TR7 2nd
IF/Audio board variations and S-Meter Calibration
This page details my investigations of variations of the TR7 2nd
IF/Audio board, especially the S-Meter circuit. I will not cover the
product detector section as this has been adequately covered elsewhere
on the internet. This adventure began with my attempts to calibrate
the S-Meter circuit, and led me to some known variations in this area
of the board. I also found other variations that had not been
previously
documented.
There seem to be three general types of design changes applied to the
TR7 over time:
1. Early production changes - modifications made very early in
production that are reflected on the schematics, in some cases
correcting PCB layout errors.
2. Schematic changes - changes made that are not necessarily
found on production boards, usually implemented after production was
well underway and sometimes shown only on a later revision schematic.
3. Undocumented changes - changes that appear to be factory
modifications, but are not shown on any schematics.
These changes may or may not have been made in production or at the Drake repair
center, depending on the vintage of the board and whether the change was considered critical in nature. I have found all three
so far in the two TR7s that I have been working with. These units both
have the Version 1 2nd IF/Audio board. One unit is serial number 32xx,
the other 34xx. My first exposure to these changes was a mailing list
discussion HERE,
between Carel PA0CMU and Garey K4OAH. That was only the beginning.
Below
are the variations I found on the two boards that I had at my disposal.
Component Designation
Board 32xx
Board 34xx
CR1101 AA119
1N4148
AA119
R1112 22K
Connected to supply side of
RFC1104 (+10) by factory modification.
Connected to supply side of
RFC1104 (+10) by factory modification.
R1127 1K
Connected to +10R through
RFC1101 by factory modification.
Connected to +10R through
RFC1101 by factory modification.
R1187 10K
1.5K added in series with anode
of CR1119 by factory modification.
1.5K added in series with anode
of CR1119 by factory modification.
R1188 180 Ohm
Not present
Not present
R1120 8.2K
Present
Missing - possibly removed to provide more IF gain
R1187 and R1188 were added to correct loud pops in the speaker when
switching between CW and AM modes. The value of R1187 does not seem to
be critical. 10K will cause the rig to "wake up" somewhat slower when
switching to CW mode than 1.5K. The change to 1.5K may have been a
service change to resolve customer complaints about slow switching to CW
mode. The positions of R1187
and CR1119 as factory installed on the Version 1 board are reversed
from what is shown on the schematic, with the resistor on the anode
side instead of the cathode side of CR1119. This is shown in the
Version 1 service manual photo. The R1112 connection was changed
to correct a pulsing of the T/R relay when switching to CW mode,
although in my opinion it should be connected to the circuit side of
RFC1104 instead of the supply side for proper supply decoupling. The
R1127 connection to +10R appears to have been changed in production to
correct a PCB layout error. These changes should be made to any old
Version 1 boards that do not have them.
Comments from Joe KC9LAD:
"That 1k resistor in the picture is on the
back of the board because that revision of the board had an error in the
traces and to fix it the resistor was put on the back of the board
between the correct points. The resistor is R1127 1k and it goes between
R1128 S-Meter zero pot and +10V receive. This was corrected in later
revisions of the board so the resistor is on the top of the board. On
some of the later revisions you will find a 4.7k resistor on the back of
the board. This goes between the junction of R1111 and CR1105 to R1105
S-Meter sensitivity pot. This was done to bring the adjustment into
range. I have not seen the 4.7k mod on any schematic.
"On version 2 of the 2nd IF/Audio board there was a mistake in the
foil layout which creates a direct short across R1187 10k resistor. The
fix was to cut the trace on the bottom of the board to put R1187 back in
the circuit. This will stop the pop when switching to CW.
Thanks Joe!
R1112 Relocation by Factory Modification
PCB Trace Cut and Added Jumper for R1127 Correction by Factory Modification
An alternate version of the R1127 mod found in K8AC's article on TR-7 Key Clicks R1127 has been relocated to the bottom of the board.
The S-Meter Circuit
Version 1 S-Meter
circuit
Version 2 S-Meter
circuit
Documented variations in the S-Meter circuit between Version 1 and
Version 2 include R1101 being replaced
by another pair of 1N4148s in series, CR1113 and CR1114; CR1101 is
removed and replaced with a direct connection. Carel PA0CMU reports a
unit with CR1101 replaced with a 4.7K resistor. R1111 has been changed
from 2.2K to 3.3K. Confused yet?
OK, back to the S-Meter calibration. As I noted above, all this started
with an inability to properly adjust the S-Meter sensitivity. The TR-7
service manual states that a "properly aligned" receiver should read
S-9 with a signal level of approximately 30 uV on all bands (page 3-3).
Furthermore, the document Service
Group TR7 Alignment Procedures
posted on the Drakelist
web site states that S-9 will typically be -73 to -77 dBm (30 To 50
uV). I was seeing more like 100+ uV using either alignment procedure.
I am usually very reluctant to "modify" a circuit to achieve proper
operation that should have been there in the first place, but both
boards that I had on hand exhibited similar behavior and I was
relatively confident that there was nothing "wrong" with either board.
I modified the Version 1 2nd IF/Audio boards to reflect the
S-Meter circuit changes that Drake made in later revisions as follows:
R1101 - Replaced by 2x 1N4148
R1111 - Changed to 3.3 K
I also replaced CR1101 with a 1N6263 Schottky diode.
This is the only non-Drake change that I made. In later revisions,
CR1101 is eliminated entirely, but I like the idea of having some
isolation between the wattmeter circuit and R1105/Q1102. Of the two
boards I had on hand, one had a Germanium diode and the other used
Silicon. The schematic and parts list specify a Germanium part, so I
used a 1N6263 Schottky as a compromise; similar voltage drop to Germanium,
better switching characteristics than Silicon.
After the changes were made, I was able to get much closer to the
desired 50 uV for S-9 figure, which seems to be the de-facto standard
for "modern" rigs. Final results using a combination of the two
alignment methods noted above are S-9 = 42/59 uV (AM/SSB) for radio
SN34xx, and 40 uV for radio SN32xx (I did not check both AM and SSB on
this sample). Close enough ... for now.
Update July 26 2016
I was never totally happy with the results of my S-Meter experiments so I
have reverted to pretty much the original configuration in the main
rig. My current practice is to set S9 = 30 uV with the IF Gain
adjustment (R1136) after the other S-Meter/AGC adjustments are
completed. The IF Gain is only vaguely mentioned in the service manual
and conventional wisdom seems to have been to just set it to max. I have
found that tends to make the RX too "hot". The sensitivity is not
improved by setting it to max. That just makes the radio SEEM more
sensitive because there is more system gain. It also makes the S Meter
very generous. I prefer to have the S-Meter reasonably close to a
standard and use a preamp on the high bands if I want more gain so that weak signals develop some AGC action.
Here is the adjustment procedure that I currently use. This is the
service manual alignment procedure from page 3-1, with my modifications highlighted:
3-2
S-METER ADJUSTMENT
The S-meter adjustments are located on the 2nd IF/Audio Board (Section
2-11). These adjustments must be made in the AM mode (to prevent errors
due to BFO leakage), and must be made in the order outlined below.
Failure to do so will result in improper S-Meter and AGC operation. (Only the pedestal adjustment is
critical for proper ACG operation). Most adjustments should be made
with the shield cover in place; you will need a long, narrow
straight-blade screwdriver to reach the adjustments near the bottom of
the board. The screwdriver shaft should be insulated with heat shrink
tubing or electrical tape to prevent accidental shorts. A small
flashlight will prove useful to sight on the adjustments through the
provided holes in the shield cover. NOTE: The +10 volt regulator
adjustment must be properly set before any adjustments are performed,
or improper AGC action may result.
a) Adjust the IF Gain
control, R1136, for maximum (fully CCW).
b) Install the shield cover unless you have an early version that does not have holes for all of the adjustments,
and ensure that all screws are in place and snug. Select the AM
mode and remove any antennas. Connect
a calibrated signal generator to the antenna jack, set to 0 output.
c) Turn the zero adjust, R1128, fully CW.
d) Turn the ACG pedestal adjust, R1129, fully CW
e) Rotate R1129 CCW until the S-Meter reading increases 2 S-units.
f) Adjust R1128 for an S-meter reading between 0 and 1. Be sure
that the S-meter is slightly above 0.
g) Turn the front panel RF Gain control fully CCW
h) Adjust the sensitivity control,
R1105, for a reading of S-9 +80 dB.
i) Rotate the RF Gain control fully CW and check
to be sure the S-Meter reads between 0 and 1. If not, repeat steps c
through i.
j) Apply a 30 uV (-77 dBm) signal
from the signal generator. Adjust R1136 for a reading of S-9. R1120 may
be removed if more IF gain is needed. Remove the signal and check to be
sure the S-Meter reads between 0 and 1. If not, repeat steps c through
j..
As a final note, here are some comments from Fabio I5LHY reagardng
his own S-Meter experimets (August 2019). It seems there is more than
one way to skin this cat! Thanks Fabio!
"The best configuration I've found is 4x1N4148 in place of R1101, R1111 at its original (here) value of 2.2k.
Calibration pedestal must be set at S4 instead of S2, I do twice the
trimming if the first time the full scale S9+80 is far away.
Final S is in the middle between S0 and S1.
I can set S9 for a signal of 100uV (-67dBm).
The four diodes shiift up the S meter range placing the S1...S9 scale in a better AGC window.
Now I have a S1...S9 measured range of 41 dB instead of the ideal 48 dB.
The missing 7 dB are in the lower S zone.
41 dB is very good!
Going over S9:
S9+10 and S9+20 are aligned
S9+30 at +35dB
S9+40 at +44dB
S9+50 at +48dB
S9+60 at +51dB
S9+70 at +57dB
S9+80 at +62dB
acceptable for me up to 9+50.
Using only 2x1N4148 it matches better over S9 but worse below,
S1..S4 are much worse. 3x1N4148 are better than 2x but worse than 4x.
The minimum signal engaging AGC is always the same, at about -108 dB. So
if we set S9 at 30uV (about -70 dBm) we will have about 10 dB less in
the S meter dynamic wrt S9=100uV case ( -67 dBm).
This ranges are theoretical because we have to see how the AGC circuit
generates the DC voltages from RF signals and how good the S-meter
isdriven.
So the result is a tradeoff between the theory and the real circuit behavior.
Now the noise floor on 7 MHz (10m long vertical) is a S3..S5, before all was much quieter, the signals too!"
Standard disclaimers apply; this is what worked for me, it may not work
for you. Your mileage may vary. I'm not responsible if you damage your
radio. User experiences, comments, and suggestions are always welcome
and may be added to this web page; flames will be ignored.
Acknowledgments: I would like to thank Carel PA0CMU and Garey K4OAH for
their previous work uncovering revisions to the 2nd IF/Audio
board, Thom K3HRN for his tireless work on the Drakelist web site
and email reflector, and Floyd K8AC for the definitive article on
resolving key clicks in the TR-7. Also many thanks to KC9LAD and I5LHY
for thier helpful comments!
all-in-one internet application suite