Tech Talks and Tips by K4KYV

(Cont'd from previous page)

Oldham Coupler
​

...is the name of the little flexible shaft coupler found in the Collins 75A- series receivers, R-390 series and other equipment. The coupler, a mechanical device for transmitting torque between two shafts that are not perfectly collinear, was invented in 1821 in Ireland by an engineer named John Oldham, to solve a problem in a paddle steamer design.

Animation of the coupler in action

 

75A-4 Mechanical Filter Shunt-Feed: Precaution

The original circuit in the 75A-4 receiver has the full B+ flowing through the input coils of the mechanical filters
to the plate of the 6BA7 2nd mixer. If the plate of the tube ever shorts internally, the wiring to the plate circuit
shorts to ground, or as in at least one reported case the (now 60 y.o.) insulation inside the filter apparently had
deteriorated with age causing the coil to short to the case, the full 200 volts B+ from the power supply will appear
directly across the input terminals of the filter, resulting in its instant destruction.

In later production runs of the receiver, the circuit is changed to shunt-feed, using an RF choke as a plate choke
to carry the B+ to the tube, with B+ voltage isolated from the filters with a 1000 pf disc ceramic blocking capacitor,
C144. A 62pf mica capacitor is wired in parallel with the rf choke to broadly resonate at 455 kc/s. The cold ends of the
mechanical filter input coils in the revised circuit are grounded directly. This upgrade was designed to prevent
damage to the filters in the event of any one of the above-described failures. See the attached schematic.

Collins regularly published "service bulletins" to the receiver, incorporating the latest production changes,
and often a set of components to retrofit the service updates could be purchased as a kit directly from Collins.
This production change appeared on the schematics in later users manuals, but was never the subject of a
service bulletin nor were the components offered as a kit. However, descriptions of the upgrade quickly
circulated within the amateur community, and many 75A-4 owners purchased the components and
incorporated the change in their receivers.

Click on the attachment to compare schematics of the original Collins circuit (left) and the circuit after
the production change (right). L32 is the mixer plate choke, and C145 is the resonating capacitor.
C144 is the plate blocking capacitor.

The original Collins production change suffers a fatal flaw: a shorted blocking capacitor (C 144)
would produce exactly the same disastrous result that the upgrade was designed to prevent. The capacitor
used in the later production is a nondescript .001 mfd 500v disc ceramic, and more than once I have seen
similar capacitors in various equipment fail by developing a dead short. In my 75A-4 I replaced the original
blocking cap with a .001 mfd 3 kV disc ceramic, after first testing it at 2 kV for leakage or short circuit,
using a hi-pot tester. The new disc ceramic is the same diameter as the old one, but several times the thickness.
The likelihood of a 3 kv capacitor that tested good at 2 kv, developing a short at only 200v DC should be small,
although not beyond the realm of possibility. A more secure approach would be to wire two .002 mfd 3 kV
disc ceramics in series, since simultaneous failure of two separate capacitors operating at less than 10%
their nominal working voltage would seem unlikely.

In summary, if your 75A-4 does not already have the shunt-feed upgrade, this modification should be installed
sooner rather than later. If it does have the revised circuit, replace the original .001 mfd disc ceramic capacitor
(C144) with another one rated at 2 or 3 KV nominal working voltage. Better still, replace C144 with two .002 mfd,
2-3 kV disc ceramic capacitors connected in series.

Note:
The 62 pf mica capacitor and 2 mHy RF choke used by Collins in the upgrade may be difficult to obtain.
Readily available standard component values, a 2.2 mHy (2200 μHy) RF choke and 56 pf dipped mica capacitor,
should work OK. A small terminal strip is needed for mounting the parts.

 

Four-sided vs Triangular Towers

You may have noticed that many older towers dating back before WWII are four sided, while most newer towers constructed postwar are triangular. Classic examples of four-sided construction are the famous diamond-shaped Blaw-Knox towers at WSM, WLW, WFEA in Manchester NH and others. Photos of towers dating back to the 20s and earlier in old magazines and textbooks nearly always show four-sided construction. Here is the Rohn tower company's take on three vs four sided tower construction.


Square Angle vs. Triangular

Following are some points to consider when specifying towers. As you will see, the square angle design has no advantage whatsoever over the triangular design, and the triangular design has many advantages over the square design. This is evidenced by the fact that the majority of the world's major tower/mast manufacturers operate in an environment where the finest quality raw materials are available, and are not limited to a certain steel supply. These modern manufacturers provide triangular towers as their primary product.

1. Square tower design is old fashioned and antiquated. Some designs date back to the 1800s and were based on availability of material, existing design capabilities and ease of manufacture. Almost any steel fabricator can design and manufacture square angle towers or masts. Square angle towers are specified in many cases because of the long standing designs that have not been updated over the years. These structures have been the accepted standard, since many departments and users do not have the in-house structural engineering capacity to evaluate modern designs, and take the time to update their requirements.

2. Square tower design is very popular in developing countries where updated technology and sophisticated machinery are not available yet. It exists only because there is no other choice, and what was good enough 100 years ago, is all that is available today.

3. Towers and masts constructed from angle material have a much higher wind load than the more sophisticated triangular round member tower. A triangular round member tower is much more aerodynamic and therefore has lower wind resistance.

4. Because of the higher wind load on the structural members, more reinforcing pieces are necessary, and therefore the structure when completed has many more components and connections than a triangular tower.

5. A square tower with all of this extra material, is no stronger than a triangular tower designed for a similar load. There are international standards developed for tower design such as ANSI/TIA/EIA-222-F-1996 that govern proper tower/mast design for the communications industry.

6. As a result of the square angle design, there is more labor involved to assemble the material, more possibilities of pieces not fitting, more connections to become loose and require maintenance.

7. Triangular towers however are lighter in weight thus saving freight costs, and are constructed of fewer pieces. This is possible because of the higher strength steels that are currently available for the more high-tech tower/mast designs.

8. Triangular towers only require 3 foundations, square towers require 4. There are considerable cost savings in civil works and concrete using a triangular design.

9. From the standpoint of deflection and twist, the triangular pipe tower is stronger and more rigid pound for pound.

10. With round main members (legs) equipment such as dish mounts, platforms etc. are mounted with 'U' bolts, and therefore can be moved from location to location without drilling additional holes in the structural members of the tower. Antenna mounts for example can be added to the structure without any field punching, drilling or welding.

11. There is an old fashioned argument that pipe members corrode from the inside, and since the corrosion is hidden, it cannot be maintained or corrected. Back when pipe members were first used in construction, the material was not hot dip galvanized inside after fabrication. With today's modern fabrication procedures and galvanizing technologies, this condition does not exist. Back to back angle members can also corrode from the inside, and cannot be maintained. The secret is in the fabrication/galvanizing procedure. See more in FAQ.

12. Due to the availability of larger sizes of higher strength round structural steel shapes, round member pipe/solid towers can be designed with single piece main structural members. Angle towers require 'back to back' bolted or welded members ("built up" sections) to provide the strength required for some of today's tremendous antenna loads and tower heights.

http://m.rohnproducts.com/technical-info/square-angle-vs-triangular.html

 

Better use of the envelope pattern on an oscilloscope to monitor modulation.

One of the limitations to using a typical ham radio quality "monitor scope", such as those made by Heathkit, Kenwood and Yaesu, is that the image is squeezed into the screen of a tiny 3" CRT, resulting in mediocre resolution at best. This makes it particularly difficult to clearly see negative peaks breaking baseline, or flat-topping of positive peaks.

This can be greatly improved with any oscilloscope, since the envelope pattern of both AM and SSB signals is symmetrical; 100% of the information about the shape of the modulation envelope and percentage of modulation is redundantly displayed both above and below the base line. Simply adjust the vertical position so that the base line appears near the bottom of the screen, but far enough away that it remains clearly visible. Now, increase the vertical gain until the crests of positive peaks reach close to, but not quite at, the top of the screen. The image can then be proportionally stretched horizontally by increasing the sweep rate. Expanding the symmetrical display so that only the top half (above baseline) fills the screen allows a more distinct view of positive and negative peaks without losing any of the information about the envelope of the signal, since this produces a 2X magnification over the conventional setting.

To see this, as photographed from oscilloscope patterns, go to:

https://www.thebdr.net/measuring-am-modulation/

Scroll down to "Using the Oscilloscope") and then to "Higher Resolution". This appears a little more than halfway down the page.

 

Bandwidth, Occupied vs Necessary

Transmitted "bandwidth" falls under two distinctly exclusive technical definitions, Occupied and Necessary.

Occupied Bandwidth means the actual measurable-with-instruments bandwidth of a given transmitted signal.

Necessary Bandwidth means the absolute minimum bandwidth that would just barely be sufficient to transmit the content of a signal at the desired information rate and with the desired quality under specified conditions.

How do these two definitions apply to amateur transmission, and why is this distinction important, particularly whenever the issue of regulation-by-bandwidth is periodically raised? As you may remember, the infamous FCC Docket 20777 from the mid 1970s, which proposed to outlaw AM on all amateur bands from 160m through 15m, would have imposed strict limits to occupied bandwidths of signals transmitted within the various mode sub-bands. Below 28 MHz, transmissions in the "CW bands" would have been limited to 350 Hz, and in the "phone" bands to 3.5 kHz, under specifically defined strict standards enforceable by the FCC.

If some form "Regulation-By-Bandwidth" were to eventually replace modes-of-emission to define the amateur subbands, as long as it remained based on necessary bandwidth, this would not in itself impose any specific limitation to transmitted bandwidth. Necessary bandwidth would simply replace the present-day list of emission designators to define band segments, which must be defined somehow. For example, if the phone segments were to be re-defined in terms of 6 kHz necessary bandwidth, that definition would accommodate AM, since it is generally understood that a minimum of 6 kHz is required for intelligible AM voice transmission. But that would not preclude a signal whose audio frequency response fails to roll off precisely at 3 kHZ; it would still be perfectly legal to transmit a signal whose modulation rolls off at 5-plus kHz, thus generating an occupied bandwidth of 10 kHz or greater. "6 kHz necessary bandwidth" would merely be an alternative accommodation to AM phone, but this would also include any other mode that happens to use the same necessary bandwidth, such as narrow-band FM. It would also allow any other yet-unspecified mode with a necessary bandwidth of 6 kHz to operate without having to amend the rules to allow this type of emission. Technically speaking, to allow SSB transmission in that same band segment, a separate designation of "3 kHz necessary bandwidth" might have to be included, since 6 kHz of bandwidth is not the absolute minimum necessary for that mode. In the CW bands, tone-modulated CW would once again become legal, as long as the necessary bandwidth was not greater than whatever is specified for those segments. In the phone bands, digital hash and RTTY would become legal as long is it met the standards of necessary bandwidth.

In summary, "necessary bandwidth" would be a way to designate modes assigned to various band segments without having to specifically list every permitted type of emission. This would have no bearing on the actual occupied bandwidth of any signal, other than the current standard of "good engineering and amateur practice". Any enumerated limits to occupied bandwidth would have to be specified separately in the rules. Up to now, with only a few special exceptions, the FCC has purposely declined to specify any numerical limits to occupied bandwidth, beyond "good engineering and amateur practice". This allows amateurs the maximum flexibility for experimentation and self-instruction in the radio art, and relieves the FCC of the burden of monitoring and enforcement of signal bandwidth. Hams don't have to worry about receiving a citation because their signal bandwidth happens to exceed some arbitrary standard, or possess an expensive spectrum analyser to make sure they are in compliance. In other words, we are expected to use common sense. Contrary to a piece of misinformation long circulated within the amateur community, there is no FCC regulation that limits AM to 6 kHz bandwidth.

See the definitions below:

Occupied bandwidth: The width of a frequency band such that, below the lower and above the upper frequency limits, the mean powers emitted are each equal to a specified percentage B /2 of the total mean power of a given emission. Unless otherwise specified by the CCIR for the appropriate class of emission, the value of B /2 should be taken as 0.5%. [NTIA] [RR] (188) Note 1: The percentage of the total power outside the occupied bandwidth is represented by B . Note 2: In some cases, e.g. , multichannel frequency-division multiplexing systems, use of the 0.5% limits may lead to certain difficulties in the practical application of the definition of occupied and necessary bandwidth; in such cases, a different percentage may prove useful.

https://www.its.bldrdoc.gov/fs-1037/dir-025/_3627.htm

Necessary bandwidth: For a given class of emission, the width of the frequency band which is just sufficient to ensure the transmission of information at the rate and with the quality required under specified conditions. [NTIA] [RR] (188) Note: Emissions useful for the adequate functioning of the receiving equipment, e.g., the emission corresponding to the carrier of reduced carrier systems, must be included in the necessary bandwidth. (188) (See Annex J of NTIA Manual of Regulations and Procedures for Federal Radio Frequency Management for formulas used to calculate necessary bandwidth.)

https://www.its.bldrdoc.gov/fs-1037/dir-024/_3495.htm

 

Distortion with a Common Modulator and RF final Plate Supply

One often overlooked problem may occur when a single HV power supply is used
for the class-B modulator and RF final. This was especially true back in the days
when HV filter capacitors were expensive and most rigs got by with only a couple
of mfd filtering.

Each class-B modulator tube draws its maximum plate current at the peak of the
half-cycle when it conducts, thus the modulator draws peak current from the supply
twice during each audio cycle as the two tubes take turns drawing plate current.

Thus the current flow to the modulator is predominantly at the 2nd harmonic of the
the audio signal, and this fluctuating load may modulate the DC power supply voltage
enough to be audible in the modulated signal if there is insufficient isolation between
the class-B modulator and the RF PA. This may be the source of seemingly inexplicable
distortion in the modulated signal.

This can be corrected several ways. If a modulation reactor is used, the reactor tends
to filter out that 2nd harmonic audio component before it can reach the PA. If a
two-section filter is used, the class-B modulator may be run off the 1st L-C section,
and the PA off the 2nd. This isn't an issue if a large enough filter capacitor is used
at the output of the +HV supply. Some broadcast transmitters use only a few mfd,
for example 8 mfd in the Gates BC-1 series; this is sufficient since a modulation
reactor is used.

In order for the modulation reactor to effectively filter this or any other incidental noise
riding in on the high voltage supplied to the final, the bottom end of the modulation
transformer secondary must be returned directly to ground through the blocking
capacitor, not returned to the +HV from the power supply.

 

Bandpass vs Passband

To clear up a frequent point of confusion, the term“Bandpass” (adjective) describes a type of filtering process. A Bandpass Filter is a device that passes frequencies within a certain range and rejects or attenuates frequencies outside that range.

“Passband” (noun) is the term that describes a range, or "band" of frequencies that can pass through a Bandpass filter without being attenuated. It refers to the actual portion of affected spectrum.

For example, the standard AM mechanical filter in the 75A-4 is a type of bandpass filter. Its nominal passband is 6.0 kHz.

http://rfcec.com/RFCEC/Section-3 - Fundamentals of RF Communication-Electronics/33 - TRANSCEIVER/Transceiver - Bandpass and Passband (By Larry E. Gugle K4RFE).pdf


For complete table of contents of Tech Talks, go back to page 1, message #2