SB-104 & SB104A

[Kechkaylo, David]

Dave:

I'm intrigued by some of the responses you've received regarding the
SB-104 and the SB-104A, especially from Keith Rowland, and having been a
factory service tech there in the amateur radio group at Heath Co who
worked almost exclusively on the both the '104 & the 104As, I do take
several exceptions to his comments.

First of all, for Paul K. Dean's comment: The HW-2036 was a cleaned-up
version of the HW-2026 synthesized 2-meter rig. It is true that the
2026s were recalled, and this was due to the high level of output
spurious being generated in the multiplier stages following the VCO.
There never was a planned recall of the 2036. In fact, the 2036A allowed
one to cover the ENTIRE 2-meter ham band, not just a slice as in the
2036, plus a couple of other mods. If you look at the VF-7401, you'll
see that the receiver and transmitter sections are still very much the
2036 design, excepting that many shields around several of the coils
were added, presumably to reduce interstage & external coupling. The
VF-7401 took a very short time to get it to market, since there were
many similarities to the 2036. It is a nice radio, and the
synthesizer/display circuitry was a good design (hat's off to Terry
Perdue).

Anyway, Keith sounds very much like the typical frustrated SB-104 owner
in the early days of that radio. In many instances, and I'm not saying
this was Keith's fault here, but many owners of the radio were careless
in their construction & operation of the SB-104, some almost abusive.
Given the complexity of the '104, it became tiring to construct it, and
with this, many of the problems I've discovered with the '104 WERE
CONSTRUCTION ERRORS, period, and believe you me, I worked on a lot of
them before I left Heath and finished my MSEE degree in 1984.

Given all this, this is a short list of some of the pitfalls of the '104
(and the 104A):

1).  A large majority of the '104 problems were receive sensitivity. The
original circuit design was grafted from the SB-303 front-end, but added
the banded (diode-switched) bandpass filters. Many of the problems on
the original "G" board were due to poor gain distribution and the
inability to align the 8.65 MHz 1st IF BPF correctly. There were several
modifications to this board (including the SBM-104-1 modkit) to increase
sensitivity. But although it could be said that it was "OK" for 80 & 40
meters, the 10 meter band certainly needed help. So in 1976, this board
was again redesigned. The orginal '104 front end boards were:
#85-1459-1, then #85-1666-1, then #85-1708-2. The "A" updated radio was
#85-1931-1. This redesigned board worked much better, but due to the
discrete design of the balanced mixers and the way the 1st IF BPF had to
be aligned, it was never offered as a kit form; instead, it was always
furnished as a factory assembled & aligned  board.

2).  The PA board had a few problems in that improper mounting of the
studded transistors there would actually crack the thermal bond of the
stud to the die base when the board was assembled onto the aluminum
plate. "Torque" down these four nuts too tight when the stud bases were
not flush with this plate caused them to crack. Therefore, these finals
would "go out to lunch" when thermally stressed during transmit,
especially during a brief high VSWR. The ALC cutoff time was too long,
and yes, these devices couldn't handle the stress. This was the problem
in the early HF transistors available at the time the '104 was designed.
The PA board was designed from the TRW application note No. CT-113-71.

3).  The DC-to-DC coverter was situated right behind the ALC/Lowpass
filter circuit board. This is where the weak signals from the antenna
passed, and this is also where this switch-mode power converter sat,
spewing harmonics of 25KHz up the band. This is one of the famous birdie
tweets that you here slowly creep up the band and then suddenly quit.
The "A" version tempted to reduce this by an additional shield between
the 180-volt power converter and the ALC board.

4).  The transmit IF "C" board, aside from the front end sensitivity
problems, was probably one of the biggest problems in the '104 as well.
The diode-switched bandpass filters after the predriver amps were
particulary troublesome in that the tolerance of these filters didn't
allow full band coverage;i.e., power would roll down from one end of the
band to the other, so there were numerous update mods for these as well.
The engineer in charge at the time I think wanted to re-design this
board, like that of the "G" board, but instead, it was given a face lift
and eventually factory built and aligned for the "A" version of this
radio. It was never re-designed completely, and many of the close-in
spurious that you see in the output spectrum is due to the somewhat
dirty on-channel mix from the discrete diode-ring mixer there on the "C"
board. The following diode-switched BPFs cannot remove those in-band
spurious products. Better control over the RF & LO levels with a
diode-matched mixer would have helped.

5).  The VFO had several problems due to FM-ing of the 5-5.5 MHz signal
when the unit was to be used in mobile applications. Orginally, the VFO
supply was regulated by the reverse-biased E-B junction of a 417-118
transistor, then was replaced by a 10V zener, then later to a 9.1V
zener. Finally, after enough FM-ing problems persisted, it was changed
to  7.5V along with a few other minor mods on both the VFO OSC board.
The  VFO BUF board was updated to increase VFO level, too. There was, in
only a few cases, some excessive drift, but these TC caps were updated
again in the "A" version.

6).  The original HP-1144 power supply actually worked quite well, and
did not suffer any RF rectfication problems as Keith mentioned. It
wasn't until Heath added the quick shut down circuit to blow the 20A
fuse should the output pass transistors short. It was then that the RF
rectification problems in the HP-1144A began. Liberal amounts of RF
bypass caps were added, but sometimes the problem persisted.

7).  The jittery display was a problem in the old 104 especially, and
was due to a too-low premeix level caused by a too-low level VFO signal.
The old 85-1516 VFO filter board injected too high a VFO level on the 10
meter band on purpose, but was found later to cause too high level of
spurious on 10 meters, and therefore was concluded it wasn't needed.
This bypassing circuit that increased the VFO injection level on 10
meters was later removed. The setting of the VFO level was somewhat
critical. In the '104, it was increased higher to help with the receive
sensitivity, but birdies resulted and transmit spurious increased. A too
low level helped with the spurious, but caused the display to have the
jitters ( the counter missed several pulses to count, as the level was
too low). The VFO filter board was completely re-designed in the "A"
version.

8).  There were other problems such as the keying waveform, ceramic
capacitors which were under rated in the "B" board and which had too
high leakage current, especially in the VOX circuit. Talk back was also
a problem in which one could hear youself in the station speaker due to
feedback on the ALC and 3395 IF circuits in transmit mode (this was a
bigger problem actually in the "A" version!). Additionally, there were
(and still are) noted instabilities that occur when switching between
transmit and receive, and visa versa, but the addition of several
capacitors varied the switching times enough to reduce this problem.

The "A" version was an improvement, but was only actually a re-design of
I'd say about 15-20% of the radio. The "A" version SB-104 was never
totally a redesign of the older '104.

Some of the "A" version updates actually made thing worse, and most
technicians would install just enough updates to make the SB-104 perform
well. And in actuality, the SB-104 worked quite well once all the
regular '104 updates were installed. The "A" version offered the 400 Hz
CW filter switchable while in CW mode and had improved sensitivity, but
really, other than this, both the SB-104 & the SB-104A worked quite well
otherwise. Spurious output levels were also reduced in the "A" version
design.

The worst problems, and I mean this sincerely, was when a customer
installed the SBM-104-2 update kit to upgrade his/her radio from the
'104 to the '104A. This was almost always a disaster to work on. These
upgraded radios never really worked as well as just the plain old (but
updated '104 excepting the receive sensitivity) or an original '104A
from the get go. The customer would always botch up the printed circuit
foil and burn wiring insulation together, etc., just to install the
upgrade kit. They were always touchy after that. This is not to say ALL
upgraded '104s to the "A" version were done poorly, but then the factory
service only saw the "botched" ones anyway in my opinion.

Most of the updates and circuit improvments came either from the guys in
tech service (factory service and field Heath Electronic Centers) and
from engineering. They appeared in bulletins made available to us in
factory service, and these improvments were incorporated into the next
lot or run of these kits as soon as possible. Heath did I think a pretty
good job of keeping customers informed of circuit improvments, and in
many cases I know, bent over backwards to help customers out. I always
added updated components and improvments to the SB-104s for customers,
and never, ever charged them.

The HW-104 series faired better performance-wise, mainly because by the
time the HW series made its debut a year after the SB-104, many of the
circuit improvments were known and incorporated at the time. The only
help the HW had over that of the SB counterpart, was the omission of the
5 to 180 volt switching power converter, thus less birdies!

There were, admittedly many more circuit changes and updates on the
SB-104 that I haven't mentioned, but the fact of the matter is that, if
built carefully, the SB-104 was a pretty good radio. Most of the
frustrations, tho were really from careless construction and soldering,
and supplied defective or out of tolerance components. There certainly
were design-related pitfalls, and my opinion was that the SB-104 was
certainly pushed to production for the 1974 Christmas buying season too
quickly; many of the problems discovered in the manual and proofbuild
stages were not caught or were not known about in time. This was
certainly true of the receiver front end board. Part of this board was
redesigned again between the proofbuild & production stage (between July
& August of 1974) without enough time for evaluation.

Anyway, this is my 2 cents worth on the SB-104. Enjoy...I think the
designers and technicians in engineering at Heath of the SB-104 (whose
names I won't mention without their permission) did a pretty good job
given the design goals of such a complex radio in kit form.

Regards,

David Kechkaylo

--- --- --- --- --- --- --- --- --- --- --- --- --- --
To subscribe: listserv at listserv.tempe.gov
and in body: subscribe HEATH yourfirstname yourlastname
To unsubscribe:  listserv at listserv.tempe.gov
and in body: signoff HEATH
Archives for HEATH: http://www.tempe.gov/archives
--- --- --- --- --- --- --- --- --- --- --- --- --- --