be aware that Jacques may or may not be able to provide support for
this project and your understanding on this matter is appreciated.
Feel free to ask me (via the link at the bottom of this page)
if you have any questions.
The "Alternate Firmware" for the Doppler II is still available
- go here
for more information on pre-programmed chips
Jacques Brodeur, VE2EMM, has designed a microprocessor-based
type DF unit described atVE2EMM's
pagethat drives the antenna switching
updates the display, and does signal processing to determine the
This unit works along the same principles that the so-called "Roanoake"
DF units do - but the addition of a microprocessor (to do the
of switching signals and processing of the received audio) adds
flexibility. For a bit of information about the newer
III Doppler unit, go to the "Doppler
3" page.This page is not maintained by
Jacques - see the disclaimer below.
This page describes some of the technical details of the VE2EMM Doppler
DF unit, as well as a few more topics such as:
While this unit is not available as a complete kit, a complete parts
list is available on the Doppler
page and the key components are available from Far
Circuits as a "partial kit" that includes three ICs (the
microprocessor, a low-voltage rail-to-rail op amp, and the RS-232
IC) as well as a 20 MHz crystal, the circuit board for the DF
and a circuit board for the PIN diode antenna array switch.
This is not a "kit" for a novice builder: If you
plan to undertake construction of this project you should either have
soldering skills as well as plenty of experience with this sort of
- or a lot of help from those who do!
Construction isn't particularly difficult for the experienced
The use of the schematic, parts location map, and parts list makes it
easy to populate the board, but there are a few things that should be
in mind when component selection is made:
Pay close attention to the types of capacitors used for C5, C6,
C8. These capacitors are the main frequency determining
in IC2, a 500 Hz bandpass filter. Do not even think
about using any ceramic or electrolytic capacitors for
and C8 (1 uf) as they simply are not suitable in terms
temperature stability. For capacitors C5 and C7 one may use NPO
if you have them, but polystyrene or polyester are preferred for these
units, as well as for C6 and C8. The key here is temperature
stability! If you skimp here, the unit will drift terms
bearing accuracy and usability over temperature. I had on-hand
0.47 uf high-stability polyester capacitors and used them in pairs,
in 0.94 uf of total capacitance for C6 and C8 while I used 470 pf
units for C5 and C7.
Pay close attention to the "Construction Hints"
text on the Doppler II page - particularly the
labeled "Adjusting the Peaking Filters." You really must
do this step to get optimum results. In my case, I ended up
470k resistors across R4 and R5 to "dial in" the frequency precisely -
but the precise values needed will vary with each construction.
Putting it in a box
A front panel view of an
Doppler-II unit Click on the image for a larger version
Pictures on VE2EMM's page show the DF unit housed in a food tin, but
that should be considered only to be a "serving suggestion."
sort of enclosure will do. The only (minor)
about a nonmetallic enclosure is that having to do with RFI - either RF
getting into the unit (not likely to be too much of a problem,
a local transmitter would probably overload the DF receiver anyway,
readings) or RFI from the unit, causing interference to
signals. Actually, I have not noticed a tendency for problems of
either sort. Anyway, if there is a problem with RFI it
more likely be a result of RF being conducted along a wire or cable
to the unit rather than from direct radiation from the circuit board
My "copy" of the 'EMM box is built as pictured, put in a plastic
While it would have been ideal to place the buttons below the
(to more-closely match the items on the various menus) I miscalculated
exactly how large the LCD display was (I got the largest 16 character
2-line display that I could find) and had to settle for the arrangement
pictured. The upper switch on the right-hand side is the power
while the lower switch selects Bright-Dim-Off for the LCD's LED
on the backlight later... Even though the unit is in a
box, I have observed no RFI problems.
A rear panel view of the same
unit Click on the image for a larger version
The other picture shows the rear panel, along with the various
cables, and connectors. These are: Speaker volume control
left), input level adjust (upper left) and headphone jack (next to the
input level control.) Next to this is the "damping control" for
added switched-capacitor filter (more on this later.) The
gray cable is the audio input cable, the red/black cable is for power,
and the connector on the left is a high-density 15 pin "D" connector
the antenna controller (wired to allow the use of up to 8 antennas) and
the right-hand connector is a 9 pin "D" connector for the serial data.
Having had quite a bit of construction experience, I had no problem
assembling the parts and constructing the unit. Some fellow
are constructing their own copies of this unit, so I'll ask them about
experiences when they are done...
The only problem that I experienced appeared to be related to
of the processor: When set for continuous serial data
of the bearing, it no longer provided any data that actually resembled
the bearing: Semi-random bearings were spewed out on the LCD and
on the serial port. Jacques was helpful in my getting a
part and the problem was resolved. (This appears to have been
a singular incident rather that an actual design problem, so don't
Since building the unit, Jacques has been kind enough to send me a
of the source code. Originally written to use the Hi-Tech
I have since modified it to use the CCS PICC compiler and added some
features and fixed some minor bugs.
More recently, very significant rewrites and modifications have been
done to provide more features and operating flexibility.
on this new firmware, go to theUpdated
Montreal Doppler II - page.Firmware
with the same added functionality/enhancement is also available for the
newer Montreal Doppler III hardware.
The antenna array
Discussion of antenna arrays usable with this system may be
on the Doppler Antenna page at this web
Operating the DF unit:
Note: This descrption refers to VE2EMM's original
For information about the newer firmware, go here.
If you have used "doppler "type DF units before (e.g. the "Roanoake"
types) that directly indicate bearing, operation will be a generally
experience: The unit will need to be calibrated, in terms of
In a vehicle, this is typically done by setting "0" degrees (due North)
to be straight ahead of the vehicle while a unit mounted at a fixed
is typically configured to display bearings relative to True North.
Being microprocessor-controlled, a variety of configurations may be
easily set up. A few of these include:
Calibration of the bearing readout. It is very easy to set
ahead" as 0 degrees - very useful for vehicular work. Manual
(to set any arbitrary offset) may also be done.
The number of antennas may be selected. This unit supports
The switching polarity may be selected. The "active"
be selected using a logic "high" (5 volts) or a logic "low" (0
The required setting will depend on how the antenna switching diodes
The "spin" direction may be selected as either CW or CCW.
number their antennas clockwise when viewed from above, and others when
viewed from below. This selectability facilitates either scenario.
A facility is provided to test each antenna individually.
it easy to verify that the PIN diode switches are working and/or to
identify each coax.
Serial data output is available. The current bearing - and
quality - is output in the (so-called) "Agrelo" format (ala the "DF
[tm].) This format is: "%bbb/q<cr><lf>" where
a 3-digit bearing and "q" is the signal quality. A baud rate of
or 4800 baud may be selected (with the possible selection of 9600 baud
on newer software.) This bearing data may sent continuously (at
to approx. 10 bearings/sec.) for logging or compass rose displaying or
in timed intervals - for APRS packet use.
This unit will also accept serial input from a GPS
This may be done to facilitate the use of both GPS and DF information
a single serial port on a computer or with APRS.
A "minimum quality" display threshold may be selected.
quality lower than this will not update the display.
An "S-Meter" reading may also be displayed. This takes a
that goes from 0 to between 2 and 5 volts for a numerical "S-meter"
Configuration for settings for up to 3 radios/installations may
in nonvolatile memory. These settings include bearing
number of antennas, their polarity and spin direction and make it
easy to move the unit between, say, two vehicles and their radios and a
Readings may be "averaged" to improve the stability of the
Internally, one reading is taken for every 24 "revolutions" of the
array, and since this unit "rotates" the antenna 500 times per second,
20.83 (500/24) bearings are available each second. The display
be updated this quickly (i.e. no averaging) or 2, 4, 8, 16, 32 or 64
may be taken between updates. This averaging is very useful for
"jitter" from the derived bearing - a condition often encountered due
modulation of the carrier itself and/or signal degradation due to
mobile flutter, and/or multipath.
As good as it is, there are a few minor drawbacks that must be worked
Unlike the traditional "Roanoake" DF box, this unit uses a 16
by 2 line LCD display for showing the DF information as well as
data. This includes a sort of simulation of a "36 segment"
rose showing the direction of the bearing. While a clever
this display is not as useful as one might hope - particularly if one
attempting to use this unit in a vehicle: It is very small and
be difficult for some people to see. It is also not at all
so it is more difficult to translate the displayed bearing to the
in which to drive or for an observer to look. While the bearing
shown numerically, this is also not as intuitive as the "traditional"
rose in its usage. To this end, an external compass rose may be
via the serial port - more on this later. Jacques has
this issue in his more recent version of the Doppler DF unit, the
III." More on info on this may be found at his web site.
While the LCD-based faux compass rose shows the last valid
3-digit bearing display does not (it shows "BAD" - a status also
by a "0" for the signal quality indicator.) If this signal has
away before you happened to glance at the reading, you will have to
based on the faux compass rose display. Having a serial port on
unit, however, make is perfectly practical to log the readings.
While not really a complaint, it would have been nice if the
of the reading were a "sliding" average rather than simply taking a
of N readings ("N" being the number of averages) and
spitting out the result (and resulting in a much slower display update
rate.) Of course, to have a "sliding average" requires having
memory to store N readings - as well as a bit of extra
horsepower - so it is perfectly understandable that this was not done.
A few operational details
In real life, one would rarely operate using "1" average (that
averaging - 10 readings/sec.) First of all, this update rate is a
bit too fast to be truly useful to a human operator and bearings may be
severely affected by modulation or multipath on the received
Experience shows that an averaging of 4, or 8 readings (correlating to
2.5 or 1.3 readings per second respectively) "stabilizes" readings
well while still maintaining a comfortably fast update rate.
Volume setting appears to be critical. One should set the
so that the display shows a "signal level" of about 90 on "average"
An occasional "OL" (overload) indication may result - but this doesn't
appear to affect accuracy. What can affect accuracy is a
in audio level - and this seems to cause a slight shift in bearing. (This
that Jacques is working on addressing in the Montreal
III doppler.) The workaround is to set the volume at a given
level and not change it.
The schematic suggests a rheostat to set the speaker level.
this is typically set, when turned all of the way up the volume of the
speaker mounted in the DF unit's enclosure will be at the loudest
level, with reduction done with the rheostat. Chances are, this
overdrive the DF unit, so the use of another potentiometer is suggested
in order to independently set the drive level to the input of
A 1k potentiometer is suggested for this, as this would still be low
to the input impedance of the first section of IC2. This is one
the controls visible in the "rear panel" picture.
The DF unit is designed to work with a standard 16 character by 2
LCD and an obvious nicety is to select an LCD with a backlight.
are several options available for backlights:
None at all. This is fine if you never plan to use the unit
dark or subdued light - unless you have a flashlight, that is...
Electroluminscent. This type of lighting is very efficient,
requires an inverter to generate the required voltage (approx. 100
at 100-400 Hz or so.) Unless one already has a surplus unit
the inverter is often as expensive as the display itself. This
of lighting also has a relatively short lifetime - only a few thousand
hours until the brightness dims to 50%.
LEDs. Integrated within the background of the display, an
LEDs provides backlighting. Typically, a parallel array of two
LEDs in series is used, requiring a voltage of 4.2 volts (nominal) at a
maximum current of between 100 and 200 mA. A color of green or
is highly recommended even though these colors are not as "snazzy" as a
red or blue because the human eye is not "designed" to resolve
details at either end of the color spectrum, and the looking at a
illuminated with red or blue can cause excessive eye fatigue.
In the case of the unit pictured, a green LED backlight was chosen for
the display. This has the obvious advantage that no inverter is
(as would be required for Electroluminscent) and the further advantage
that the display brightness may be easily adjusted - simply by
One suggested scheme is shown in the schematic. In this
the "dim" setting is implemented by putting the LED array in series
switch in the center position) with the input of the 7805 regulator on
the VE2EMM DF board. With the typical current consumption of the
DF unit being on the order of 50-60 mA or so, the LED is operating
near its 200 mA maximum continuous current rating. Additionally,
the current to operate the backlight in this mode is "free" as the
operated the LED would otherwise be dissipated as heat by the voltage
The only caveat to method of powering the backlight this that the
supply voltage must be at least 4.2 volts (or so) higher than
minimum voltage at which the 7805 will regulate (about 7 volts) so one
needs more than 11.2 volts to maintain voltage regulation when
the backlight this way - an effect of the 4.2 volt drop across the LED
array. Another effect of this operation is that the brightness of
LED backlight may fluctuate very slightly according to the load of the
unit itself (this effect being mitigated somewhat by the 220 uf
This load can vary slightly, depending on the operation of the RS-232
and how the drive to the antenna unit is provided.
There is also the "bright" mode. This is accomplished by
a 100 ohm 1 watt resistor to ground, increasing the current through the
LED backlight. The "off" mode is implemented simply by shorting
the LED backlight. (If you feel nervous about shorting out the
with the switch, you may put a low-value from 1 to 4.7 ohms -
in series with its center terminal.)
Antenna drive circuit:
If you look at the schematic for the DF unit and read the text at
VE2EMM web site, you'll see references to using the output of the PIC
drive the PIN diodes. This is perfectly practical, as the PIC is
capable of sinking or sourcing 20 mA or more - plenty to drive a
Although this "Direct drive" works, it makes me a bit nervous:
With direct drive, one connects the PIC's output pins directly to
world. In terms of RFI, this is just asking for trouble, as the
(or any processor) can easily generate RF noise clear into the VHF/UHF
spectrum. This is why ferrite beads are recommended with this
There is little protection offered to the PIC: With the
method, one relies on the PIC's own protection circuitry to protect
static discharges, etc. While this circuitry is effective against
the casual minor "zap" it will do nothing to protect against potential
damage from nearby lightning strikes or the accidental shorting of an
drive output to, say, the +12 volt supply. If you do
make the latter mistake, you have just bought yourself another PIC...
For these reasons (and a few others that I could come up with) I chose
to put series resistors on the PIC's output and use an external antenna
drive circuit. Fortunately, the circuit board has been designed
accommodate these resistors and values from 1k to 10k should work well
- depending on how your external circuit is driven. Having these
resistors will allow the PIC to tolerate momentary faults to (even to
on these connections as well as providing very effective RFI
(the typical ferrite bead would only have 100-200 ohms of "resistance"
at VHF, while the 1k resistor will have, well, 1k of resistance...)
There is one disadvantage of doing it this way,
The amount of current available from the PIC's 5 volt output through
1k resistor is not likely to drive the switching diodes
and an external driver circuit will likely be required - but that isn't
really too much of a problem.
Note: Neither the author or UARC officially endorse
any vendors mentioned above. The level and satisfaction of
of any of the above circuits is largely based on the skill and
of the operator. Your mileage may vary.
Note: This page (and other pages
on this site) are not
"official" pages of VE2EMM and he cannot reasonably be expected to
answer questions about everything that is covered here. These
are simply set up to aid
those who have built or might build the described equipment.
Do you have any questions on this or other DF-related
topics? Go here.