[GreenKeys] TTY-Connect FAQ
gil smith
gil at vauxelectronics.com
Thu Jul 29 19:39:03 EDT 2004
Hi folks:
I was hoping to tell you that the firmware doc and file were ready today,
but alas, I still need to do some more work on them. Sometime next week, I
suppose. Below is a FAQ I started from questions I have received and
answered. I will upload this to the site next week as well, and update it
as new questions come in.
I'm off to the mountains until Tueday, to get away from this 105+ degree
heat for a few days, so don't look for any email replies until then.
thanks for all your kind words, by the way.
gil
-------------------------------------------------------------------------------
TTY-Connect FAQ
7-28-04
gil smith
gil at vauxelectronics.com
480-354-5556
------------------------
Q: Where is the documentation for TTY-Connect?
A: Files/pictures/docs are
at: http://www.vauxelectronics.com/gil/tty-connect/
Q: Do you have parts kits, completed boards, or full systems for sale?
A: Sorry, only the pcb and micro are available from me.
Q: Do you have dimensions for the PCB and mounting holes?
A: I kept all mounting hole and connector-centers on a 0.1" grid, to make
it easy to quickly measure
the board and transfer locations to the chassis. Yeah, a dimensioned
drawing for chassis
holes would be nice. Have not had time for such a thing.
Q: Why did you spec the Sescom chassis (it seems too small)?
A: The Sescom rack-mount box is nice for a couple of reasons:
a) it has flat panels that slip into extrusions, so if you want to
change
a hole layout later, you can just change the panel.
b) the panels are 1/16" aluminum (not steel), so it is easy to
drill and nibble holes.
c) since the panels are flat, you can use a punch press (if
available).
d) they are available in a variety of sizes -- I happen to be
using the
2U-high-by-7-inch-deep version -- if you want more room
inside, get a bigger one,
like the 10-inch-deep version. But you can mount the unit in
anything you'd like.
Note that the Sescom box's lower-rear extrusion needed to be notched
where the two pcb mounting
standoffs are located. This is easy to do since the extrusions are
also aluminum.
Q: Why is the assembly drawing cut off around edges?
A: That is all of the assembly drawing that is available in pdf form. It
is really
only for the pcb components anyway, and the "cut-off" stuff merely
shows tentative
locations for transformers etc.
Q: The parts list mentions an optional $83 LCD -- what is it for?
A: Well, the Matrix-Orbital lcd is not supported in firmware at this time,
but the pcb has a
header that provides the signals (it uses I2C, not a simple
logic-level uart, like other
serial lcds). The objective is to provide a front panel control for
the unit, and maybe
a monitor mode. But I am not committing to anything at this point (I
should not have
put the lcd in the docs). It is a nice lcd module, but don't buy one
yet.
And a front-end on a PC will do the same thing.
Q: I am poking through my junk box for parts, and I have various
optoisolators. Can I use them?
A: The places where the 4N37 isolators are shown could use other parts,
but I'd rather not get
into that. The 4N37s are cheap and easy to find. DO NOT substitute
the high-voltage H11D2.
Q: I'd like to use sockets where possible. Do some parts need to be
soldered, or can I use sockets?
A: Not necessary, but I would socket all chips, including the 6-pin dip
optos. Easier test/repair.
Q: What's the story on the two ways to wire up loops HV1 and HV2, using
either 80VDC or 160VDC?
A: The high-voltage loops may be configured for either 80VDC or
160VDC. To use 80VDC, the
transformer primaries are wired in series, and to use 160VDC, the
transformer primaries
are wired in parallel. The power resistor values change too, of course.
I really recommend the 80VDC option, since things run much
cooler. With the heatsinks listed
in the parts list, the unit set to full duplex (which burns additional
power), and the loop
set for 60-mA (for max power), temperatures read:
config xfrm heatsink
--------------------------------------
80VDC 100 degrees 125 degrees
160VDC 110 degrees 175 degrees
175 degrees (F) is just too darned hot, in my opinion. If you really
must use a 160V loop,
it should have a larger heatsink and/or fan. If you use 80V, the
heatsink listed in the parts
list should be fine with vertical mounting and simple convection cooling.
Q: I am curious as to why the transformers run so hot especially in the
160-volt mode? You show
the transformers to be 43 VA and I calculate that the current at 115
volts should be 370 ma.
Only drawing 60 ma should mean that the transformer is loafing.
A: It is not the transformers that run hot, it is the
power-resistor/heatsink that gets toasty.
The xfrms are oversized at 43VA, but a smaller xfrm is about the same
price, and I like to be
conservative, especially on a power design. Which is why I don't like
a heatsink running
at 175 F (160V loop). With an 80V loop, the heatsink is only 125 F,
and the lower voltage
is safer as well. Also, to get a worst-case scenario, these
measurements were made in Arizona,
during summer, in a garage where the ambient air was over 100 degrees.
Q: Depending on the tty-connect chassis size, and what is stacked on top
or bottom of it
in the rack, will heat be a problem?
A: If you are using 160V loops, a fan (and bigger heatsinks) would be a
good idea.
With 80V loops it should not be necessary. Good point about what is
above/below the
unit in the rack. You would want to mount tty-connect at the top of
the equipment so
it does not heat anything else up, and you would want to leave a rack
space or two
above/below it for convection air.
Q: I was checking parts list against schematic and found two C10
capacitors. Also U11 and U12
are not on the parts list. Would you clarify?
A: Oops. Yes the schematic should show C11 (not C10) near U11; it is a
bypass cap (0.1 uF)
for U11/12. The assembly drawing and silk are correct.
Actually, the schematic should show U11, U12, and C11 as optional,
which is why they are
not on the parts list. U11 and U12 are external eeproms (non-volatile
memory) -- there
are many sizes of eeprom that fit into these sockets. But there is a
small eeprom built
into the pic, which stores the tty-connect machine configuration. At
this time, U11 and U12
are not used.
Q: I thought the pic micro was already programmed, but apparently not --
what's up with that?
A: I programmed the micros with the bootloader, but did not load the
firmware. You will need
to load the firmware from a PC after your unit is built (it just takes
a few minutes).
You need a serial cable (DB9-F to DB9-M, straight-wiring) from the PC
to TTY-Connect.
Get the application file on the website called
PIC-downloader.exe. This little standalone app
runs on winblows xp/98, and allows you to send the actual tty-connect
object code to the pic.
To use, PIC-downloader.exe, you need to also get the tty-connect.hex
file for download to the pic.
If (or when), I make a change to the code, I will upload the latest
version to the website,
and folks can use PIC-downloader to update their tty-connect system
with the latest version.
All of this magic made possible by flash memory in these 16Fxxx pics
(as opposed to the
one-time-programming of the 16Cxxx parts). Details are in an appendix
of the firmware doc.
Q: I did not get the micro from you as I already have one -- what do I
need to do to program it?
A: For you folks that want to burn your own PICs, and have access to a
programmer to do so, you
need to obtain a PIC16F876-20/SP chip (mouser part 579-PIC16F87620SP
is $5.44 ea). On the
website, is a file called bootloader.hex. This small file gets burned
into the pic -- it
resides in the upper hundred or so bytes, and provides bootload
capability over the serial port.
This is the only time the pic needs to be in the programmer. Since
most folks don't have a
programmer, I am doing this step for them when they get the pic from
me. Both the bootloader
code and the downloader are simple freeware goodies -- I have used
them for years, and
they perform flawlessly. After burning the bootloader into the pic,
put it into the tty-connect
board, and refer to the above question for loading the actual
tty-connect firmware.
Q: Are you providing source code for the pic?
A: Sorry, no.
Q: Is there a PC GUI for using the system?
A: No. TTY-Connect uses a simple command/message 232 interface for
control and programming.
It is easy enough to use a terminal emulator like HyperTerminal to
enter commands with a
few keystrokes. It is also feasible to embed these commands into a PC
program,
and provide a GUI front-end. Be my guest, if you are so inclined.
Q: I have seen more than a few tty machines that have a ground path on the
keyboard side. It's not
common to see it on the magnet side but it can happen. It's probably a
real good idea to have
people ohm out their machines before hooking tty-connect to them. A
ground short on either
side will smoke the board or at least blow a fuse, won't it?
A: All loop circuitry is floating from chassis/earth ground, and is
opto-isolated from the
232/TU section. If you have just a single half-duplex loop to a tty,
and one side of the
loop (either side actually) is grounded to earth, it will NOT affect
operation, will not
blow a fuse, nor will it damage the board.
But, if you are running two loops in full-dup mode, I suppose certain
ground combinations
could end up causing problems. However, I DID presume that your tty
loops are floating.
Checking machine wiring to know what you have is a good
idea. Bringing out a pair of
wires for the selmags, and a pair for the keyboard contacts is the
simplest.
;-------------------------------------------------------------------------
; vaux electronics, inc. 480-354-5556
; www.vauxelectronics.com (fax: 480-354-5558)
;-------------------------------------------------------------------------
More information about the GreenKeys
mailing list