30 Apr
2016
30 Apr
'16
10:22 p.m.
Has anyone used any modern hardware to interface with HP-HIL gear? (HIL is Human Interface
Link, not to be confused with HP-IL.) I?d like to interface an HP 46085A Control Dial Box
<http://www.hpmuseum.net/display_item.php?hw=684> with my Mac via USB.
But before I go any further in this?I?ve already put a bunch of effort into trying to get
it to work?I realized I should probably ask here to see if anyone else has done it. Has
anyone?
HP-HIL uses what *should* be a pretty basic serial protocol with available docs. (Thanks,
Bitsavers!) The only odd things about it are (1) a data rate of 100Kbps (rather than say
115Kbps) and (2) the use of 12/O/1 framing (rather than say two 8/E/1 frames per packet
and (3) a 12V supply line for what?s otherwise a TTL-compatible bus.
Getting the actual host interface IC is just about impossible without removing it from
existing equipment, and the client interface IC doesn?t really seem like it would act as a
host. The weird framing means virtually no common UART will work, and the data rate means
something like an Arduino only has 160 cycles/bit to work with.
It looks like I might be looking at either using either an Arduino or a CPLD to implement
a 12/O/1 UART to talk to the control dial box, and using a second Arduino to communicate
with that (via a parallel interface) to actually implement the USB HID via the builtin HID
stack functionality on some models of Arduino.
-- Chris
5 May
5 May
9:47 p.m.
HP-HIL uses what *should* be a pretty basic serial
protocol with available d$
I haven't, but my own tendency would be to build a UART from a clock
generator (divided down) and some shift registers for the hardware side
of it, then add some logic for the host interface - I might use a
115200 serial port or I might use a parallel port, I'm not sure.
I would not mess with things like Arduinos. I would prefer a bunch of
TTL for multiple reasons: (1) because I already have it; (2) because I
understand it better and thus can build and repair it better; (3) it is
likely to be better in a bunch of respects - faster, lower power, maybe
even cheaper, etc.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
6 May
6 May
9:39 a.m.
faster, lower
power, maybe even cheaper, etc.
TTL could never claim "lower power" :-)
11:46 a.m.
On 05/05/2016 09:40 PM, Lyndon Nerenberg wrote:
On 2016-05-05 6:47 PM, Mouse wrote:
Hmmm, I don't know. (Don't remember the earlier part of
discussion.)
Some discrete transistor implementations were actually
pretty high power, they ran off 12 V rails or even more. I
know some old computer I poked into (SEL 12-bit, maybe) must
have been using 24 V rails, they had 2 W load resistors on
the transistor collectors, and the boards were badly burned
around those resistors! Some systems use a variant of ECL,
which of course really burned power, too.
Jon
faster, lower power, maybe
even cheaper, etc.
TTL could never claim "lower power" :-)
7 May
7 May
1:35 p.m.
On 05/06/2016 10:26 AM, Paul Koning wrote:
This application sound like a perfect candidate for a CPLD.
--Chuck
On May 6, 2016, at 1:11 PM, Fred Cisin <cisin
at xenosoft.com> wrote:
I think a well chosen hearing aid tube can outperform a classic (not
L or LS) 7400 series IC. > TTL could never claim "lower power"
:-)
than vacuum tubes ("valves")?
3:05 p.m.
On 05/07/2016 12:35 PM, Chuck Guzis wrote:
On 05/06/2016 10:26 AM, Paul Koning wrote:
The Xilinx 9500 series of CPLDs are not low-power, by any
means. They have a formula in the data sheet and a power
calculator program, that are both WRONG! They underestimate
power by a factor of 3!! I reported this to them, and they
said I was right, but they were not going to fix the
documentation! Sheesh!
(The internal architecture is 36-input diode gates with
pull-up resistors. You have two choices of resistor,
low-power and fast.)
The Xilinx CoolRunner II is quite low power, basically a
CMOS FPGA architecture.
Jon
I think a well chosen hearing aid tube can outperform a classic (not
L or LS) 7400 series IC.
This application sound like a perfect candidate for a
CPLD.
4:46 p.m.
On 05/07/2016 12:05 PM, Jon Elson wrote:
The Xilinx 9500 series of CPLDs are not low-power, by
any means.
They have a formula in the data sheet and a power calculator program,
that are both WRONG! They underestimate power by a factor of 3!! I
reported this to them, and they said I was right, but they were not
going to fix the documentation! Sheesh!
I don't believe that Xilinx has produced the 9500 series for at least a
couple of years.
I believe that Atmel still makes some 5V CPLDs (ATF1500 seies).
--Chuck
5:03 p.m.
On 05/07/2016 03:46 PM, Chuck Guzis wrote:
On 05/07/2016 12:05 PM, Jon Elson wrote:
Oh, yes, definitely discontinued. There is stock of either
NOS or counterfeit parts available from China.
But, if you REALLY need a 5V-powered CPLD, there are only a
few choices.
The Xilinx 9500 series of CPLDs are not
low-power, by any means.
They have a formula in the data sheet and a power calculator program,
that are both WRONG! They underestimate power by a factor of 3!! I
reported this to them, and they said I was right, but they were not
going to fix the documentation! Sheesh!
I don't believe that Xilinx has
produced the 9500 series for at least a
couple of years. I believe that Atmel still makes some 5V CPLDs
(ATF1500 seies).
I wonder for how long those will remain available. The big
problem is the foundries are shutting down those 5 V processes.
Jon
5 p.m.
On 05/07/2016 02:05 PM, Jon Elson wrote:
On 05/07/2016 12:35 PM, Chuck Guzis wrote:
Oh, but only the 9500 series is a 5V part. The 9500 XL is 5
V tolerant, but needs 3.3 V supply.
Jon
On 05/06/2016 10:26 AM, Paul Koning wrote:
The Xilinx 9500 series of CPLDs are not low-power, by any
means. They have a formula in the data sheet and a power
calculator program, that are both WRONG! They
underestimate power by a factor of 3!! I reported this to
them, and they said I was right, but they were not going
to fix the documentation! Sheesh!
(The internal architecture is 36-input diode gates with
pull-up resistors. You have two choices of resistor,
low-power and fast.)
The Xilinx CoolRunner II is quite low power, basically a
CMOS FPGA architecture.
Jon
I think a well chosen hearing aid tube can outperform a
classic (not
L or LS) 7400 series IC.
This application sound like a perfect candidate for a
CPLD.
3031
days inactive
3038
days old
12 comments
7 participants
participants (7)
-
cclist＠sydex.com -
cisin＠xenosoft.com -
cmhanson＠eschatologist.net -
elson＠pico-systems.com -
lyndon＠orthanc.ca -
mouse＠Rodents-Montreal.ORG -
paulkoning＠comcast.net