AN SB-220 COMPENDIUM

MODIFICATIONS & UPGRADES

INSTRUCTIONS
SCHEMATICS
PHOTOS
MATERIAL LISTS
SOURCES of SUPPLY
PERFORMANCE MEASUREMENTS

by

John White VA7JW

Original: 17 November 2011
Rev 1: 13 June 2014
Rev 2: 3 Feb 2014

3 Feb 2014

VA7JW

AN SB-220 COMPENDIUM
INTRODUCTION
This Compendium is a collection of upgrades, modifications and performance measurements
applicable to the Heathkit SB-220 RF Power Amplifier. These may be of interest to owners and
users of this venerable and popular amplifier.
The authors SB-220 amplifier was purchased from the original owner in 1988. The amplifier was
complete, unaltered, fully functional and in very nice condition. Through the years 1992 to 1998
extensive work was undertaken to upgrade operational features, replace aging parts and tame
parasitic oscillation tendencies. All changes have been implemented by the author and have been
in service for over 10 years. No failures or performance issues have arisen.

Part 1 UPGRADES and MODIFICATIONS

This part of the document provides a number of upgrades and modifications commonly published,
plus some by the author not seen elsewhere.
CREDITS
Allen Harbach WA4DRU of Harbach Electronics http://www.harbachelectronics.com/ offers
kits which come with complete installation instructions. The kits as listed are current whereas
the kits used in this amplifier were purchased over 10 years ago. There are likely some
differences and probable improvements but functionality is expected to be the same.
Richard Measures AG6K http://www.somis.org/ provides a wealth of information on most
aspects of RF power amplifier operations, modifications and upgrades. AG6K provides
information rather than kits in general although many parts are available on his web site.
A special mention of a very nice publication called Heathkit Upgrades for the Linear Amplifier
Model SB-220 by Anton Suarez [email protected] came to my attention after this work was
essentially completed.

There are common modifications in both independent works, but there are also unique
modifications which expands the scope of the possible re-work that can be done. The manual
has been posted on eBay, for about $10 from time to time. Search under HEATHKIT SB-220
Amplifier Upgrades. The seller ID is tito_de_taboada

Part 2 PERFORMANCE MEASUREMENTS

This part of the document provides performance data as measured on this amplifier. RF carrier
power output, PEP measurements, as well as spectral harmonic and two tone tests and a
unique noise loading test are presented.

3 Feb 2014

VA7JW

AN SB-220 COMPENDIUM
INDEX
Part 1 Upgrades and Modifications
Items 1.3 through 1.12 are stand-alone modifications; they can be done independently of each
other.
Items 1.13 and 1.14 describe an integrated QSK feature and require additional skills.
Section 1.1
Section 1.2
Section 1.3
Section 1.4
Section 1.5
Section 1.6
Section 1.7
Section 1.8
Section 1.9
Section 1.10
Section 1.11
Section 1.12
Section 1.13

Skills and Safety Notes

A Quick Overview
Parts Notes
Parasitic Suppression t
Filter Capacitor
Rectifier Metering Board t
Inrush Current Limiter
AC EMI Line Filter
Hour Meter Filament time
Pilot Light
Adjustable ALC Standby Operate & Fan switch
120V Bias Supply Upgrade

Section 1.14 QSK Circuits

Section 1.15 High Speed Vacuum Relays
Appendix I
Appendix II
Appendix III
Appendix IV
Appendix V

Harbach PS-220 kit or per AG6K kit

Harbach FB-220 kit or per AG6K parts
Harbach RM-220 kit or per AG6K parts
Harbach SS-220 kit or per AG6 design
VA7JW modifications
VA7JW modifications
VA7JW modifications
VA7JW modifications
VA7JW modifications
VA7JW modifications
AG6K design as implemented by VA7JW
AG6K design as implemented by VA7JW

QSK Keyer Schematic

Current Source Detail
Current Source / Keyer Construction Detail
Kit Listings
Bill of Materials for QSK System

Part 2 Performance Measurements

Section 2.1
Section 2.2
Section 2.3
Section 2.4
Section 2.5
Section 2.6
Section 2.7
Section 2.8
Appendix VI

3 Feb 2014

Instrumentation
Carrier RF Power Output
PEP RF Power Output
Harmonic Distortion
Two Tone Test
Noise Loading Test
SB-220 Specifications
A Little History
Power Tutorial

VA7JW

AN SB-220 COMPENDIUM
PART 1 - UPGRADES and MODIFICATIONS
1.1 SKILLS and SAFETY NOTES
The modifications listed herein will variously require removal of covers, side panels, the faceplate
and disassembly of mechanical parts. There will be reworking of electrical circuits, drilling of
holes for new mounting screws in the chassis, and re-dressing of wiring. Some modifications are
more complicated than others. In all instances, you should be experienced and comfortable with
bench work, have tools for drilling, cutting, filing, soldering etc, possess good workmanship skills,
and be able to read schematic diagrams. Note that the QSK section requires some scratch
building experience. Use a DVM to check wiring for shorts, opens, continuity, correct resistances,
and voltages under powered up conditions. Do NOT try to directly measure the 3.5 kV High
Voltage with your DVM. The author does not provide step x step instructions for the mods and
rebuilds herein. Portions of the original Heathkit schematic are presented in black with the
modifications indicated in RED.
FIRST and ALWAYS - UNPLUG the 120/240 VAC LINE
Then remove other cables
LAST - Remove the Green Wire Safety Ground from the Chassis Ground bolt.
The amplifier is heavy. Removal of case can be simplified by cutting a 5 long block off a wooden
2 x 6, place and holding it on the rear panel at the back of the power transformer, tip the amp up
on end, on the block, until it is upright. The amp will balance on the block because of the heavy
power transformer and the connectors on the back will be clear of the bench surface.
Remove the four feet, but before removing the last foot, prevent the cabinet from sliding down off
the chassis on to the bench it may be difficult to lift the cabinet back up and off as the cabinet
tends to snag on the chassis. Place a 2 x 2 wooden block on the bench under the edge of the
case. Remove the last foot and lift the cabinet straight up off the chassis. The chassis can then
be tilted back down to the horizontal for work. Reverse the process for re-installing.
Before proceeding with any modifications, be advised to remove the 3-500Zs from their sockets
using a clean cloth to keep finger oils and other contaminates from being deposited on the glass
bulb. Then wrap each tube in a protective cloth (towel) and put somewhere out of harms way.
The glass bulbs need to be squeaky clean on reinstallation.
Do not leave clippings, drill debris or any other conductive material loose inside the amp. Take
care that there is no possibility of short circuits developing between components or circuit boards
to chassis or to each other. Ensure that the integrity of wiring insulation and coaxial cable
dielectrics are maintained.
There is a High Voltage (HV) safety shorting switch in the tube compartment. Do not power up
amp with RF compartment screen off as the HV will be shorted to ground. If you defeat this switch
when servicing, understand the peril to which you expose yourself should you contact the HV.
LEATHAL AC and DC VOLTAGES EXIST
USE EXTREME CARE

Disclaimer
Author only provides best of knowledge information only.
Author makes no warranties or guarantees regarding these modifications.
Author is not liable for costs, loss, damage or injuries due to undertaking such work.
This work should only be done by those who are competent to do so.
High voltages exist and troubleshooting is dangerous.
Check your work before powering up.

3 Feb 2014

VA7JW

AN SB-220 COMPENDIUM
1.2 A QUICK OVERVIEW
Figure 1 gives a quick insight as to the nature and extent of the under-chassis changes described
in this document.

Figure 1.1
UNDER CHASSIS VIEW of UPGRADES and MODIFICATIONS

1.3 PARTS NOTES

A list of materials is provided mostly using Digikey http://ca.digikey.com/ part numbers. All parts
are stock items at the time of writing. Digikey is recommended due to the extensive and diverse
stock available and quick, typically overnight shipping by courier.
If parts are not available from Digikey, alternate sources of supply are provided.
Parts used were selected over 10 years ago. Many are still the same with some being slightly
different. Form, Fit and Function for current parts are specified to the same requirements.

3 Feb 2014

VA7JW

AN SB-220 COMPENDIUM
1.4 THE PARASITIC SUPPRESSOR KIT
It is a common belief that the SB-220 is vulnerable to uncontrolled high power VHF parasitic
oscillations that may occur when the amp is keyed into transmit. Instantaneous destruction of
certain components in the cathode circuits follows. The author has experienced this failure mode
and application of available VHF Parasitic kits has remedied this problem. Whether it is an
oscillation or not is not altogether clear and there is differing opinion as to the actual cause.
AG6K supports the VHF position whereas W8JI has differing thoughts on the subject,
http://www.w8ji.com/Amplifiers.htm . Many other interesting topics are discussed.
The cause will not be debated here but the effect of the failure is not subtle a loud bang,
possibly some smoke, and an amplifier that may have too high an idle current or none at all. The
components prone to failure are ZD-1, a 5.1V stud mount zener and R3, 0.82 ohm / 2 W resistor
on the rectifier board. They usually blow open circuit but may also short. The stud mount zener is
expensive and is most often replaced with a series string of rectifier diodes (Section 1.6 and 1.14,
pg 19).
Part
5.1V Zener 1N3996A, (obsolete)
10 watt, 5%, DO-4 case, cathode to stud
0.82 ohm / 2 W resistor
Rectifier Diode, 1N5408G (3A /1000V), qty 8

Digikey p/n
None listed
NTE-5177AK *
P0.82W-2BK-ND
1N5408GOS-ND

Approx Cost
$14+
0.39
0.60 each

* NTE supplies replacement parts to distributors for obsolete components. Newark Canada,
http://canada.newark.com/ can supply the NTE part under their own part number 29C4650.
The Harbach PS-220 kit is shown below. The plate circuit discrete resistor / coil suppressors are
shown going to each plate cap from the plate choke and to the TUNE cap is another L and R
suppressor although the R in this case it is a Nichrome (resistance) wire.
Plate L and Discreet Rs

Output L

Nichrome R

Figure 1.2
PS-220 Parasitic Suppressor Kit Plate Circuits
Additional suppression components are installed on the tube sockets under the chassis as shown
in Figure 1.3. These consist of various Rs and Cs in the grid and cathode circuits. Note that
W8JI and others recommend DIRECT grounding of the grid to chassis using short, wide, multiple
copper straps to chassis rather than installing Rs and Cs.

3 Feb 2014

VA7JW

AN SB-220 COMPENDIUM
0.01uF / 10 ohm 2 places
820 pF

820 pF

27 ohm / 820 pF two places

Figure 1.3
PS-220 Parasitic Suppression Grid- Cathode Circuits
Both kits from Harbach and AG6K have been installed in other SB-220 amplifiers. They are
somewhat different from each other but both are supplied with excellent instructions and both
have worked reliably.
1.5 HV FILTER CAPACITOR KIT
The SB-220 amplifiers were sold between 1970 and 1978. The follow-on SB-221 was sold
between 1978 and 1983. If you are still using the original filter caps it is time for a change before
they fail. These caps have a ripple current flowing through them as a result of rectification which
results in internal heating due to their ESR (Equivalent Series Resistance). External heating is
also prevalent in this closed space. Heat is very detrimental to the longevity of components in
general and particularly electrolytic caps. They slowly loose capacitance with increasing ESR and
may eventually rupture due to heating, possibly explosively. Expected life time of caps of this
vintage is in the order of ~ 25 years.
Harbach supplies a complete kit of properly sized capacitors and voltage equalizing - bleeder
resistors. AG6K also offers replacement capacitors and resistors as individual components. Be
certain the capacitors ordered for the SB-220 amp will fit the Heathkit capacitor block.
This mod is requires delving into the power supply band switch compartment. Good idea to do
both capacitors and rectifiers, next section, at the same time if never replaced.
1.6 RECTIFIER METERING BOARD KIT
Harbach offers the RM-220 kit, an improved PWB metering board with protective diodes for the
meters, modern low leakage rectifier diodes, new HV resistors and a diode bias string to replace
the 5.1V Zener diode. AG6K also offers components.

Figure 1.4
RM-220 Rectifier and Metering Board Upgrade

3 Feb 2014

VA7JW

AN SB-220 COMPENDIUM
1.7 INRUSH CURRENT LIMITER
The SS-220 kit offered by Harbach limits the inrush current to the filaments of the 3-500Zs. The
tubes are rated ~ 15 amps at 5.0 volts each (R = 0.33 ohms) when hot. Total current is twice that
as there are two tubes. The cold resistance is lower than the hot resistance. On power-up, there
is some concern that a high inrush current may shorten the life of the tubes or damage them.
AG6K reports that the Heathkit filament transformer T2 has an inherent current liming capability
which limits inrush current and so the usefulness of this modification is debatable. Regardless,
the limiter was installed to reduce the stress on components throughout the power system
circuits. AG6K also suggests a circuit for reducing inrush current.

Figure 1.5
SS-220 Inrush Current Limiter
The SS-220 kit is mounted close by the input 120 / 240 VAC terminal board. The board is shown
strapped for 240 VAC operation with terminal 2 and 3 connected. The two red insulated leads
are across the 120 VAC portion of the circuit go to the Hour Meter module, Section 1.9.
Ensure the PWB is well insulated from chassis; use tape or insulating paper. Mounting hole
drilling of the chassis required to mount the PWB.
1.8 AC EMI LINE FILTER
Overview
There is no effective RF AC line filtering on the SB-220. RF fields and currents inside the chassis
can couple to the AC wiring and can be conducted out on to the AC power circuits where the RF
can radiate or be conducted elsewhere to cause RF problems within the shack or the household.

Figure 1.6
Curtis RF Line Filter
It would seem prudent to install readily available RF filters in the amplifier to ensure that the AC
power systems are eliminated as a source of RFI. A Curtis brand filter rated at 20A / 220 VAC,
model number F1700AA20 was used at the time.

3 Feb 2014

VA7JW

AN SB-220 COMPENDIUM
As an additional note, all other leads external to the amplifier ought to be of high quality shielded
wire; flexible coax such as RG-58 works well. Bypassing them at the amp apron is a good idea.
Refer to section 1.9 following.
Common Mode attenuation is > 30 dB at 1 MHz rising to 40 dB at 30 MHz, and somewhat higher
in between these end points. Differential Mode attenuation is 60 dB falling to 55 dB respectively.
A better choice would be a (Tyco) TE Connectivity Corcom filter, p/n 20VSB1. Attenuation
performance is better than Curtis and about half the price.
Part
Curtis Filter F1700AA
Corcom Filter 20VSB1

Digikey p/n
364-1086-ND
CCM9069-ND

Approx Cost
$42
$24

Also installed are Metal Oxide Varistors (MOVs). These devices are surge rated, transient
voltage protection devices. They will clamp voltage transients to relatively safe values should
they appear on the AC line from external sources. The 240 VAC household branch circuit is not
protected / filtered at the wall socket as are the 120 VAC circuits in this shack. The intent is to
clamp high voltage AC spikes from over-voltaging components within the amplifier. The
packaging is in the form of a ceramic disk and they look like a disc ceramic capacitor, just thicker.
The original MOVs used were type ZNR14K391 where ZNR is an industry recognized prefix for
MOVs. The part number deciphers as 14 = diameter of the MOV in mm, and 391 = clamping
voltage. Closest equivalent today is the Panasonic ERZV14D431 which is of the same diameter
and the clamp voltage is 431volts rather than 390 volts.
Part
Panasonic ERZV14D431

Digikey p/n
P7267-ND

Approx Cost
Qty 3 @ $1.04 each

Schematic

Figure 1.7
Complete Circuit Diagram for MOV, RF Line Filter and Current Limiter

3 Feb 2014

VA7JW

AN SB-220 COMPENDIUM

Installation Photo of Inrush Current Limiter & EMI Filter

The Inrush Current Limiter is on the left and the EMI filter is on the right. Further to the right are
the original circuit breakers. The MOVs are on the back of the filter, one just visible as a black
object. The three MOVs are mounted on the filter terminals.

Figure 1.8
Inrush Current Limiter & RF Line Filter Installation
Original AC Circuit breakers on right hand side
1.9 HOUR METER
Overview
This device records the hours accumulated on the tube filaments, that is, cumulative time when
the amp is powered ON. It is not RF hours which are much less.

Figure 1.9
Curtis LCD Hour Meter
The Hour meter is a Curtis Instruments LCD Hour Meter, Model 701, 3 wire. Hexagonal case
code F. Nominal voltage operation range from 100 to 230 VAC.

Part
Curtis 701FR001048150D100230A
Ferrite Beads, Amidon Assoc. FB-73-801

Part number
DigiKey 267-1004-ND.
RPE 73-801

Approx Cost
$32
Qty 8 for $10

Order Ferrite beads direct from Amidon Associates. https://www.amidoncorp.com/

Product is also available from RP Electronics in Burnaby. http://www.rpelectronics.com/
Beads plus 0.01 uF disk caps to ground make for a reasonable RF filter. The link in Section 1.13
regarding capacitors is a great place to shop for many different capacitor types (mica, disk, poly,
electrolytics etc) and values. All new, reasonably priced and some hard to find anywhere else.
The beads were not used to remedy malfunctions due to RF, but were installed as a cautionary
measure.

3 Feb 2014

10

VA7JW

AN SB-220 COMPENDIUM
Schematic

Figure 1.10
Hour Meter Wiring and Bypassing
The wiring is run down through the chassis fan cutout and along underneath the chassis to the
input voltage strapping terminal board. Bundle wires with existing filament and power wires for
neatness.
Installation Photo
The meter is mounted on rear of amplifier. The screen has to be cut away to fit the body through.

Figure 1.11
Hour Meter Mounted on Rear Panel
1.10 ON / OFF PILOT LIGHT
A green LED lamp is installed on the lower, left hand side of the front panel. The lamp is ON
when AC power is ON. It is not a Transmit On lamp.
Lamp voltage is taken from the filament AC supply, same as the meter pilot lights. The LED is
series connected with a 200 ohm resistor with a 1N4005 diode across the LED to ensure that the
LED does not breakdown on AC reverse polarity. The resistor limits the LED current. This circuit

3 Feb 2014

11

VA7JW

AN SB-220 COMPENDIUM
is connected to the cold (RF blocked) side of the filament choke. Junk box parts were used to
make this LED lamp but if you buy a lamp assembly, a series resistor is usually incorporated in to
that assembly. The product listed below operates from 6 VDC which is fine even though the
filament supply is approximately 5 VAC. The lower voltage may result in a slightly dimmer LED.
The AC reverse diode should still be installed across the LED assembly to protect it. Ensure that
the pilot light case is long enough to mount through the thickness of the front panel + the chassis,
about 5 mm is required,
Schematic

Figure 1.12
LED Pilot Light
DigiKey has a line of Dial Light product under part numbers 350-19XX-ND that appear to have a
sufficiently long threaded body (~ 12mm) to pass through the thickness of the faceplate + chassis
(~ 5 mm). Look at others in that series for style and color to suit.
Part
Dial Light Green LED / Black Body

Digikey p/n
350-1908-ND

Approx Cost
$5.27

Installation Photo
Pilot light is mounted on center line of power ON-OFF switch and CW/TUNE SSB switch, and half
way between TUNE and LOAD controls

Figure 1.13
Pilot Light Placement

3 Feb 2014

12

VA7JW

AN SB-220 COMPENDIUM
1.11 ADJSTABLE ALC
The ALC output voltage from the SB-220 is a negative going voltage which is compatible with
most rigs. Many modern transceivers have an ALC adjustment on their rear apron making this
mod unnecessary. Check your transceivers ALC input specs for compatibility.
This is a simple addition to the back apron of the amp. A 100k pot is wired to the ALC circuit
output and the wiper is connected to the ALC jack. The pot can be adjusted such that the
amplifier does not saturate with peak drive from the exciter. Saturation is the level at which the
amplifier will no longer increase output power levels when more input power is applied.
Schematic

Figure 1.14
ALC Control
This amplifier was seen to saturate at about the 1400 watt level (Refer to Part 2). By convention,
the onset of saturation is the point at which the output no longer rises equally, dB out for dB in. At
this point the intermod and harmonic content rise very rapidly with increased input (overdrive)
causing the distortion products to occupy a greater bandwidth (sideband growth). You may get
reports of being wide or causing splatter as voice peaks overdrive the amplifier.
The ALC control on the amplifier is set to limit the output power to about 1300 Watts, just less
than the measured 1400 W saturation point on the authors amplifier. It may be that you do not
have to worry about overdriving the SB-220 if your exciter delivers < 150 watts at the amplifier
input. It takes something just over about 150 Watts to induce saturation.
Installation Photo

Figure 1.15
Pot Location, Rear Apron, beside RF Input Connector

3 Feb 2014

13

VA7JW

AN SB-220 COMPENDIUM
The pot has to be of a small diameter to fit the rear apron, preferably less than 3 /4 diameter and
be a single turn, 100k value.
Part

Digikey p/n
987-1312-ND
or 987-1327-ND

Potentiometer 100k Single Turn

Approx Cost
$0.84
$1.94

1.12 STANDBY OPERATE & FAN SWITCH

Overview
When the SB-220 is powered ON, the fan runs at full speed, and are noisy and bothersome.
The STANDY OPERATE switch when in the STANDBY position, reduces the fan speed by ~
50% which lowers the noise to a near inaudible level, yet keeps the temperatures well under
control. The author measured < 5 degrees C rise on top of the SB-220 cabinet over the tubes.
The ANT RELAY input (keying line) is deactivated in STANDBY so that the amplifier cannot be
switched to transmit with reduced air circulation. When switched to OPERATE, however, the fan
goes to full power and the Antenna relay / keying line is enabled.
The switch is DPDT, ON-ON, toggle Power Switch with a threaded bushing. Note there are no
high currents involved and only 115 VAC to switch. The switch controls both fan speed and the
keying line. A switch is suggested in the parts table following but the user may wish to obtain at a
switch at a local supplier based on a personal preference for appearance and feel.
A resistor would be able to reduce voltage to the fan but it is lossy and generates unwanted heat.
Instead, an inductor is placed in series with the fan providing the required voltage drop due to its
reactance at 60 cycles. The inductor is essentially lossless.
Schematic

Figure 1.16
Operate Standby Switch and Fan Circuit

3 Feb 2014

14

VA7JW

AN SB-220 COMPENDIUM
Installation Photo, Switch
The switch is mounted on center line of the power ON-OFF switch and CW/TUNE SSB switch,
half way between the LOAD and BAND SWITCH controls

Figure 1.17
Position of Operate Standby Switch
Installation Photo, Fan
In this top view at the back of the amplifier, the inductor is mounted on the left side of the
amplifier, along side the fan, and in between the RF compartment and the back plate. The
inductor on the left side. The fan motor is in the middle, and the rear of the Hour Meter can be
seen on the right. The author used a custom built inductor but the Stancor product is a form, fit,
function compatible inductor of the same inductance and current rating.

Inductor

Figure 1.18
Mounting of Fan Speed Inductor
Part
Stancor-Walden, Choke C-2343
Standby Operate Switch suggestion

Supplier
Allied Radio p/n 928-0010
Digikey EG4823-ND

Approx Cost
$18
$5.31

Allied Radio, http://www.alliedelec.com/

1.13 +120VDC BIAS SUPPLY

The bias supply is a half wave rectifier with a 20 uF filter capacitor, suitable for the original relay
coil, cut-off bias and ALC threshold. For use with the QSK circuits, a better regulated, lower
ripple source of +120V was desired although not altogether necessary. It would however be
prudent to replace C4 with a 47uF / 160 electrolytic for the same reasons cited in Section 1.6
regards lifetime of electrolytics. The single diode rectifier was replaced with a diode bridge for
full wave rectification.

3 Feb 2014

15

VA7JW

AN SB-220 COMPENDIUM
Schematic

Figure 1.19
Bias Supply Modification
Part
Rectifier Bridge 1.5A 400 V

Digikey p/n
W04GDI-ND
4180PHBK-ND
*

Capacitor 47uF / 160 V axial leads

Approx Cost
$0.53
$4.82
$1.10

* Alternate less expensive source of capacitors for all types is a Canadian source in Ontario at
http://www.justradios.com/orderform.html. There is a minimum order of $15.
Installation Photo
The modified Bias Supply is built on the original tie strip per Figure 1.20. The exact wiring is
obscured in the photo and so Figure 21 shows how it is done.

Figure 1.20
Component Placement

Figure 1.21
Component Wiring

3 Feb 2014

16

VA7JW

AN SB-220 COMPENDIUM
1.14 QSK CIRCUITS
Introduction
The original Heathkit open frame, RF un-friendly, clunky, mechanical relay is unsuitable for Break
In operation. The relay coil is connected to the ANT RELAY jack on the amplifier and presents
120VDC to the exciter keying device on Receive. On Transmit, the antenna relay coil current is ~
25 mA. More importantly, inductive kickback voltage is developed when the keying line goes
open as the coil discharges in to an open circuit where destructive voltages can develop
threatening semiconductors. This is OK if the exciter has a relay output for keying amplifiers but
may not be acceptable to modern solid state keying circuits in the exciter.
The QSK circuit will allow fast switching of new antenna relays which replace the Heathkit relay.
As well, the keying line has only ~ 1.5V present on receive and < 1 mA of keying current on
transmit.
The QSK feature will allow Break-In CW operation. Relay switching times are specd at 8 ms
maximum with typical < 5 mSec for the relays chosen per Section 1.14. Receiver recovery time
and keying delay in the transceiver will also affect the ultimate Words Per Minute (WPM) that this
system will support.
According to W4JBM, http://www.hamuniverse.com/w4jbm/emission.html , WPM = 1.2 / (dit
length). A series string of dits of 30 mSec (0.030 second) duration each would be equivalent to
~ 40 WPM. The CW keyer was set up to deliver dits of ~ 30 mSec. A Tek 465 scope was
connected to the dummy load to view the RF with the scope triggered from the keying line.
Figure 1.22 shows the timing of the keying line and the RF.

+V
Keying Line

Ground

RF

Figure 1.22
Relay Timing
The scope sweep was set to 10 mSec / div. Top trace is keying Line and the bottom trace is RF
out of the amplifier. Dit length is about 30 mSec. Note that the keying line activates first, the +V
to ground transition and about 5 mSec later, the RF output appears and looks clean on the
leading and trailing edges as it should. The exciter has an embedded 5 mSec delay from onset of
keying to RF out and this appears to be reflected in the traces above. Similarly, when the key is
opened, the RF is turned off and the keying line is delayed about 5 mSec as well. This system
allows the relays to cold switch, i.e. no RF on the contacts either on leading or falling edges.

3 Feb 2014

17

VA7JW

AN SB-220 COMPENDIUM
Figure 1.23 shows that signal is heard between the dits at the same speed on Full Break In.
This photo is taken from the SB-614 in-line monitor scope. The time base is not calibrated.

Figure 1.23
Full Break In Noise between dits

Figure 1.24 is a block diagram of the QSK system. This provides a top level view of the
functional blocks that will be installed.
Installing the QSK circuits is an undertaking. It requires the assembly of various purchased parts
and kits and the scratch building of the current source per Appendix II.

Figure 1.24
QSK Block Diagram
You will need to buy the Harbach Soft Key SK-220. You will also have to scratch build the
Cathode Bias Keyer as AG6K has designed. The QSK feature also requires removing the original

3 Feb 2014

18

VA7JW

AN SB-220 COMPENDIUM
Heathkit antenna relay and the installation of the high speed switching relays. The rectifier circuit
board will also require modification although replacement using the Harbach RM-220 Rectifier
and Metering Board kit is recommended as it will ease this task. Also recommended is the
upgrade of the +120V PS of Section 1.13 to provide a better quality +120V which the QSK circuits
will use.
The Antenna Relay jack on the back of the amplifier was used to switch the antenna relay from a
pass-through-on-receive position to inserting the amplifier in-circuit on transmit. The Antenna
Jack function is now changed to a Keying Line input and is connected to an electronic interface
that keys the amplifier in a QSK mode when the amplifier switch is in the OPERATE position. For
this reason, it is required to install the Standby-Operate Switch feature of Section 1.12. The QSK
system still provides the same function of switching the antenna as did the old relay.
Operation
A full schematic of the QSK circuit appears in Appendix I
The current source takes the unregulated +120 VDC bias supply and outputs a constant current
of about 80 mA in the OPERATE mode when the Keying Line is grounded on Transmit by the
exciter. The current is adjustable and is set up just once with the potentiometer on that board.
This current drives all the circuits down the QSK totem pole circuit. Refer to Appendix II for
construction and current setup detail.
Two relays are required to switch the antenna path from bypassing the amp to through the amp.
In this implementation the relays are Jennings RJ1A with the 26.5V / 335 ohm coils. The higher
coil resistance provides a higher voltage drop than would 12 V coils with 80 mA flowing. This
lowers the voltage drop across the current source transistor thus easing its power dissipation
requirements. Even so, a heat sink on the transistor is required. If 12V coils are used, then the
power dissipation of the current source transistor will increase and a larger heat sink would be
required. The diodes across the coils absorb the inductive voltage kick-back as the coils
magnetic field collapses when the relays are de-energized; that stored energy can result in very
high voltages developing if the energy has no where to go.The installation of the Vacuum relays
is covered in Section 1.15 following.
The Cathode Bias Keyer is a solid state switch that controls the bias current flowing in the tubes.
The transistor switch Q3 is turned on or off by the opto-isolator diode due to on or off current
flowing through the QSK Totem Pole circuit. This applies or removes bias to the 3-500Z tubes.
When the amp is in the STANDBY or OPERATE mode, but not keyed on, Q3 is off (open circuit).
Unlike the original Heathkit design where +120V is applied to the tube cathode (actually the
filament) to turn the tube off and prevent conduction, this circuit simply goes high impedance with
Q3 being off. However, Q3 is never perfectly off as there is always a leakage current through Q3,
which can be represented by a high resistance of some megohms. Current from the tube will flow
through this high R developing a self biasing voltage that acts to turn the tube off. This self offbiasing technique works very well. Interestingly the voltage required to cut off the tube is only in
the order of +25 VDC.
When the amplifier is in the OPERATE mode and the Keying Line is grounded, the 80mA flows
through the relays, and the opto isolator turns on Q3 (closed circuit). Q3 is connected to the bias
voltage developed on the Rectifier board. This voltage is in the range of 5 to 6 VDC and is low
enough to just put the tubes into conduction. This places the tubes in Class AB2 linear operation
for SSB. The tubes will of course fully conduct when RF is applied to the amp in the case of
carrier such as CW, RTTY and when speech is present for SSB.
Note that 5.1V Zener ZD-1 on the original rectifier board is replaced with a series diode string to
develop the bias voltage needed to establish resting anode current on transmit with no signal
present if the RM-220 kit is installed. This tube current is approximately 120 mA or 60 mA per

3 Feb 2014

19

VA7JW

AN SB-220 COMPENDIUM
tube. The higher the current the more linear is the amplifier performance. A range of 100 to 200
mA is acceptable.
The Harbach SK-220 Soft Key kit is used for the interface to the Keying Line. In the STANDBY
mode, the keying line is switched open so that the amp cannot be turned on. In the OPERATE
mode, the amp will not come on in the presence of an open circuit on the keying line as
presented by the transceiver. When the keying line is grounded, the amp will transmit RF in
response to the signal presented by the transceiver. The voltage on the keying line presented by
the SK-220 to the transceiver is about 1.5V on Receive, and when grounded for Transmit, the
current is ~ 1 mA. This greatly reduces stress on the transceiver amplifier keying line output.
The Harbach RM-220 Rectifier Metering board provides the bias voltage needed by the Bias
Keyer. It is recommended that this kit be purchased although it is not absolutely necessary. The
original board can be used but the 5.1V Zener is removed and replaced with a more robust power
diode string of up to eight forward biased 1N5408s. One can tap off between the diodes to
adjust the idle current as desired although between 7 and 8 at the top of the string usually works
out quite OK.
1.15 ANTENNA RELAYS
Overview
The original 120V relay is replaced with a pair of fast switching, RF, high speed vacuum relays.

Figure 1.25
Typical Vacuum Relay
Manufacturers are Jennings Technology, model RJ 1A, Gigavac model GH3 and Kilovac
Technologies model HC-3.
These are essentially interchangeable. This implementation used a pair of used Jennings RJ 1A
vacuum relays, SPDT. Typically all of these relays are available on eBay from about $40 to $100
US each, new & used.
Schematic

Figure 1.26
Antenna Relays

3 Feb 2014

20

VA7JW

AN SB-220 COMPENDIUM
NOTE: 26V / 335 ohm coils were used for reasons cited (higher voltage drop) in Section 1.14
One can use 12V / 80 ohm coils but a modification to the constant current circuit will have to be
made. Refer to Appendix II.
Installation Photo
Two relays are required to perform the Antenna Amplifier, In or Bypass, function.
The relays are acoustically isolated from the chassis to minimize contact switching noise. The
relays are not silent when switching and while this is not an issue with SSB or other forms of
carrier transmission, CW can be otherwise noisy if using Break-In.
A relay mounting bracket is a custom fabricated from a piece of 1/8 x 3 /4 x 1aluminum angle.
Two U shaped slots are cut into the bracket to hold the relays. The bracket is NOT screwed to
the chassis but is acoustically isolated through use of a foam pad with double sided adhesive
tape as shown by the cross shaded area at the bottom of the bracket, Figure 1.28. Refer to
Figure 1.1 for approximate positioning.
Relay 1 Input

Relay 2 Output

Figure 1.27
Relay Installation

Figure 1.28
Relay Bracket Fabrication Drawing - Typical

3 Feb 2014

21

VA7JW

AN SB-220 COMPENDIUM
The relay bodies slide into the two U shaped slots. The slots are wider than the threaded part of
the relays. An O ring is slipped on to the threaded body which sits in the slot. Flat 1 rubber
hose washers are placed on either side of the O ring such that the relay body has no metallic
connection / acoustic path to the bracket. A flat thin metal washer provides a smooth hard
surface for the relay nut to tighten up against. The assembly is then lowered in to the slot and the
nut is done finger tight only. The washer materials were sourced at Home Depot.

Figure 1.29
Washer and O Ring Components

Figure 1.30
Side View Relay Mounting
Overall, the relay switching noise is probably acceptable to most. Coincidently, the April 2011
issue of RADCOM magazine makes similar statements in their review of the ALPIN 100 HF
amplifier. The switching speed between receive and transmit was extremely fast, settling in
under 3 ms. The Gigavac vacuum relay used in the amplifier is specified at less than 6 ms
switching speed. Also stated, The relays are quiet for normal use but become rather noticeable
with full QSK.
Wiring
Wire / Coax connection to the relay contact lugs should be made using a length of flexible wire.
De-soldering braid makes such a connection. Refer to Figure 1.27. This relieves stress on the
lugs from stiff coax as well as providing acoustic isolation. The coil wiring should be made with
#22 or # 24 AWG, stranded, insulated wire.
Part
No Clean De-Solder Wick 5 x 0.080

3 Feb 2014

Digikey p/n
82-150-ND

22

Approx Cost
$3

VA7JW

AN SB-220 COMPENDIUM
APPENDIX I
QSK KEYER SCHEMATIC

3 Feb 2014

23

VA7JW

AN SB-220 COMPENDIUM
APPENDIX II
CURRENT SOURCE DETAIL
This QSK design has used 26V / 335 relay coils. Voltages and currents were measured and
are indicated in BLACK. If 12 V / 80 relay coils are used, which have not been done in this
implementation, a (calculated) resistor change is made from 4k3 to 11k to maintain ~ 5 mA
current. Calculated voltage readings are in RED. These may differ somewhat depending on
actual coil resistances affecting voltage drops in the totem pole chain. Other voltages / currents
that stay the same remain as BLACK.
The Current Source will not change output current if 12V relays are used, but Q4 will have a
much higher voltage drop across it given the lower voltage drop across the two 12V relays. This
increases the power dissipation in Q4 and a larger heat sink is required. Refer to the notes
below.

Circuit Operation
A 10 volt Zener diode is connected across a 10k pot which in turn is connected to base of
transistor Q4. There is a 100 ohm resistor in the emitter of Q4 across which is impressed a
constant voltage made up of the Zener voltage, as adjusted by the pot, plus the constant B-E
voltage of ~ 0.7V. A constant voltage across the 100 ohm resistor means that there has to be a
fixed current flowing through that resistor. That current is determined by the pot setting. Using a
DVM, adjust the voltage across the 100 ohm resistor to ~ 7.5 VDC for ~ 75 mA of current. An
additional 5mA of current from the Zener circuit plus ~ 1 mA from the pot circuit are added giving
a total current through the totem pole circuit of ~ 80 mA.. This is not a critical setting.

3 Feb 2014

24

VA7JW

AN SB-220 COMPENDIUM
APPENDIX III
CURRENT SOURCE / BIAS KEYER CONSTRUCTION DETAIL
The current source and bias keyer circuits are constructed on the same piece of holey board.
The board size is approximately 1-1/4 x 4-1/2. Component layout as shown and is not critical.

COMPONENT SIDE VIEW

CURRENT SOURCE
100 ohm

4.3k (two 2.2k)

BIAS KEYER
390 ohm 200 ohm
Opto-Isolator
51V zener
0.01 disk

Q4 with H.S
white wire

150 ohm

Current Adj
10k Pot

Q3

2 red wires violet wire green wire

Note that two 2.2k resistors are connected in series in this view as they were available
components at the time. They can be replaced with a single 4.3k 1/2W resistor
The QSK schematic has wire colors notated. The wire colors can be seen here and will aid in
assuring the connectivity. White wire to Relay 1. Red wire to +120VDC (in) & second red wire
(out) to SK-220. Green wire to SK-220. Violet wire to Relay 2.

Mounting bracket

WIRING SIDE VIEW

brown wire
green yellow wire

Mounting
Hole

CURRENT SOURCE

BIAS KEYER

Heavy Brown wire to 6V bias on RM-220 Rectifier board. Heavy Green-Yellow to T2 Filament
transformer.

3 Feb 2014

25

VA7JW

AN SB-220 COMPENDIUM
APPENDIX IV
KIT LISTINGS
This table is a summary of products currently offered by Harbach and AG6K. (as of Oct 2011)

Description

HARBACH ELECTRONICS

AG6K

Notes

P/N

USD

Stock Status

Part Number

USD

Stock
Status

Parasitic Suppression Kit

PS-220

$50

out of stock

3-500Z

$16.50

stock

Rectifier / Metering Board Kit

RM-220

$35

stock

1N5408

$0.22 ea

stock

Filter Capacitor Kit

FB-220

$130

stock

560 uF/450V

$11 each

stock

Electronic Keying Interface

SS-220

$33

stock

no kit

Current Inrush Limiter

SK-220

$27

stock

no kit

3 Feb 2014

26

check physical size. Equalizing Rs re'd

VA7JW

AN SB-220 COMPENDIUM
APPENDIX V
BILL OF MATERIALS for QSK SYSTEM
Designation
Q4

RLY 1, 2

Q2

Description
CURRENT SOURCE BOARD
Transistor, PNP 300V, 0.5A, 20 W
Diode, Zener, 10V, 1 watt. 5%
Resistor, 100 ohm 3 W 5%
Resistor, 4.3k 1/2 W, 5%
Trim Pot. 10k 1/2W multiturn
Heat Sink, Wakefield 290-2AB (see note)
VACUUM RELAY
Vacuum Relay SPDT Jennings
or Vacuum Relay SPDT Gigavac
or Vacuum Relay SPDT Kilovac
Diode, Rectifier, 1A, 600V
BIAS KEYER BOARD
Transistor, NPN 100V, 10A, 80 W
Opto Coupler, 80V, 50mA, Isolation 5kV
Zener Diode, Surge rated 51V / 5 kW
Capacitor, 0.01 uf 100V ceramic. 20%
Resistor, 150 ohn, 1 watt, 5%
Resistor, 200 ohm, 2 W, 5%
Resistor, 390 ohm, 2 W, 5%
SOFT KEY BOARD
Parts are as supplied in Harbach SK-220 kit
Values shown on schematic are for reference only
Values are for kit dated 1997
RECTIFIER and METERING BOARD
Parts are as supplied in Harbach RM-220 kit
Values shown on schematic are for reference only
Values are for kit dated 1997

Quantity

Type

1
1
1
1
1
1

MJE350
1N4740A

2
2
1
1
1
1
1
1
1

RJ-1A
GH3
HC-3
1N4005
TIP33C
PS2505-1
5KP54A

Source

Cost

DigiKey MJE350STU-ND
Digikey 1N4740A-ND
Digikey 43J100E-ND
Digikey CF12JT4K30CT-ND
Digikey CT94EZ103-ND
Dikikey 345-1021-ND

$0.60
$0.40
$0.80
$0.15
$1.43
$0.59

eBay

$40- $70

Digikey 1N4005GOS-ND

$0.35

DigiKey
Digikey
Digikey
Digikey
Digikey
Digikey
Digikey

$2.00
$0.64
$3.35
$0.32
$0.35
$0.56
$0.56

TIP33CGOS-ND
PS2505-1A-ND
5KP51ALFCT-ND
399-4262-ND
PPC150W-1CT-ND
PPC200W-2CT-ND
PPC390W-2CT-ND

Digikey price subject to change

3 Feb 2014

27

VA7JW

AN SB-220 COMPENDIUM
PART 2 PERFORMANCE MEASUREMENTS
2.1 INSTRUMENTATION
For a tutorial on Power, definitions and measurements have a look at page 35, Appendix VI. This
may be helpful in understanding the differences in power ratings for CW, PEP, RMS and
Average.
For an accurate measurement of output power, two measurements were taken to confirm each
other. Primarily, we depend on the inherent accuracy of a quality external wattmeter, and not the
rigs built-in wattmeter. Secondly, we measure the voltage across the 50 ohm dummy load and
then calculate the power to confirm the wattmeter measurement.
The voltage measurements were taken with an oscilloscope capable of accurate level
measurements at RF frequencies, i.e. 14 MHz. At a 1500 watt level, the voltage across a 50 ohm
load resistor would be 274 Volts RMS or ~ 390V peak. This is too high to apply directly to the
input of most scopes even using a 10X probe, as the maximum deflection on screen is limited to
200 V. Because of this, a precision attenuator is required between the dummy load and the scope
to reduce the voltage to a safe and accurate level. One solution to this is to use a high power, 50
ohm, lab quality attenuator. The attenuator presents a 50 ohm load and so acts as the dummy
load, and has the added feature of having an output that presents a calibrated output level, in this
case 30 dB (1000 X) less than the input.
Figure 2.1 below is the test setup for measurements presented in Sections 2.2 and 2.3
Amp
Output Power

Amp
Input Power

Icom
IC-7700

Watt
Meter

Power
Amplifier
SB-220

Watt
Meter

Total Attenuation Load to Scope 50 dB

Dummy Load
50 ohm / 2 kW
30 dB Attenuator

20 dB
50 Ohm
Attenuator

Vertical input
terminated 50 ohms

Tektronix
Scope
Model 465

Figure 2.1
Power Measurements Test Arrangement
The exciter is an Icom IC-7700 capable of delivering ~ 200 Watts of RF output.

Figure 2.2
Icom IC-7700
The amplifier under test is the SB-220 featured in this article.
The wattmeter is a Coaxial Dynamics Model 83000-A. This meter is much the same as a Bird
wattmeter using the same Bird slug except that it has the additional feature of being able to
measure Peak power as well as Average power. The meter display is much larger than the Bird
instrument which is nice. The instrument has a fast attack time and a longer decay time (seconds)
to sample PEP which makes it easy to observe RF peak power.

3 Feb 2014

28

VA7JW

AN SB-220 COMPENDIUM

Figure 2.3
Coaxial Dynamics Wattmeter
Output power from the exciter is measured in to the amplifier as a function of the RF POWER
control setting on the IC-7700. This ranges from a minimum of setting of ~ 5 watts to a maximum
of ~ 200 Watts. The dots on the control are assigned numbers, minimum (fully CCW) being 1
and maximum (fully CW) being 11. For example, the control is set to 6 in Figure 2.4

Figure 2.4
RF POWER control on the IC-7700
All power tests were conducted on the 20 meter band at 14.129 MHz. Once the Tuner on the IC7700 was matched to the SB-220 on the 20m band, it was found unnecessary to readjust the
tuner at the various power levels over the exciter range of 5 to 200W for these tests. Since the
input power levels were stable and repeatable at all output power levels, only one set of power
input levels were recorded for all tests and the wattmeter was moved permanently to the output
power position to measure amplifier output power vs RF POWER settings from 1 to 11.
The dummy load is a 50 ohm, BIRD Tenuline Coaxial 30 dB Attenuator, model 8329-300, serial #
4372. It is rated at 2kW continuous at 45 0C ambient. Thanks to Nick Massey, VA7NRM, of
Ionocom ( http://www.ionocom.com) in North Vancouver for the loan of this monster.

Figure 2.5
2 kW Dummy Load / 30 dB Attenuator

3 Feb 2014

29

VA7JW

AN SB-220 COMPENDIUM
The output of the attenuator is connected to another device, seen on the right, a Narda 20 dB / 20
watt, 50 ohm in-line attenuator that further reduces the high power transmit signal level to a level
suitable for input to the scope. The input to the scope is terminated with 50 ohms to preserve the
load impedance requirements of the attenuators. The voltage across this 50 ohm termination is
50 dB down from the voltage that appears at the 50 ohm input Z to the Bird Dummy Load /
attenuator.
The scope is a Tektronix Model 465 with a bandwidth exceeding 100 MHz so that the accuracy
of a voltage reading at 14 MHz is good.
All coaxs are 50 ohms using Belden FR-1 or Times LMR-400 product.
Formulas used to calculate Output Power as follows,
1.
2.
3.
4.
5.

Average Power, Pavg = (Vrms)2 / RL: also read as AVG on power meter.
Peak to Peak (pk to pk) voltage is read on the scope.
Vrms = [ Vpk-pk / (2 ) ] / 1.414
RL was as measured at 51. 4 ohms with a Fluke DVM.
The voltage out of the amplifier is attenuated by 50 dB. The load voltage is therefore 316
times greater (10 50/20) than measured on scope.

RMS Load Voltage = (Vpk-pk x 316) / 2.828

Load Power = (RMS Load Voltage) 2 / 51.4
Refer to Appendix VI for details on power measurement.

2.2 CARRIER RF POWER OUTPUT

This test measures the key down, steady state, maximum output level of carrier at the highest
power, SSB setting rather than the CW/Tune setting. N.B. The PA is NOT left running between
power settings. The SB-220 is not designed, or rated for, continuous Key Down operation.
Each reading represents a momentary operation where the reading is obtained.

(A)

(B)

(C)

TABLE 2.1
RF POWER OUT

3 Feb 2014

30

VA7JW

AN SB-220 COMPENDIUM
Table 2.1 (A) summarizes the Measured and Calculated CW average power output. Table 2.1
(B) lists the measured peak to peak voltage and the calculated RMS voltage across the load
resistor from which the calculated output Power is derived. Table 2.1 (C) shows the error noted
between the wattmeter reading and the computed wattage.

GRAPH 2.1
CARRIER SATURATION CURVE
Comment:
1. The amplifier will produce 1 kW of output with 100 Watts of input as is commonly assumed.
2. The (red) Linear Operation line shows where the amplifier departs the linear region and begins
to saturate at ~ 1400 W out when driven in excess of approximately 150 watts input.
3. Agreement between the wattmeter and scope is good at the higher power levels but degrades
at the lower power levels. This coincides with the changing of the slug from 250W to 2500W.
Calibration error in the 250 W slug perhaps.

2.3 PEP RF POWER OUTPUT

Peak Envelop Power (PEP) is the power measured at the peak of the RF sine wave.
It was thought that perhaps the RF peak power on SSB would be somewhat higher than the CW
peak power due to the lighter loading on the HV power supply since the average power is quite a
bit less on speech due to the nature of the peak to average levels. These tests indicate that the
peak to average is about 11 dB. Note that IC-7700 compression was ON and set up for ~ 6 dB
as measured by the IC-7700 meter function.
Using the same setup as the Carrier test, the peak voltage across the dummy load was measured
again using on the scope and using a consistent repeated speech pattern.
Capturing the peak voltage on the scope is a fleeting exercise since the Tektronix 465 is not a
storage scope, but the peak voltage results correlate well with the carrier voltages in Table 2.1 for
CW.
The wattmeter was used to measure both AVG and PEAK with the results per Table 2.2.

3 Feb 2014

31

VA7JW

AN SB-220 COMPENDIUM
Table 2.2 (A) summarizes the Calculated and Measured PEP. Table 2.2 (B) lists the measured
peak to peak voltage and the calculated RMS voltage across the load resistor from which the
calculated PEP was derived. Table 2.2 (C) shows the peak watts as measured on the wattmeter.

(A)

(B)

(C)

TABLE 2.2
PEP on SSB SETTING

GRAPH 2.2
PEP SATURATION CURVE
Comments:
1. The point at which voice peaks show amplifier saturation onset is about the same, at 1400
Watts. The peak carrier level is essentially the same for CW or SSB
2. Note that there is considerable difference between the calculated PEP based on measured
peak voltage and the wattmeter peak. The wattmeter spec states that PEP is accurate within 7%
of full scale. For a 2500 Watt slug, and running the maximum available power, ~ 1600 watts to
minimize error, this computes to a max error of 175 Watts. The discrepancy is considerably
greater than that. The reason for variance has not been determined but the greater confidence
resides with the measured voltage technique as equipment is verifiable, whereas the wattmeter,
on the Peak setting, is not. This is a probable calibration issue requiring further investigation.

3 Feb 2014

32

VA7JW

AN SB-220 COMPENDIUM
The following sections, 2.4 thru 2.6, were performed using an HP8562A spectrum analyzer and
an HP 9872C plotter in 1992 when the equipment was then available .
2.4 HARMONIC DISTORTION
This test measures the harmonic content of the carrier at 3.7 MHz. Choosing the lowest band
gives a wider view of the harmonic content throughout the higher HF bands. The amplifier was
driven hard, near saturation.

FIGURE 2.6
HARMONIC DISTORTION
Comment
There is no Heathkit specification for harmonic distortion products. The second harmonic at -43
dBc (dB with respect to carrier) is seen as quite acceptable given that the amplifier was driven
very hard.
2.5 TWO TONE TEST
Tests 2.5 and 2.6 are described in detail in the March 1992 QEX feature article Transmitter
Noise Loading by this author, who was then licenced as VE7AAL at that time.
Two audio tones of 1.2 kHz and 2.2 kHz at the same level were fed into the exciter. The amplifier
was loaded to about the 1.2 kW carrier level.
Intermodulation Distortion (IMD) products are measured in dB below the single tone level. With a
single tone, there are no IMD products to display as two tones are required to produce IMD.

3 Feb 2014

33

VA7JW

AN SB-220 COMPENDIUM

In this case, the two tones are at the zero reference level. The IMD products are easily read from
the graph but one must add 6 dB to reference these to the single tone = carrier level, i.e. the first
intermod product above the tones reads ~ 25 dB below the reference. With respect to carrier, the
IMD product is 25 + 6 = 31 dB down.

FIGURE 2.7
TWO TONE DISTORTION PERFORMANCE

Comment
The first IMD product, on the upper side, is about 28 dB below carrier. The Heathkit specification
calls for -30 dB so the amp is performing to spec.

2.6 NOISE LOADING TEST

This is a rather unique test in that white noise is loaded into the exciter MIC jack. This produces
a broadband signal more closely simulating voice. Obviously voice is not simulated by a steady
tone although the tone test is easy to do and provides specific levels that can be measured and
compared. However the noise test is informative as a simulation of actual service.

3 Feb 2014

34

VA7JW

AN SB-220 COMPENDIUM

FIGURE 2.8
NOISE LOADING INTERMODULATION

Comment
It is apparent that the sideband growth is much less that that seen in the two tone testing. At +/2.5 kHz to the side, the IMD products are down about 40 dB. This is likely attributable to the
modulating energy being spread out as a continuum rather than concentrated as two tones.

3 Feb 2014

35

VA7JW

AN SB-220 COMPENDIUM
APPENDIX VI
POWER TUTORIAL
UNDERSTANDING POWER
DC Power is easy to understand. Multiple the DC Voltage x DC Current and you have Power in
Watts.
AC Power is not so simple since the AC waveform varies with time. When the waveform crosses
zero, the power = 0, at its peak the power is highest, and in between it is something else, instant
by instant.
The equivalence between a DC voltage or current and an AC (sine wave) voltage or current is
called the effective value of the AC sine wave meaning that the AC waveform would generate
the same amount of power = heat as the DC value for a given resistive load.
The effective value is referred to as the RMS (Root Mean Square) value and is calculated simply
as,
(equation 1)

where Vp is the peak value of the sine wave. IRMS would be similarly calculated from Ip
The power developed by a sine wave is then given by
(equation 2)
where Pavg is the average or effective power.
Pavg may also be referred to as RMS power or mean power.
Fortunately we do not have to measure both RMS voltage and current. The measurement of
peak voltage along with knowledge of the value of resistance that the voltage is developed across
will give Pavg
(equation 3)
where R = 50 ohms

WATTMETER
The more common method of measuring RF power is to use an In-Line RF wattmeter such as the
Bird or Coaxial Dynamics products. The slugs or more correctly Plug-In Detector Elements
are calibrated to read Average Power.

3 Feb 2014

36

VA7JW

AN SB-220 COMPENDIUM

CW POWER
Figure VI.1 shows a CW signal, being an unmodulated RF carrier of unvarying amplitude
VP
0 to
PEAK RMS

PEAK
To
PEAK

Figure VI.1
Unmodulated Carrier
The RMS voltage of the carrier developed across a 50 ohm non-inductive resistor can be
measured by measuring the peak to peak voltage which is easy to do on a scope. The peak
voltage will be one half the peak to peak voltage and so equation 1 is used to calculate the RMS
voltage. Equation 3 will then give the average power.
MODULATED POWER
When the carrier is amplitude modulated as in AM or SSB, the RF sine wave varies in amplitude
and the above RMS calculations do not readily apply because the carrier power is varying with
time, that is, with the modulation. Since the modulation envelop being voice (SSB) and essentially
random, we cannot determine what power the modulation is developing.
Figure VI.2 illustrates this difficulty as one can see the modulation envelop is varying continuously
and variously with time.
PEAK = PEP

AVERAGE
0

Figure VI.2
Voice Modulated Carrier
We can avoid the RMS approach by introducing the Peak Envelop Power (PEP) power concept
where only the peak of the modulation envelope is measured, that is, Vpeak. Strictly speaking,
Peak Envelope Power (PEP) is defined as the average power supplied to a load by a transmitter
during one radio-frequency cycle at the crest of the modulation envelope taken under normal
operating conditions. Refer to http:www.ab4oj.con/test/peptest.html
PEP is instantaneous power and in the case of our transmitters / amplifiers, is a measure of the
peak power capability of the transmitter P.A.. This is actually of more interest and is not so hard
to measure. Again, measuring the peak to peak voltage as before, and using the peak voltage,
the peak power can be calculated at the voltage peak by simply using the familiar power formula,
Ppeak = (Vpeak )2 / R

3 Feb 2014

37

VA7JW

AN SB-220 COMPENDIUM
2.7 HEATHKIT SPECIFICATIONS

As per the authors Heathkit SSB Manual dated 1969

3 Feb 2014

38

VA7JW

AN SB-220 COMPENDIUM
2.8 A LITTLE HISTORY
According to WA7ZZE, Chuck Pensons book Heathkit A guide to Amateur Radio Products the
SB-220 was in production from 1970 to 1978. In 1972 alone, 1,341 units were sold. Pricing in
those days was $369.95 USD. To quote Penson, The SB-220 is probably the second most
popular amplifier on the planet. The most popular amateur Heathkit product was the famous
Cantenna, the HN-31. For details on this product, refer to Notes on the Heathkit HN-31
Cantenna also by this author.

17 November 2011 by John White VA7JW

Permission to use these ideas granted provided for personal non-profit use only
Reproduction granted provided acknowledgement given
Enquiries, comments, discussions, corrections, criticisms, disagreements as below,
[email protected]
604-936-2367

3 Feb 2014

39

VA7JW

AN SB-220 COMPENDIUM
2.8 Revision History, Errata, & Feedback Notes
19 June 2013
1. Error in Appendix V. QSK Zener diode was 60V, Should be 51V,
Digikey Part # 5KP51ALFCT-ND
2. Thermal question regarding 3-500Z pin temperature on Standby when fan is running at
low speed was raised. Would it exceed the rated 200degrees C due to lower air flow?
Tests indicated bulb temp on standby was 115 C max so it is OK

Bulb

FAN OPENING
Air FLow

Back of Chassis

Pin Temp

Filament 1R
Filament 2R

Filament 1L
Filament 2L

Bulb / pin Temp

Bulb

3. Feb 2014
There was an error in schematic of Appendix 1, page 23.
The common circuit (BN) of the AG6K Bias Keyer was shown to connect straight through to the
high voltage (117V) series string. This is wrong. There is not supposed to be a connection from
the HV Keying circuit to the Bias circuit as that is the purpose of the Opto-Isolator. The corrected
schematic is now shown. Apologies for the error.

3 Feb 2014

40

VA7JW