DOES THE K7DYY RIG ENFORCE 375 WATTS REGULATIONS IN THE MICROCONTROLLER DESIGN?
HOW TO SAFELY TEST DYY INTERNAL VOLTAGES
HOW DOES THE DYY PROTECTION WORK?

From QRZ forum discussion

    K5UJ said: ↑

        I guess that's the part of the rig that shut it down if the vswr went above 1.2:1 and didn't allow carrier power above the b.s. 375 watts. This is like something you'd expect from a JA rig. One reason I decided to not pursue one of those rigs was that I didn't like K7DYY's decision to be a self appointed power cop. I studied the data sheet for the FET used in the PA and concluded it was being run conservatively. Unless the microcontroller does the pulse duration modulation, I'd try defeating it and see what happens. Mike WZ5Q is pretty good at getting firmware out, modifying it and reflashing eproms. He could probably fix that box and remove the nanny state code. He's pretty smart.

W8KHK said:

    Rob and several others have made positive suggestions regarding opportunities to improve upon the K7DYY Super Senior line of transmitters. But there seem to be many assumptions about the PIC code, and why the product trips the way it does. There is also debate as to whether or not the rig will reliably provide the legal limit power output. I would like to share some technical details that may help to debunk the rumors, but in addition to try to set a path or direction toward assisting DYY owners until when (or if) someone or a group of folks is able to continue support and possibly continued production of the product line. Perhaps those who have first-hand experience and detailed knowledge of the internal design and operation could contribute collectively to an new thread, a sort of "survival guide" akin to the one that was developed over time for the Kenwood TS-530 and TS-830 series of hybrid SSB transceivers.

    First and foremost, I believe it is necessary, for safety's sake, to state that there are hidden risks to the uninformed, with regard to the so-called "floating ground", or "hot chassis" circuit design. This is in no way considered a shortcoming, rather, it is an elegant design that eliminates the need for a power, or isolation transformer at the line input. The advantages are lower cost, lighter weight, smaller package, and more efficient operation. The only downside is the need to understand the circuit design, such that tests, measurements, repairs, and adjustments may be performed in a safe manner. Much can be gleaned by studying all the sections of the schematic diagram, prior to opening the unit and diving inside.

    The design of the trip "feature" was in no way an attempt to enforce the revised power limit. While the traditional power limit was 1000 watts DC input to the final amplifier that feeds the antenna, requiring the metering of both plate voltage and plate current if the power is above 900 watts; the current regulation is simply a maximum of 1500 watts peak envelope power to the feedline. While the amateur is prohibited from exceeding this limit, the regulations do not state any required method of measuring the transmitted power on an ongoing basis. The commonly quoted algorithm for AM carrier power output of 375 watts is based upon the fact that a carrier, modulated to 100 percent with a steady sine wave, the modulated carrier including sidebands reaches a peak envelope power of four times the carrier power. The converse calculation assumes that the legal limit is 1/4 of the legal PEP, thus 1500 / 4 = 375 watts carrier. In contrast to a sine wave modulating signal, all bets are off when attempting to apply this "rule" to the dynamics of speech modulation! Further, there is nothing within the K7DYY transmitters that determines or attempts to "enforce" either maximum carrier power or peak envelope power. Instead, here is a brief summary of the trip functionality:

    The class-D transmitter relies upon broadband filter networks to achieve adequate spectral purity and ensure all spurious products are within legal limits, especially when working close to band edges. This is implemented with fixed, not tunable components, as is customary with most products of the day. Within the output section of the filter network, a reverse power detector with passive components consisting of a diode, inductor, resistors and capacitors measure the reverse power, and if it reaches a danger point, a transistor pair functioning as a bistable flip flop, is locked into the overload state, and sends a binary signal to the PIC, initiating an orderly shutdown. The overload state is reset by removing, then restoring, the primary AC power input. The sole purpose of the measurement of reverse power is to limit the peak voltage across the RF output power FETs to a safe value. There is no intent to measure forward power or enforce compliance with either a carrier or PEP output level.

    If the transmitter is connected to a 50 ohm non-inductive dummy load, it may be modulated 100 percent with a sine wave, and it will produce 375 watts carrier or 1500 watts PEP into the load without tripping off. However normal voice communication is not a sine wave, and a typical amateur radio antenna is not going to look like 50+j0 to the DYY output circuit. As we all have observed, there are several things that cause the reflected power detector to cause a shutdown trip. One common cause is negative overmodulation, which we all know causes distortion and splatter. This situation results in a large amount of reflected power, even with a properly matched RF load. The prevention methods may include tight compression, rock-solid peak limiting, or, in the absence of these functions in the audio chain, a reduction in modulation level. But to maintain 100 percent modulation without trips and the above audio chain features, the carrier power is reduced to 200 or 250 watts, such that when the occasional negative overmodulation peak occurs, the trip is avoided and the transmission continues.

    We all know that an antenna must be matched for the desired operating frequency in order to present a 50+j0 load to the transmitter. If a high-Q sharply tuned antenna system is used, straying away from the resonant frequency increases the reflected power, and if the level of reflected power reaches the threshold, the rig trips off. The detection circuit includes a trim pot to allow the specific reflected power trip point to be adjusted. Again, this is set at the factory to protect the FETs from overvoltage spikes, and is unrelated to forward power or carrier power level. the DYY rig works equally well with coax-fed or balanced line feed. Of course the latter requires a tuner, and the former benefits from a tuner. With a coax-fed dipole, the range if frequency above and below the design point is limited by the reflected power, even though the losses incurred in the coax due to higher SWR are tolerable. The Q factor of the tuner, if used, is also a contributing factor. If a low Q tuner is employed, it presents a reasonable match slightly above and below the carrier frequency. This allows moderate QSY capability, and reasonable sideband breadth, in other words, occupied bandwidth. A higher-Q tuner presents a perfect match only at the carrier frequency, and the match progressively deteriorates as one moves away from the carrier frequency, risking an overload trip. Because the excursion of the sideband frequencies with wide bandwidth modulation are a larger percentage of the carrier frequency on 160 meters, it is possible to trip the overload circuit with 6 KHz audio response (12 KHz occupied bandwidth) on this band, if the tuner Q is very high. In this case the sideband energy may cause a high enough reflected power to be detected, and the rig is shut down.

    Is the PIC in any way involved in this decision process? ABSOLUTELY NOT. The PIC simply responds to a binary input on one pin, which trips only when the reflected power reaches the threshold set by the trim pot, causing the two transistor flip flop to lock into the overload state, and this binary signal commands the PIC to set the overload status. This status switches back to receive mode, but the flip flop in the detection circuit latches until primary power is removed.

    Suggestions to improve the "operator friendliness" of this behavior might include an indicator LED, informing the operator that the transmitter has entered overload status.. Another option would be an output to an audible alert, informing the opeRator of the shutdown. Another convenience option would be a reset button, allowing the flip-flop transistors to be reset, without requiring a power down, power up sequence. These can all be accomplished with NO ALTERATION TO THE PIC CODE!

    It might also be helpful to share some info regarding the PIC. A third party assisted Bruce in the development and support of the PIC code. With all the other tasks to produce and support the product, Bruce did not have the bandwidth to manage that portion of the development. The PIC is primarily responsible for monitoring the front panel controls, managing the alphanumeric display, setting a code word in the DDS VFO chip to generate the desired carrier frequency, switching between bands (relays are used to alter the bandwidth filter components) switching between transmit and receive mode, and honoring the order to shutdown when overload is detected. Considering the generic functions of the PIC, instead of attempting to extract the code and reverse engineer the PIC, it would be rather trivial to write new code, from scratch, to manage the user interface, display, and the DDS generator.

    The PIC has absolutely no involvement in either deciding when to shut down, and it is not involved in the audio or modulation logic. The modulation is standard PWM, and it is implemented with a standard IC commonly used in switching power supplies. Instead of the varying output voltage in a regulator application, the audio is impressed upon the feedback input to the switching power supply chip, with the result being a PWM signal, which is then amplified by transistors (either bipolar or FET) then filtered to remove the PWM square wave, resulting in the modulated DC which is passed to the RF final amplifier stages. In this regard, the PWM modulation is very similar to the implementation used on all the Class-E transmitters designed by WA1QIX, and more detail on their operation may be gleaned by studying the documents at http://www.classeradio.com/ (Some of his rigs use the alternative class-H modulator, but most are PDM-based).

    In summary, the DYY transmitters do not attempt to enforce allowable power regulations, and they can operate in a manner to produce legal limit power when modulation input and RF output match are adequately controlled. While there is always room for improvement in any product, Bruce has produced an exemplary range of equipment that is appreciated far and wide. If a perfect antenna match cannot be accomplished in a practical manner, carrier power reduction is an excellent alternative, and the small fraction of an S-unit reduction will never be noticed at the receiving end.

    For more helpful information, please consider visiting the "Wireless Girl" site, that of Janis Carson, AB2RA at https://www.wirelessgirl.net/ She has worked extensively with Bruce, from the DYY product's inception, and has helped numerous amateurs attain maximum satisfaction and pleasure while operating their Bruce Franklin rigs.