drivers.
- The real drivers are not available anymore.
===>> I find it's a good compromise to use the 7438 as a replacement.
I could easily change the design to use real 8881 drivers. For me it would be
fine - I still have some of them. But I want to supply the board to other people
who might not have those chips. I want to use current hardware for that current
project.
The spare 8881 should be kept for repairs on old hardware.
And by the way I'm always open to suggestions for improvements. But only
complaining is not exactly that productive.
do NOT work
reliably, because in a heavily loaded bus asserted
signals don't always dip as
low as the threshold of normal TTL parts, or don't
do so in a timely manner.
On a lightly loaded system you can get away with using ordinary TTL receivers
and OC buffers.
I'll try it out with my lab8/e: Full lenght backplane, completely packed with
modules and a cable to a pdp8/a expansion box which contains some more core, a
third party multi serial card, my selfmade joystick interface and others.
If that works reliably, further discussion is worthless.
On the other hand: Most pdp8/e systems are not fully loaded...
The last consequence would be: Two adjustable reference voltages, two
comparators for each bus line. That results in adjustable schmitt-trigger
inputs. But - weren't you complaining about the drivers?
The 7438 are used as drivers in my design. They pull the lines low. That's all
they do. As receivers I use 74AC05 open collector inverters.
Philipp
--
Dipl.-Inf. (FH) Philipp Hachtmann
Buchdruck, Bleisatz, Spezialit?ten
Alemannstr. 21, D-30165 Hannover
Tel. 0511/3522222, Mobil 0171/2632239
Fax. 0511/3500439
hachti at hachti.de
www.tiegeldruck.de
UStdID DE 202668329