Heathkit SB-200
Tuned Input Circuit Modification
By:
Date:
email: SB-200@ne7x.com
Web Site: http://www.ne7x.com/
Problem:
There
has been lots of discussion over the years on how to modify the Heathkit SB200
HF amplifier’s input circuit to match a 50 ohm input
impedance on all bands. The discussions have gone from adding a separate low
power antenna tuner in series between the transmitter exciter and the SB200, to
very complex mathematical calculations on changing component values.
When
the input circuit of the SB200 is not at or near 50 ohm impedance, transmitter
exciters requiring a 50 ohm load Z will roll back it’s
RF output power, resulting in reduced drive to the SB200. Reduced input drive
to the amp will result in reduced RF power output from the SB200. This is
especially noticeable at higher frequencies above 14 MHz.
Solution:
To
address this problem, instead of getting my pocket calculator out and doing
lots of complex math, or adding extra antenna tuner hardware on the operating
desk, I decide to utilize a Tektronix spectrum analyzer and tracking generator
to troubleshoot the root cause of the problem. Using the method described in
this article, I was able to visually see exactly what was going on with each of
the pass-band circuits. This includes all environmental conditions and
component tolerances which affect circuit characteristics that math
calculations do not take into consideration. Utilizing a Spectrum Analyzer and
Tracking Generator took all the “guess work” out of getting the SB200 input
circuits matched to 50 ohm impedance.
Disclaimer:
If
you perform this procedure to your Heathkit SB200, perform it at your own risk.
I am providing this web site for informational purposes only. I accept no
liability on your behalf if you screw things up.
15 & 10 Meter Input Considerations:
The information contained in this document worked for
me, however you may need to experiment more with the 15 and 10 meter input
circuits values. There is a problem with the input rotary
switch. If you remove all the 15 and 10 meter components from the circuit
(coil/caps) you will still have capacity coupling in the circuit due to the
capacitance between the switch contacts. The switch Heath used is the wrong
type (contact spacing) for 15 and 10
meters. The proper way to address this would be to replace the switch or come
up with some sort of circuit that cancels out the capacitance which the switch
contacts inject into the circuit.
What is happening, the distance between the switch
contacts is too close. It's acting like a capacitor in parallel with the coil
and the fix mica capacitor. This is very pronounced at higher frequencies,
above 21 MHz. This added capacitance is detuning the circuit. Below 21 MHz, the
added capacitance is so small, it has no effect on the
tuned circuit.
The
way I addressed this was to add some bypassing
from each grid socket pin to the nearest ground tab. I used 0.01uF, 1000pF,
220pF (already there), and 120pF. Each
time I added a capacitor, the SWR and gain improved. The last capacitor I added was 120pF. Since
it already had 200pF, I didn't think 120pF would make a difference, but it made
just enough difference to bring the input SWR to a level that my Yaesu FT-897
would accept. All the caps I added were
ceramic disc. I kept the leads short,
all parallel with each other, and flowed solder between them to make very wide
conductors. Since the additional capacitance was small in value, this
additional circuit modification did not affect 80, 40 and 20 meters. However a
slight adjustment of the 80, 40 and 20 meter coils maybe necessary to bring the
center frequency back into the pass-band.
Low Output on 80 Meters:
Check the 100pf 5KV square
red (pink/rose color) 80 meter fixed HV mica cap that is in the output tank
circuit. That cap is switched in and only used for 80 meters. When it goes bad
it normally blows a small pin-hole in it. Replace this with any HV fixed HV
mica cap +/- a few pf. I find HV pf type caps all the time at local Hamfests.
You find them in the “junk” boxes or coffee cans among all the stuff that no
one looks at. Don’t go by just the physical appearance, you can use doorknob
type also out of old military surplus or air-craft antenna tuners. Note the
brown round cylinder looking cap next to the output tank coil in the pictures
below.
Test Equipment
Used:
· Tektronix 496 1.8 GHz Spectrum Analyzer
· Tektronix TR503 1.8 GHz Tracking Generator
· Tektronix DC502 550 MHz Frequency Counter connected to tracking generator AUX RF out
SB200 on the workbench connected to Tektronix spectrum analyzer and tracking generator
Default
Heathkit SB200 Configuration:
Per
the Heathkit SB-200 manual, dated 1964, part number 595-682-03, here is the original
tuned input configuration and values.
|
Band |
Configuration |
L |
C1 |
C2 |
|
80 meter |
L |
23 turns |
470 pf |
|
|
40 meter |
PI |
16 turns |
310 pf |
510 pf |
|
20 meter |
PI |
10 turns |
200 pf |
360 pf |
|
15 meter |
L |
5 turns |
75 pf |
|
|
10 meter |
L |
3 turns |
66 pf |
|
C16 .02uf
What
I Found with Default Configuration:
- 80 meter pass-band was
centered at 12 MHz
- 40 and 20 meter pass-bands
were both within only a few 100 KHz of their respective bands.
- 15 meters pass-band was
centered low by almost 4 MHz
- 10 meter pass-band
center was down at 45 MHz.
Procedure:
1) Disconnect SB200 from all power
2) Remove from case
3) Test equipment set to 50 ohm load Z
4) Removed both 572B tubes
5) Connected tracking generator ‘out’ coax cable to SO-239 input connector of SB200
6) Connected Spectrum Analyzer ‘in’ coax cable across pin #1 of 572B tube sockets (filament) and chassis ground using coax with alligator clip ends
7) Use Heathkit default stock input coils, un modified, no turns added or removed
8) Placed the iron core slug inside all coils at the middle (half way) point
9) Begin with lowest frequency (80 meters) and work upward to highest frequency (10 meters)
10) Start with default capacitor values, then added/subtracted capacitor values, watching the spectrum analyzer screen as the pass-band moves up/down in frequency
11) Once the pass-band is inside the desired amateur band, adjusted the coil slug to fine-tune the pass-band “peak” for the desired optimum center pass-band operating frequency. Use the frequency counter to make this final adjustment. (note: use non-metallic adjustment tool)
12) Repeat steps 9, 10 & 11 for each band
13) Disconnect spectrum analyzer and tracking generator
14) Insert 572B tubes
15) Reassemble into case
New
Values:
The following values were found to bring each band dead on at 50 ohms impedance for the SB200 which was under test. Your values and findings may vary slightly due to environmental and component tolerances. All capacitors used for the new configurations are silver mica +/- 5% tolerance 1KV.
|
Band |
Configuration |
L |
C1 |
C2 |
|
80 meter |
PI |
23 turns |
910 pf |
910 pf |
|
40 meter |
PI |
16 turns |
310 pf |
510 pf |
|
20 meter |
PI |
10 turns |
200 pf |
360 pf |
|
15 meter |
L |
5 turns |
220 pf |
|
|
10 meter |
L |
3 turns |
100 pf |
|
(See spectrum analyzer screen shots below)
Items
Unchanged:
40 and
20 meters configurations did not required any changes of components from the
factory defaults. The only thing required was a small tweak (turn) of the coil
slugs to bring center of pass-bands to middle portion of bands.
Now
of the coils were modified, no turns removed or added. Utilized
the same factory supplied iron slugs.
Other
Considerations:
I
wonder what effect it would have to the circuit since I did not have the 572B
tubes in their sockets with the filaments conducting current. Would this have
any affect on the output impedance of the tuned input circuit? It seems to me
the input tuning (matching) circuits should be terminated with the same
impedance when testing as during operation. So I temporary placed different
resistor values across the filament pin #1 of the tube socket to ground (47
ohm, 100 ohm and 200 ohm). Watching the spectrum analyzer scope, all it did was
"shift" the frequency up about 10-20 KHz. I though this was quite
trivial, so I left the resistors out of the equation during the rest of the
testing/alignment. However during my fine-tuning coil slug adjustments, I
compensated a pinch-of-a-turn for this shift in frequency.
Block
Diagram of Test Equipment Connections:
SB200
Input Circuit:
- Spectrum Analyzer coax
cable with clip lead ends. (end leads should be as short as possible)
- Connect positive side to pin #1 (red boot) of
572Bs tube sockets and negative side to chassis ground (black boot)
Spectrum
Analyzer Screen Shots:
|
The following spectrum
analyzer scope screen shots are showing pass-bands with the new capacitor
values and circuit configuration. |
3.5 MHz, new PI configuration using 910pf capacitors
each side of pass-band circuit
7 MHz, original default values
14 MHz, original default values
21 MHZ, new 220pf capacitor on IN side of pass-band L
circuit
Note the capacity bumps to
the right of main pass-band peak for 21 and 28 MHz. This is caused by the
capacity of wire leads connected to the “OUT” side of coil. This additional
capacity becomes very noticeable above 14 MHz. When attempting to make the 21
and 28 MHz circuits into PI configurations, this “OUT” bump peak becomes equal in
vertical amplitude to the “IN” peak, resulting in a very broad tuned input
which is not 50 ohms Z. This is why the 21 and 28 MHz configurations are L
instead of PI.
28 MHz, new 100pf capacitor
on “IN” side of pass-band L circuit
Summary:
The tests which I have performed matches the gurus advice in
Feel free to email
me (
Additional
Information:
I have received several
emails asking for 6 meter tuned input information. The following two links are
a great source for this information:
- Carl KM1H provides a 6 meter conversion kits for the SB200.
Email Carl at: KM1H@juno.com for more information.
- The “50 MHz Modification
Pages” web site has multiple articles on converting SB200 for 6 meter
operation.
- Some good circuit modifications and improvements can
be found at: http://www.palacenet.net/home/mules/sb200.html
- Restoring tips by: N4JK can be found on his web
site. http://members.cox.net/n4jk/sb200.htm
73s,