What started out as some experiments on how to use ICOM's expansion
modules (UT-115, UT-118,
UT-121 and UT-123 now) to create a
home-brew D-STAR solution, eventually has become a widely known and often-used project:
the DV Node Adapter board, designed by Satoshi Yasuda 7M3TJZ.
The first versions were based around a PIC16F877 chip from
Microchip, and basically talked the low-level command language required by the
UT-xxx modules. It is not known whether any of these are still available,
but fortunately, some pictures could be found on
Satoshi's website that show what they
The prototypes were actually veroboard-type projects, where
the components are placed on the through-hole PCB, and then a lot of manual wiring
must be done on the back side. This looks easy, but it isn't, especially not if
you know that the UT-xxx modules have rather annoying
connectors, with a very fine pitch. This is fine for regular PCB's, but not
However, things worked, and the more-or-less final versions can be seen
Later on, a real PCB was made as well, although it is unknown whether this
was actually distributed or not:
Still not a very clean system, but hey, we're experimenting, and
it sure is quite an improvement over having to do the veroboard thing!
The next version was based around a (much bigger and
faster) PIC18F2550 processor, which also implemented a USB device-side
bus so it could be connected directly to a PC (or Mac.) The new design
also no longer required the use of one of the ICOM UT-xxx modules on the
board, it now had its own chips on board.
Although the exact
reasons for this change are not known, it is assumed by most that
ICOM, not amused by Satoshi reverse-engineering their modules so
they could be re-used in the hobby projects, kindly (or not..) let
him know he'd better stop doing that. We'll probably never
know for sure.
Because of the totally new design, with a new processor, and
the fact that the hardware on the UT-xxx modules effectively had to be
re-produced on the new board, it became a more or less standard setup, given
the fact that the entire system was based on the CMX589A GMSK modem chip
from CML. Their datasheet is quite clear in how it should be used in a circuit,
and the first versions of the new Node Adapter design were true followers of
those reference implementations.
As time went by small changes occurred, mostly in component
values around the analog frontend which connects the modem chip to the target
radio. ICOM did not have to worry about things working on many possible target
radios, so although one could 'sneak a peek' in their service manuals and
schematics, it does not really help in the broader light of things.
And so, here it is... the new DV Node Adapter board,
equipped with both CPU's (the master CPU, and the optional one when used as a
repeater for D-STAR systems.) The main components can be clearly identified-
modem chip in the upper left corner, the two PIC processors on the right-hand
side (including their USB connectors), and the analog front-end in the lower
Quite an upgrade from the veroboard-based dv877 versions, and even
from the PCB version of the dv877. The PCB had some minor
design mistakes, such as using A-type USB connectors for a device
(that is normally supposed to be B (or even mini-B) type, a resistor was added, and so
on, but, in general, this is what many people are using right now- usually
combined with the PC Windows-based
HOTSPOT software written by Mark McGregor.
The bad news, however, is that this board has been sold out
completely, and it is not clear whether the future will bring new boards to the
people. Several people have (each for their own reasons) made their own boards,
and some of them were kind enough to make the PCB production files
and/or pictures available.
The simplest way to handle that, is by looking closely at an
original board, and then manually create a copy, using film-transfer methods, or
even by using transparent transfer methods. In either case, you'd end up with a
prepared, dual-sided PCB which you can subsequently etch, drill, and assemble.
It will not look very nice (about the same as the "PCB prototype" version of the
dv877 project mentioned earlier on this page), but it WILL work, of course.
A helpful HAM sent us scanned images of his un-soldered
board, so people can see (clearly) where all the components must go.
||Obviously, these cannot be used for making your
own boards, but they do help in case you'd like to try something on
veroboard, or even feel like digging out their etching materials!
here for a large version!
And if all else fails....
One HAM sent us pictures of his own solution, writing "I was
unable to get a real board, so I ordered the parts, and then started to make my
own PCB, based on how Satoshi's board looked in the pictures. It works great,
thanks to all who have helped !!"
OK, so let's have a look at those pictures, shall we?
Wow, that does look quite familiar, eh?
We see pretty much the same layout as the one used on the
original board (then again, he did tell us that, we should have paid
attention..) and at roughly the same positions, too.
Funny is, that even this "design" uses the better-suited
B-type USB sockets.
Although we don't know for sure, we assume that the DSUB-9
socket at the front is for programming, or maybe it can be used to extract the
'slow data' bytes out of the system. We'll ask and report back here.
The second picture gives a better look at the board, and now
we clearly see the DSUB-9 socket as well. In the background, we can see part of
the original schematics, as published by Satoshi Yasuda.
All this looks really neat and tidy, so hey... what's the
Ahh, so that's the secret.. all the scary stuff happens on
the back side of the board. A true spider-web, but: somehow, it does look
pretty, doesn't it?
Well, that's it for now, folks More information can be found on Satoshi's
We cannot check because Satoshi has
decided to block our access to his site at the IP level, for unknown reasons. Still, if you're
interested in the project, go and check it out !