Saturday, February 11, 2017

A novel APD-based speech bandwidth optical receiver

In a previous posting I wrote about a novel application of a JFET (Read about that in the article "Gate current in a JFET - The development of a very sensitive, speech-frequency optical receiver" - link) in which the flow of gate current was integral to the operation of a photodiode-based optical detector.  In testing this circuit, which included an indoor "photon range" and out in the field, it was observed that the sensitivity of this circuit was, at "audio" frequencies, on the order of 8-20 dB better in terms of signal/noise ratio than any of the more conventional "TIA" (TransImpedance Amplifier - read about that circuit here - link) circuits that had been tried.

In the analysis of this circuit it was determined that several factors contribute to the ultimate sensitivity, including:
  • The intrinsic noise of the JFET.  This can be minimized by hand-selection of the device itself for the lowest-possible noise as well as selecting a device that can operate at a higher drain current to reduce the "bulk noise" - or even the use of several JFETs in parallel.
  • The contribution of noise by other circuitry.  In the design this was minimized through the use of a cascode circuit topology as well as the use of a low noise, high impedance current source to supply the bulk of the drain current and the complete avoidance of other components being connected to the photodiode-JFET circuit junction.
  • The capacitance of parasitic circuit elements, such as capacitance (including the Miller effect) that reduces the amplitude of the signals from the photodiode, particularly as the frequency increases, effectively reducing the signal-noise ratio.
  • The contribution of the photodiode itself.
Of these factors, the majority of the noise would appear to be due to the JFET itself, particularly above the low audio frequencies frequencies (e.g. below 100Hz or so) where 1/F noise would dominate. One of the possible approaches to get better noise performance is to cool the circuitry, but this is fraught with difficulties related to condensation which would require that the device itself be sealed in an atmosphere (e.g. dry nitrogen) in a manner similar to that used to cool CCD imagers for astronomy.
Figure 1:
The outside view of the completed APD-based optical receiver.  Because
of its extreme sensitivity it must be well shielded to minimize the pick-up
of stray fields such as those from AC mains and radio transmitters/phones.
Click on the image for a larger version.

What else may be done to improve the performance?

Perhaps counter-intuitively, the use of a smaller photodiode can help a bit and provide at least as much signal output as a larger one, provided that the optics can focus the given amount of light from the distant source of light efficiently onto its active area:  A smaller device will have lower self-capacitance and thus will shunt a smaller amount of the AC currents being produced in response to the impinging, modulated light in addition to having a lower intrinsic noise contribution.  In the case of an optical receiver the active area of the device is less important than in some other applications as lenses and mirrors may be used to concentrate the light from the distant source onto the photoactive area.

When reducing the size of the device one must assure that the optics themselves will resolve the distant spot of light to an area that is not larger than the active area of the device as well as taking into account additional constraints with respect to the accuracy and stability of the aiming and pointing mechanisms.  For example, using reasonable-quality molded Fresnel lenses of common focal lengths (e.g. an f/D ratio of approximately unity) one can expect only to resolve a spot with a "blur circle" of approximately around 0.2mm at best while high-quality glass optics should be able to reduce this by an order of magnitude or better assuming a suitably-distant source, a corresponding small, subtended angle and proper paraxial alignment and focus.  If the resolved spot of light is much larger than the active area of the device - perhaps due to the device being too small for the optics ability to resolve or due to the quality and/or misalignment of the lens(es) - there may be an additional loss of available optical energy and signal-noise ratio as some of the light from the distant source is being "wasted" when it spills beyond the active area of the photodetector.
For more information on "spot sizes" using inexpensive, molded plastic Fresnel lenses see the article "Fresnel Lens Comparison:  A Comparison of inexpensive, molded plastic lenses and their relative 'accuracy' and ability to produce collimated beams" - link.
Aside from the reduction of the size of the photodiode or cooling, where else may one eke out greater performance from this circuit topology?

The Avalanche Photodiode:

The Avalanche Photodiode (APD) is a type of photodiode that contains an internal mechanism for amplification.  Simply put, instead of a single photon having a given probability of mobilizing a single electron when it impinges the active area of a standard PIN photodiode, in an APD, what might have been a single electron being loosed in a normal PIN diode that same electron event can cause the mobilization of many electrons via an "Avalanche" effect and providing amplification of the optical signal, hence the name.  The result of this intrinsic amplification is that the output signal from this diode from a given photon flux can be much higher than that of a standard PIN photodiode.

Because the signal from the Avalanche photodiode itself is amplified internally it is more likely to be able to overcome the effects of the capacitance on frequency response as well as the noise intrinsic to the JFET amplifier, support circuitry and components, providing the potential of producing a greater signal/noise ratio for a given signal. Typically an Avalanche photodiode is incorporated into a TIA (TransImpedance Amplifier) with good effect, but what about its use in the previously-described "Version 3" photodiode receiver circuit that utilizes JFET gate current?

The basic design:

From the previous article (link) one can see the basic topology of the "Version 3" circuit using a "normal" PIN photodiode depicted in Figure 2, below.
Figure 2:
A diagram of the "Version 3" optical detector that utilizes JFET gate current.  In this circuit Q1 and Q2 comprise a cascode
circuit with Q3 providing the majority of Q1's drain current while U1b is configured as a differentiator to compensate
for the low-pass effects of the intrinsic capacitance of D1, the photodiode and Q1.  Resistors R1 and R2 along with
C1 provide a filtered reverse bias for D1 which not only decreases its capacitance, but it also biases Q1 to
its operating state where it is drawing maximum drain current.  In this circuit the connection between the Photodiode (D1)
and the gate of the JFET is made in air and not on a circuit board to minimize capacitance, stray signal pickup and
most importantly a source of leakage currents and related noise.
Click on the image for a larger version.
In this design PIN photodiode D1, a BPW34, is reverse-biased via R1 and R2.  One of the main benefits of doing this is that the capacitance of D1 decreases from approximately 70pF at zero volts to around 20pF at the operational voltage, reducing the degree to which high frequency signal are attenuated by this capacitance.  A somewhat less tangible benefit of this is that in addition to photovoltaic currents produced by the impinging light, the bias also allows photoconductive currents to flow from the bias voltage, through the photodiode and into the gate of the JFET.  As noted in the original article, it is the presence of the gate-source junction of the JFET (Q1) and its conduction that limits the gate-source differential to around 0.4-0.6 volts, permitting D1's reverse bias to become established without the need of any additional noise-generating or lossy components.  In this configuration the drain current of the JFET is still proportional to the gate-source voltage (but with an offset of drain current greater than the "zero bias" drain current) and like a bipolar transistor's base voltage and current, the relationship between gate voltage and gate current is logarithmic.

What about replacing D1 with an avalanche photodiode?

Testing with an Avalanche Photodiode:

Like its more-sensitive distant cousin, the Photomultiplier tube, the avalanche photodiode requires a rather high bias voltage in order to function at maximum gain.  Rather than requiring a kilovolt or so as is needed for a photomultiplier, typical photodiodes may operate with up to "just" a few hundred volts.  Like the photomultiplier, the current required for "dark" operation is minuscule - a few hundred microamps is more than enough.

In perusing the various component catalogs I noted that Mouser Electronics carried some avalanche photodiodes - but as expected, there was a price:  Around US$150 at the time for just one APD.  In a compromise between size, availability and cost I chose the AD1100-8-TO52-S1 by First Sensor  (previously known as "Pacific Silicon Sensor") - a device with a round, 1mm2 (1.128mm diameter) active area - a reasonable compromise.  This device, which came with its own test sheet, indicated a maximum gain ("M" factor) of approximately 1000 occurring at 134 volts at a temperature of 25C.

In most ways using an APD is just like using a reverse-biased PIN photodiode - except that the reverse bias voltage will be much higher.  Perusing the literature and manufacturer's specifications one will note that many designs depict a temperature-compensated bias voltage supply, but further investigation reveals that this is necessary only if the device is being used at/near maximum gain (and maximum voltage) and/or if it is necessary to precisely maintain a certain gain over a wide temperature range.  For our application, we don't really care if the gain changes with temperature, so an arbitrarily adjustable supply is fine - and actually preferred.

In my initial research I noted that the internal action of any APD suffers an inevitable, but expected, effect:  As the gain goes up with increasing bias voltage, the intrinsic noise of the device itself increases at a faster rate than the gain.  What this means was that there is going to be a point at which a further increase of device gain will cause the signal to noise ratio to decrease even though the actual signal level continues to increase with bias voltage - but at what voltage might this happen, and would this "crossover" point occur at a point where we can expect the overall "gain+noise" to offer a net advantage over a PIN photodiode?

Building a prototype receiver similar to that depicted in Figure 2 I substituted an APD for D1 using a string of sixteen 9 volt batteries and a 1 megohm potentiometer with a 100k resistor in series with the wiper (and some bypass capacitors to ground) in lieu of R1 to set the bias voltage.  Placing this prototype in my "Photon Range" - a windowless room in my house where there is an LED mounted to the ceiling - I compared the sensitivity of this prototype to both my "standard" TIA receiver (the VK7MJ design) and an operational exemplar of my "Version 3" design.

Varying the voltage from 10 volts to around 140 volts I noted that at a bias voltage comparable to the reverse bias applied in Figure 2 (approx. 8 volts) the apparent sensitivity was roughly on par with that of the Version 3 unit after the signal levels were corrected to compensate for the smaller area of the APD as compared with the BPW34 (e.g. 1mm2 of the APD versus 7mm2 of the BPW34 - the larger size gathering proportionally more light in this lens-less system).  At around 130-135 volts, the output of the APD-based prototype was very high, but the weak, optical signals from the test LED were lost in the noise.  In the area of 35-45 volts I observed that while the overall signal levels, while significantly higher than they were at 8-10 volts, were a fraction of what they were at 130 volts but the signal/noise ratio was roughly 6-10dB higher than it was at the lowest voltage when the differences in active area of the APD versus the photodiodes in the test receivers were taken into account.

Comments:
  • The test receivers used BPW34 PIN photodiodes with an active area of 7mm2 while the APD has an active area of just 1mm2.  Because there are no optics in front of the photodiodes there will be 7 times as many of the LED's photons hitting the larger device, resulting in an approximate 8.5 dB difference in signal/noise - assuming all other parameters being equal.  It is when using the device in this "lens-less" configuration that this factor must be accommodated.
  • While it is theoretically possible to use a photomultiplier tube (PMT) in lieu of an APD, there are several practical concerns.  Even though an "S-1" type of photocathode has a peak in the red-NIR area, its low quantum efficiency makes it a rather poor performer overall.  The "931A" PMT - widely available surplus - has a more typical blue/violet peak response (type "S-4") in which the longer red wavelengths suffer greatly in terms of quantum efficiency and field testing of these devices by British amateur radio operators has shown that they offered no obvious advantage over the "Version 3" PIN photodiode design for "red" wavelengths.  As of the time of this writing the use of PMTs with more exotic photocathodes (such as multialkalai and GaAs) that are better suited for "red" wavelengths (but much more difficult to find surplus!) have not been field-evaluated.

A practical design:  The high voltage APD bias supply:

First, a few weasel words:
Even though the currents are very low, there is some risk of injury with the voltages involved (e.g. several hundred volts) and it is up to you to educate yourself about high voltage safety!  If you wish to construct these circuits, be aware of possible hazards and always assume that any capacitors are charged, even after power is removed.

You have been warned!
Because it is not convenient to carry around a lot of 9 volt batteries, a simple high voltage converter was designed to provide the  microamp-level current required for the APD bias supply and is depicted below in Figure 3.
Figure 3:
High voltage supply for the APD receiver.  U101a is an oscillator that drives Q101 to produce a high-voltage,
low-current bias for the APD.  The output is regulated via U101b and associated components to the voltage
set by potentiometer R111.
This design is a simple "boost" type switching converter using a high voltage transistor and an inductor to produce the needed bias.  In this circuit U101A forms an oscillator that drives the high voltage transistor Q101, and when Q101 switches off, the magnetic field of L101 collapses, producing a high voltage spike that is rectified by D101 and filtered and stored by C102, R106 and C103.  To regulate this high voltage a sample is divided-down by R108 and R109 and compared with a 5-volt reference from U102 that is made variable with R111:  If the output voltage is too high, U101b turns on Q102 to pinch off the drive for Q101.  Because I used an "ordinary" op amp that could not go all of the way to the negative supply rail, LED101 was put in series with the transistor's base to provide a drop of around 2 volts to assure that Q101 could be shut completely off.
Figure 4:
Inside the high voltage (bias) supply for the APD receiver.  Potentiometer
R111 and the indicator, LED101, are mounted in the front of the
case.  Both the high voltage generator and the receiver itself are powered
from a single 9 volt battery.  The typical combined current consumption
for the both sets of circuits is less than 35 milliamps.
Click on the image for a larger version.

LED101 also provides two other features:  It functions as a "power on" indicator, and since it is in series with Q101's base drive it is modulated at approximately 6.5 kHz (determined by experiment to be the frequency at which Q101 and L101 produced the highest voltage with the best efficiency) and can be used as an optical signal source to verify that the receiver is working.  Worth noting is that R112 is placed across the "hot" end and the wiper of R111 to "stretch" the high voltage end of the linear potentiometer's adjustment range a bit to compensate somewhat for the fact that near the maximum voltage, the gain goes up exponentially with the bias voltage, making fine adjustments at this setting easier.

The APD (optical) receiver:

The actual optical receiver section is depicted in Figure 5, below:
Figure 5:
The optical receiver which works in a manner very similar to that depicted in Figure 3.  In this implementation
the high voltage bias is applied to the cathode of D201, the APD, which has its anode connected to the gate of the JFET,
Q201.  Q201 and Q203 comprise a self-biasing, AC-coupled cascode amplifier while Q202 provides the a high-
impedance source for the bulk of Q201's drain current.  The components in the sections marked "HV Filter"
and "LV Filter" are used to keep the residual switching frequency energy from being conducted into these circuits.
As with other circuits of this type, the connection from the photodiode to the JFET's gate is made in air and not via a
circuit board trace to minimize capacitance, leakage currents and noise.
Click on the image for a larger version.

Not surprisingly this looks very similar to the "Version 3" optical receiver of Figure 2.  Notable features include an R/C filter consisting of R201, R202, C201 and C202 to remove traces of the 6.5 kHz power supply ripple from the high voltage supply while L201, C211, R215 and C212 do the same for the 9 volt supply that the receiver circuitry shares with the high voltage generator.  The two sections - high voltage supply and optical receiver sections - are separate, connected by a 3 foot (1 meter) umbilical cable, both to provide isolation of the extremely sensitive optical receiver from the electrostatic and electromagnetic fields of the high voltage converter and also to remotely locate the controls on the high voltage supply away from the lens assembly on which the receiver portion is mounted so that adjustments can be made without disturbing it.
Figure 6:
Inside the receiver portion of the APD receiver.  This section is physically
separated from the high voltage converter to prevent the switching energy
from getting into these extremely sensitive circuits.  In the center is
a small sub-board with the APD and JFET that is mounted on short pieces
of 18AWG wire to allow its position to be adjusted in all three dimensions
to provide both paraxial alignment and focus.
Click on the image for a larger version.

The APD itself is mounted on a small sub-board along with Q201 (the JFET) and the other capacitors noted in the box in Figure 5.  Most of Q201's drain current is provided by Q202's circuit, a current source, that operates at high impedance while Q203 is the rest of a cascode amplifier circuit that is designed to be self-biasing at DC and to provide gain mainly to AC signals.

The output of the cascode amplifier is passed to U201b, a unity gain follower amplifier.  This signal then passes to the circuit of U201a, a differentiator circuit that is designed to provide a 6dB/octave boost to higher frequencies to compensate for the similar R/C low-pass roll-off intrinsic to the APD and JFET itself:  Without this circuit higher frequency audio components of speech would be excessively rolled off, reducing intelligibility.  By design, the frequency range of the differentiator and its surrounding circuitry is intentionally limited so that low frequencies (below several hundred Hz) are strongly rolled off to prevent AC mains related hum from urban lighting from turning into a roar as are very high frequencies - above 5-7 kHz - which would otherwise become an ear-fatiguing "hiss" were the differentiation allowed to continue to frequencies much higher than this.

An interesting property of this circuit is that the "knee" related to this 6dB/octave roll-off occurs varies somewhat with the bias voltage and thus amount of device capacitance and, to a certain degree, its gain.  Because of this the frequency response of the APD/JFET circuit and the differentiator don't match under all operating conditions but experience has shown that it is better to have a bit of extra "treble boost" than not when it comes to making out words when the distant voice is immersed in a sea of noise.

A sample of the output from U201b, before differentiation, is also passed to J20, the "Flat" output.  The audio taken from this point, lacking differentiation, will sound a bit muffled under normal low-light conditions and it is not subject to either the high or low pass effects of the U201a differentiator which means that it will pass both subsonic and ultrasonic components as detected by the APD amplifier itself.  On the low end, the sensitivity is limited by 1/F noise which becomes increasingly dominant below a few 10s of Hz while on the high end it is again the capacitance associated with the APD and JFET circuits.  In testing it was observed that at this "Flat" output it was possible to detect signals from an LED modulated up to several MHz, albeit with significantly reduced sensitivity.  The main purpose of this output is to provide a signal point suitable for both subsonic digital communications as well as ultrasonic for experimentation with low/medium rate data, FM carriers and SSB signals.

In this circuit the amount of drain current in the JFET will vary depending on the individual properties of the JFET itself, the bias voltage, and the amount of impinging light.  Under "dark" conditions the "standing" JFET current was set to approximately 7-10 milliamps by the current source and the drain-source voltage varied from around 0.21 volts when the APD bias was just 12 volt to around 0.155 volts when the APD was operating at its maximum rating of 135 volts.  The specified JFET, the BF862, is typically capable of handling more drain current than this - and to do so would likely reduce its noise contribution slightly - but it was set at this level (with R205) to moderate battery current consumption.

Although it may have risked component damage, the APD circuit was "torture tested" to check ruggedness.  In a completely dark room a xenon photo flash was set off just inches away from the photodiode with the bias set at 135 volts.  While the receiver was deafened for a second or two, the time it took for the various circuits to recover (e.g. power supply, re-equalization of various capacitor, etc.), repeated tests like this did not do any detectable damage to the receiver sensitivity or noise properties, indicating that the APD and JFET were more than rugged enough to handle any conceivable event that might happen in the field, aside from directly focusing the sun on the photodiode!

This circuit has also been successfully used in broad daylight:  While the receiver worked, the background thermal noise from the sunlit landscape was the limiting factor for sensitivity, the recovered audio had quite apparent nonlinearity (distortion) with an altered frequency response and the ambient light and resulting photodiode conductivity effectively shunted the high voltage bias and device capacitance.  In short, in such high, ambient light conditions this circuit has no advantage over other optical receiver topologies such as the original "Version 3" or even a more conventional TIA (TransImpedance Amplifier) but its ability to be useful under such conditions is indicative of its versatility.

The results of in-field testing:

This receiver was first field-tested on a 95+ mile (154km) optical path during the September 2012 segment of the ARRL "10 GHz and up" contest:  For detail on this communication, read the blog entry "Throwing One's Voice 95 Miles on a Lightbeam" - link
 
Figure 7:
My end of the 95+ mile optical path during the session where the APD-
based optical receivers were first field-tested.  As seen in the picture
the optical path passes over urban lighting which tends to slightly raise
the noise floor due to both Rayleigh and lens-related scattering
effects.
Click on the image for a larger version.

During this test the optical (voice) link was first established using the "Version 3" PIN Photodiode receiver depicted in Figure 2.

With the reasonably clear air and the moderately long path we noted that we could reduce the LED current to a tiny fraction of the maximum before significant signal/noise degradation was noted.  At this lower LED current each station at opposite ends of the path switched from the PIN photodiode to the APD receivers and after tweaking our pointing and reducing the LED current even more we observed what turned out to be between 6 and 10 dB improvement in the signal-noise ratio - about what was observed on the indoor "Photon Range" with the initial prototype circuit.  It is likely that the actual improvement in sensitivity was greater than this, but because our respective optical paths passed directly over populated areas (see Figure 7) our ultimate noise floor was degraded by light pollution which included a thermal "hiss" and a low-level, harmonic-rich 120 Hz hum.

As was determined in the lab, the best signal-noise ratio in the field occurred with the APD biased in the 35-45 volt range where the "M" (amplification) factor was in the area of 3-10 (approximately 10-20dB gain).  At this rather modest bias voltage the "Gain+Noise" from the APD itself was sufficient to overcome much of the intrinsic noise of the JFET.  At higher voltages the gain continued to increase but the signal-noise ratio decreased at a faster rate until the APD's own avalanche noise drowned out the desired signal.


* * *

For more information about (speech bandwidth) optical communication, check out these links from my "Modulated Light" web site (link):

Be sure to check out the "ModulatedLight.org" web site's other pages as well!

[End] 

This page stolen from "ka7oei.blogspot.com". 
 


Wednesday, January 25, 2017

An A/B Battery replacement for the Zenith TransOceanic H-500 radio, with filament regulation

A friend recently gave me an old Zenith TransOceanic (ZTO) H-500 and after re-aligning it to get it into proper working condition I decided that I wanted to build a battery pack for it - both for "completeness" as part of making it look as original as possible and to allow the radio to be used outdoors, away from interference sources.  While it might be said that the GoogleWeb is lousy with options to replace the obsolete "A/B" battery used to power the Zenith TransOceanic, that wasn't a deterrent for me to design and build yet another one.

It is easy to use a lot of 1.5 volt cells to get the 9 and 90 volts required to operate the radio, but I decided to make do something different.

Figure 1:
The faux A400 "AB" battery, installed and working in the Zenith Trans
Oceanic H-500.  Contained therein are eight "D" type cells and circuitry
to produce the 90 volt "B" voltage and a regulated 9 volts for the
filament supply.
Click on the image for a larger version.
I threw a computer at it.

While it might seem odd to wield a microcontroller to solve a relatively simple problem on an antique, tube-type radio, it does make sense in a few ways as I'll outline below.

Design goals:

There are several things that I decided that this voltage converter should do:
  • Automatically power up when the radio is turned on and shut down when it is turned off. 
  • Cause no interference to radio reception.
  • Consume minimal current when the radio is turned off.
  • Produce a regulated B+ voltage so that radio performance is consistent.
  • Regulate the filament voltage so that the radio functions properly even when the battery is mostly discharged so that maximum use can be made of its total capacity.
While I was at it I decided that it should be able to do a few other things as well:
  • If the radio is on for a very long time (e.g. more than about 2 hours) do a "power save" shut down to (hopefully) prevent the batteries from being completely flattened.
  • "Lock out" the operation of the radio if the batteries are already extremely low.  Avoidance of completely killing the battery may reduce the possibility of their leaking.

Generating the "B+" voltage:

The "B Battery" (high voltage) needs of the ZTO are rather modest - approximately 90 volts at 5-20 milliamps.  Aside from using a battery of sixty 1.5 volt cells or ten 9 volt batteries in series there are two common ways to generate this sort of voltage electronically:
  • Use a step-up transformer to take the low battery voltage to the appropriate B+ potential, typically using a low-voltage mains transformer in "reverse" (e.g. applying drive to the secondary, rectifying high voltage from the primary.)
  • The use of a simple boost-type converter using a single inductor.
The first method has the advantage that it is possible to design it such that the switching of the driving transistors is "slow" enough (at a modest efficiency loss) that it does not produce harmonics that may be picked up by the receiver - even at the lowest receive frequencies, and without shielding.  If you are interested in a good discussion of this method visit Ronald Dekker's excellent page on the subject (link).
 
Figure 2:
Test circuit to determine the suitability of various inductors and transistors
and to determine reasonable drive frequencies.  Diode "D" is a high-speed,
high-voltage diode, "R" can be two 10k 1 watt resistors in parallel and
"Q" is a power FET with suitably high voltage ratings (>=200 Volts)
and a gate turn-on threshold in the 2-3 volt range so that it is suitable
to be driven by 5 volt logic.  V+ is from a DC power supply that is
variable from at least 5 volts to 10 volts.  The square wave drive, from a
function generator, was set to output a 0-5 volt waveform to
make certain that the chosen FET could be properly driven by a 5 volt
logic-level signal from the PIC as evidenced by it not getting perceptibly
warm during operation.
The second method - and the one that I chose - uses a boost-type converter as depicted in Figure 2.  The switching frequency must be much higher than one would use with an ordinary mains transformer, typically in the 5-30 kHz range if one wishes to keep the inductance and physical size of that inductor reasonably small.  With these higher frequencies and the typically "square" drive signals (which are rich in harmonic content) needed to obtain good efficiency there is a much greater likelihood that it will interfere with reception - particularly in the AM broadcast band.  While a bit of a nuisance, the interference potential may be easily mitigated by putting the entire circuit in a metal box and appropriately bypassing and filtering the leads in and out.

Raiding my inductor drawer I picked a few "power" devices (those capable of handling at least half an amp) in the range of 100μH and 1 mH and threw together the circuit in Figure 2, consisting of a high-voltage FET (Q), the inductor under test (L), a high voltage, high speed diode (D), a 22μF, 160 volt capacitor (C) and a 5.6k, 2 watt load resistor (R).  Connecting the FET's gate to the square wave (50% duty cycle) TTL-level output of a signal generator I measured each one in terms of output voltage, total output power and overall power conversion efficiency with respect to frequency.

As would be dictated by the plethora of design articles on the subject, not to mention data sheets of switching regulator chips, I noted that neither the value of the inductance or switching frequency was particularly critical to achieve the desired results.  In general, higher inductances produce a bit more output at the lower frequencies (a few kHz) while the lower inductances worked a bit better in the 10-30 kHz range, but all of the inductors did work over the entire range to a greater or lesser degree.  Settling on a decent-sized 330μH inductor - a value that is not particularly critical - I proceeded with the circuit design.
Figure 3:
Schematic diagram of the voltage converter.  See text for details.
Click on the image for a larger version.
The circuit:

Rather than go through a lot of theory I'll just describe the circuit that I designed and built - See Figure 3, above.

When the radio's power switch is turned on its filament circuit is connected and a voltage appears across the "Batt-" and "A-" leads and R7, a 10k resistor connected across switched-off FET Q4 which are in series with the filaments.  When this happens transistor Q3 is turned on, pulling the base of Q1, a PNP transistor in the high side of the BATT+ line, toward ground and turning it on and applying power to U3, a 78L05 voltage regulator, and microcontroller U1, a PIC12F683.  After a short initialization delay the microcontroller activates the "PWR_SW" line which turns on Q2 which assures that Q1 is always turned on even if the filament switch is turned off abruptly and Q3 turns off or, as we shall see, when the battery voltage is at or below the filament regulator's set point.

At this point the microcontroller enables interrupt-driven code to produce the high voltage (B+) output by monitoring it via resistor divider R18/R19/R20:  If the voltage is below the threshold, the duty cycle of the PWM signal output on the "SW_DRIVE" line is increased to force more energy storage in the inductor (L1) - up to a maximum limit of around 80%, set in software.  If the voltage is above the threshold, the duty cycle is decreased - down to zero and even into "discontinuous" mode (e.g. the PWM signal intermittently turned off and on) if necessary as would be the case if there were no load on the output.  In this way the output voltage is appropriately regulated, typically to 90 volts as set by R19.  In this circuit when the PWM signal turns off Q5, the high voltage FET, the magnetic field in L1 collapses and induces a high voltage across it.  The current resulting from this field collapse is rectified by high-speed, high-voltage diode D2 and stored and filtered by C8 and additionally filtered and smoothed by R21 and C9.
Figure 3:
The (mostly complete) converter board.  The high-voltage FET (Q5) is
in the lower left corner while the filament regulator FET is in the lower-
right corner.  In the upper right corner is U2, the rail-to-rail dual op-amp
that is part of the filament regulator.  Because of the very small amount of
heat being dissipated by any component, no heat sinks were required.
The high voltage filtering components and the optoisolator are in the
upper left corner.
No circuit board is available - but if you design one, I'd be happy
to post information about it and give you credit! 
Click on the image for a larger version.

Because the battery voltage could be as high as 16 volts if ten fresh "1.5" volt cells were used it is necessary to regulate the filament voltage down to something around 8.5-9 volts.  Op amp section U2b is configured as a "difference amplifier" (a.k.a. subractor) that measures the voltage difference between the "A-" and the "A+" lines (the filament supply to the radio) and this calculated voltage difference is output from U2b and applied to the inverting input of U2a via scaling potentiometer R14.  The voltage at the inverting input of U2a as set by this potentiometer is compared to the "reference" voltage applied to its non-inverting input and if the voltage is low, its output voltage is increased so that FET Q4, which is placed between the A- and BATT- connections, conducts more to increase the filament voltage.  Conversely, if the voltage is too high, the output voltage of U2a to Q14's gate is reduced, decreasing its conductivity.

The use of the circuitry of U2b is necessary because neither the A- or A+ (filament) leads are referenced to the circuit ground (e.g. they are sort of "floating") which makes it necessary to measure the difference between those two leads to ascertain the actual filament voltage.  If the battery voltage does get low enough that Q4 is completely "on", the voltage across R7 will disappear and Q3 will turn off:  It is because this can happen that we must have activated Q2 to keep the microcontroller's power turned on and this is also why we cannot use the voltage drop that we used to tell if the radio was turned on to also detect if the filament current has ceased to flow when the radio is turned off.

Note:  It would have been possible to have used a microcontroller to regulate the filament voltage in a manner similar to that in which the high voltage is produced, but a programming bug or crash could cause the fragile, expensive tubes to be exposed to the full battery voltage whereas a malfunction of the high voltage generator is unlikely to cause damage to the radio.

A short time after the high voltage converter is enabled the "FIL_SW" line is set high.  Because the microcontroller has low-impedance FET output drivers, this pin's voltage is essentially that of the 5 volt regulator and it is used as the filament voltage reference.  Similarly, if the microcontroller sets the "FIL_SW" line low (zero volts) this will shut off the filament supply.

With the use of a MOSFET (e.g. Q4) as the filament control device, the series regulation of the filament has a very low drop-out voltage - that of the voltage drop across the FET, limited by its own "on" resistance, and the wiring used to carry the filament current - and this drop can be as low a few 10s of millivolts.  What this means is that if the filament voltage is set to 9.0 volts by R14, as long as the "A" battery voltage exceeds that by a few 10s of millivolts, the filament will always be maintained exactly 9.0 volts but if the "A" supply (battery voltage) drops below 9.0 volts, Q4 will be turned fully on and the filament voltage will be within 10-20 millivolts of that battery voltage.  Compared with the operation to a typical "low dropout" regulator IC that has around 0.15-0.3 volts drop, the circuit used here offers a lower voltage drop and better radio performance in those situations, particularly when even a few tenths of a volt can make a lot of difference!
Figure 4:
Inside the completed voltage converter.  All leads going in and out are
bypassed with low-ESR electrolytic capacitors and further filtered with
series chokes as shown in Figure 3.  The use of a completely shielded
enclosure (top not shown) is necessary as direct E-field radiation from the
circuit will otherwise be heard on the radio.  This box is made from
cut pieces of circuit board material, soldered at the seams inside and out,
with cut-in-half nickel-plated brass standoffs soldered to the board being
used to support the circuit and at the corners to attach the lid.
Click on the image for a larger version.

A second or so after the application of the filament voltage - enough time for the tubes to warm up - the microcontroller starts to "look" at the current drawn on the B+ lead as detected by U4, an opto-isolator that is in series with this supply.  Once the tubes warm up and begin drawing current, U4's internal LED turns on, activating its internal phototransistor which then pulls the "HV_IMON" (high voltage current monitor) line low, indicating to the microcontroller that the radio is now operating.  At this point the microcontroller is in a mode where it will repeatedly check to see that current is drawn by the radio on the high voltage line.

When the radio is turned off the current on the B+ line will disappear due the loss of the tubes' emission caused by the filaments being turned off and, possibly, the B+ line being disconnected.  When this happens the LED in optoisolator U4 will turn off, its phototransistor will stop conducting, and the "HV_IMON" line will be pulled high indicating to the microcontroller that the radio has been turned off.  After a short "debounce" period to verify that this loss of current wasn't due erroneously detected the microcontroller will shut off the high voltage generator, set the "FIL_SW" line low, powering down the filament regulator, and then set the "PWR_SW" line low which then disconnects the microcontroller's power source from the BATT+ line, removing load from the battery.

Why use eight 1.5 volt cells rather than just six to get the filament voltage?

Why not just use six 1.5 volt cells to get "9 volts" for the filament string?  As it turns out only a set of six fresh 1.5 volt cells will actually produce 9 volts - and the voltage drops from there.  If one consults the manufacturers' specifications for alkaline cells it will be noted that the majority of the useful life of typical "1.5 volt" cells occurs with their voltage being in the range of 1.2-1.3 volts and it isn't until a cell gets all of the way down to 1 volt (for a total of 6 volts to the filaments for a six cell battery) that just 80% of the cell's capacity has been exhausted.

In this radio I noted that below an "A" battery potential of 8 volts (e.g. 1.33 volts/cell for 6 cells) the sensitivity started to drop and by the time it has dropped to around 7.5 volts (1.25 volts/cell for 6 cells) the radio was practically deaf with the oscillator abruptly stopping just below this.  Poking around inside the radio I noticed that at 9 volts, the series voltage drop across each of the tubes' filaments was very close to that shown on the schematic diagram in the service manual, but by the time it dropped to 7.5 volts it had become unequal, with the 1L6 converter tube being disproportionately affected and its filament voltage at or below 1 volt.  Interestingly this drop-off in sensitivity did not appear to be related to frequency:  The radio still worked at all frequencies with a filament voltage just above where it cut off, but it was just as deaf on the low bands as it was on the high.  Because the 1L6 tube is the component in this radio that is the most difficult to find, it would also make sense to construct the battery supply in such a way that it would allow the best operation from a "weak" tube, anyway.

For this reason I decided to use the battery voltage of eight 1.5 volt cells rather than the "9 volts" obtained from six cells for two important reasons:
Figure 5:
Inside the faux "AB" battery box for the Zenith TransOceanic.  Eight
"D" cells are used in four holders (one 4-cell,  one 2-cell and two 1-cell) which,
along with the converter box, are screwed down to some plywood (3 layers of
3.2 mm "luon") which itself is glued to the bottom of the box.  The cover,
made from the same circuit board material as the box containing the circuits,
has both of its surfaces electrically connected using thin, copper foil soldered
to each side to assure that an electrical connection is made to the box
itself when the cover screws are tightened.  The authentic-looking replica
battery box and radio connector were obtained from "Edsantiqueradios.com".
Without having made the voltage converter smaller, there is room only for
eight "D" cells in the box.
Click on the image for a larger version.
  • The higher voltage of eight 1.5 volt cells (12+ volts when fresh) would allow the total filament potential ("A" voltage) to be regulated down to 8.5-9.0 volts.  (For longest filament tube life, one should run the filament string in the 8.0-8.5 volt range - the lower end being somewhat preferred if the radio's performance is still acceptable.)
  • The use of an extra two cells will allow the use of more of the battery capacity.  For example, with 8 cells discharged to 1 volt, each, around 80% of the cell's useful life has been utilized with the ending voltage still being 8 volts.  Contrasting this to the use of just six cells, at a total "A" voltage of just 7.75 volts (approx. 1.3 volts/cell for 6 cells) 40-60% of the life of the cells will remain, but the radio will likely be getting deaf or even may not be usefully operational!
  • In theory, ten 1.5 volt cells could be used.  Because the voltage of a "fresh" 1.5 volt alkaline is around 1.6 volts, this could expose some of the devices, particularly the electrolytic capacitors and U2, to voltage at or above the official maximum rating.  Practically speaking these devices will likely survive this, particularly since the voltage will very quickly drop under the load presented by the radio into the "safe" range.  The use of one or two additional 1.5 volt cells (e.g. 9 or 10) won't add more than 10-15% of "run time" to the radio so it is not likely to be worth using more than eight 1.5 volt cells.  (Nine cells would work as well, provided there was space in the battery box and that you were OK with using an odd number of cells.)
  • The typical filament current of this radio is on the order of 50 milliamps.  At a battery voltage of 12 volts where 3 volts is dropped by the series regulator, approximately 150 milliwatts is dissipated as heat - about 25% of the total filament power, or around 8% of the radio's total power consumption.  Were a switching regulator used for the filament its efficiency would likely be in the 85-90% range and increase of efficiency over the linear regulator would likely not be worth the added complexity.  Considering that the average voltage of the battery over its life will be around 10-10.4 volts (approx. 1.25-1.3 volts/cell) with a dissipation of only 70 milliwatts the difference in loss will be even lower.
With a fresh set of eight 1.5 volt "D" cells the current consumption was measured at 140-150 milliamps at very low volume and peaking to well over 250 milliamps when the volume was set to maximum on a strong station (lots of audio distortion!) with the filaments accounting for around 50 milliamps of the total.  While it has not been empirically tested (it's not particularly cheap to buy eight "D" alkaline cells just to run them down!) the estimated run times at "room" temperatures and normal receive volume to 1 volt per cell for various sizes of alkaline cells, based on manufacturers' data sheets are:
  • For "AA" size:  15-20 hours with reduced performance for an additional 1-2 hours.
  • For "C" size:  30-40 hours with reduced performance for an additional 3-5 hours.
  • For "D" size:  70-90 hours with reduced performance for an additional 6-10 hours.
If just six cells were used the filament voltage would drop below 7.5 volts in about half the time noted above and by then, the radio's performance will have likely diminished considerably.  In contrast, using eight cells and a filament voltage regulator the performance will remain essentially unchanged until the cells are about 80% discharged (around 1 volt/cell) and the radio's performance will drop from there.

Note that this circuit can be powered directly from a 12 volt supply or battery - just heed the warnings below about NEVER allowing the "Batt-" line to come in contact with the "A-" lead - or any part of the radio's chassis.

Additional comments about the circuit:

It should be noted that the "BATT-" and "A-" lines are isolated from each other.  These two lines should never be connected to each other as that would prevent the closure of the filament switch from being detected when the radio is turned on and it would bypass the filament regulator, exposing the tubes' filaments to the full battery voltage, likely destroying one or more of them!  The reason for putting the filament regulation in the negative lead is allow an N-channel FET to be used and to avoid the use of a P-channel device in the "high" side and the complications required in driving this device and keeping its circuit stable (e.g. avoiding spurious turn-on events and momentary loss of voltage regulation) when the unit is powering up or down.

Even more circuit comments:
  • Resistors R8 and R17 are used to bias their respected FETs "off" by default.  This is necessary as the outputs of the microcontroller are high-Z unless/until it is operating and these FETs could randomly turn on due to leakage currents without them.
  • Similarly R15, on the "reference" voltage for U2's filament regulator circuit from the microcontroller, pulls that output down before the processor initializes its outputs from their default "Hi-Z" state, eliminating a possible "glitch" of the filament voltage during circuit start-up and shut-down.
  • U2, the filament voltage regulator, MUST be a rail-to-rail input and output op amp.  An "ordinary" op amp such as the '1458 or '358 WILL NOT WORK PROPERLY under all conditions.  Some parts suggestions for suitable op amps are included in the schematic diagram of Figure 3.  In other words, if you use a "normal" op amp it is possible that this circuit will misbehave and expose the filaments to excessive voltage.
  • Resistor R9, a 470 ohm resistor in series with the output of U2a and FET Q4, isolates Q4's gate capacitance, preventing instability of the op-amp while C6 provides frequency compensation for the regulator circuit.
  • When powered down the quiescent current of this circuit is approximately 7μA and is a result of the battery voltage (minus the drop of D1) always being applied across the B+ voltage divider string R18, R19 and R20.  This amount of current is comparable to the self-discharge rate of modern alkaline cells and can generally be ignored.  If this amount of current were to really bother you, the  voltage converter circuit could be powered from the "V+_SW" line and transistor Q1 could be replaced with a P-channel power FET as noted on the diagram.
  • LED1 and LED2 are optional.  LED1 will glow when the microcontroller activates the "PWR_SW" line and can be used for troubleshooting.  For example, if no current is being drawn from the B+ line - or the converter is not working - the software will continually cycle:  It will turn on the high voltage, wait for current to flow and when not seeing it, it will turn off the high voltage again and retry after a few seconds causing LED1 to turn on and off.  LED2 is also optional and is used to indicate when the circuit is powered up by Q1, either by the microcontroller turning on Q3 and/or Q2 being activated by the voltage drop across R7 when the filament switch is closed.  If desired, it may be mounted to the battery box so that it is visible when radio's rear cover is open.
  • Transistor Q5, used in the high voltage "boost" converter, must be rated for at least 200 volts and it should have a "logic level" gate threshold appropriate for turning the FET (more or less) fully on at just 5 volts:  Some suggested device types are noted on the diagram (Figure 3).  An additional device worth considering is the ON Semiconductor NDD02N40-1G, a 400 volt, 1.1 amp FET that has a suitably low turn-on threshold - and it's pretty cheap.
  • Components TH1, a 1 amp self-resetting fuse, and diode D1 protect the circuit against shorts or accidental reverse polarity by limiting the current to a reasonable value should this occur.  TH1 may be replaced with a 0.75-1 amp fast-acting fuse if so desired.
  • The PWM (switching) frequency is approximately 15.625 kHz and is based on the microcontroller's internal 8 MHz clock.  Both 7.8125 and 31.25 kHz were tried and the conversion efficiency was slightly lower (e.g. approx. 1-5%) with the 330 μH inductor value chosen - an indication that the actual value of L1 isn't particularly critical.
  • The value of L1 may be anything from 220μH to 470μH - and even a bit beyond this range.  Make sure that the inductor used has a current rating of at least a half an amp or else internal resistive losses will significantly impact conversion efficiency.  If available, a toroidal inductor or other shielded type is preferred as it better-contains its magnetic field than solenoid styles.
  • The measured efficiency of the boost converter of the prototype was greater than 80% despite the power lost in R21, the "filter" resistor in series with the B+ output.
  • The 15 volt maximum supply voltage limit is set by the voltage rating of op amp U2 and possibly the ratings of the electrolytic capacitors exposed to the battery voltage.
  • If one chose to use just six 1.5 volt cells instead of eight, never supplying more than 9 volts, the "FIL_SW" line would be connected directly to the gate of Q4 and the circuitry related to U2 would be omitted.  Do note that six "fresh" 1.5 volt alkaline cells could initially produce a bit over 1.6 volts/cell and expose the filament string to over 9.6 volts.
  • The diagram and pictures show the use of feedthrough capacitors (4000pF) to pass the voltages through the shielded box.  Feedthrough capacitors are somewhat difficult to get, but good results may be obtained by using good-quality monolithic ceramic (NOT disk ceramic) capacitors instead, placing them - using very short leads to a solid ground plane - where the wires pass through the hole in the shielded box.  These capacitors are typically square in shape and rather compact and available in both leaded and surface-mount form.  Remember that for the B+ output a capacitor with a rating of at least 100 volts must be used:  Any value from 0.0022μF to 0.1μF may be used.
  • If you build this sort of circuit make absolutely certain that you simulate the filament string with a 150-200 ohm 1/2-1 watt resistor and the B+ load with a 10k, 1-2 watt resistor and verify that the circuits are working properly BEFORE connecting it to a radio.  While a brief bit of over-voltage on the B+ line (to perhaps 130-150 volts) will likely not harm the radio, more than 9 volts on the filament line, even for a moment, will probably ruin one or more of the fragile and expensive tubes!
  • About that "auto power save" feature mentioned at the top of this article?  After two hours of uninterrupted operation the microcontroller will modulate the filament line with an intermittent tone and drop the B+ voltage to about 50% causing the radio to partially mute with the alarm tone sounding in the speaker.  This "beeping" will continue for about a minute before the microcontroller turns off the filament and high voltage supplies, dropping the current consumption from around 150 milliamps to about 6-12 milliamps - the quiescent current of the remaining circuitry.  Turning the radio off for 5-10 seconds and then back on will reset this timer at any time.  The down-side of this is that if the radio shuts down in this way, one may forget that the radio is even on, still drawing a few milliamps, except for the fact that the front lid of the radio will still have been in its upright position!  If the battery voltage is less than around 7.5 volts (0.9375 volts/cell) the radio will be "locked out" and will not even turn on, but at this voltage the batteries are not only quite discharged, but their internal resistance will be rapidly increasing as well and little run time would have been left.
Figure 6:
A handy "map" showing where the various RF adjustments may be found.
This doesn't really have too much to do with the article, but since I made it
when I was aligning the radio I thought that I might as well post it here!
Note that locations of some of the trimmer capacitors - particularly those
in the lower-left corner - will vary with different production runs.  Some of
the alignment points shown in this picture are also omitted in the
"official" H500 service manual and thus have no parts designations:  These
adjustments are peaked at the frequencies indicated on the drawing.
Click on the image for a larger version.
How well does it work?

As can be seen in Figure 1 the circuit board and the eight "D" cell battery is concealed in a replica battery box that is situated exactly where an original "AB" battery would be placed.  Then the power switch is turned on it takes a bit over a second for the computer to power up, do its checks and for the tubes to warm up and the radio begins playing while the power-off is detected within two seconds of the radio power switch being turned off.

With the shielding of the circuity and bypassing of its leads there is no detectable interference caused by switching voltage converter.  With the filament and B+ voltage being regulated to the same as a "fresh battery" or AC mains voltage, the sensitivity and audio output capability are maintained until the battery is more than 80% depleted.

In other words, it works just as it should!

* * * * * * * * * * * * * * * * * * * * * * *

If you are interested in the code for this (written in "C" using the PICC compiler) or just a .HEX file so that you can program a PIC12F683 yourself, or if you are interested in getting an already-programmed PIC12F683, let me know via a comment.

And before you ask:  Sorry, but I can't build you one at this time...

[End]

This page stolen from "ka7oei.blogspot.com".
 

Thursday, January 12, 2017

A low power PSK31 transmitter using a Class-E power amplifier and envelope modulation

Back in 1999, not too long after the first appearance of PSK31, I decided that I wanted to construct a beacon transmitter that would operate using this mode, but at the time the only practical means of generating PSK31 was with a computer, a sound card and an SSB transmitter.  Not wanting to tie up that much gear for this purpose I set about to use the PIC16C84 microcontroller which was popular among the homebrew builders at the time.

By this time the AM broadcast band had (relatively) recently been expanded up to 1705 kHz but very few stations occupied the new 1605-1705 kHz segment.  In perusing the FCC rules I noted that part 15§219 had been modified to allow low-power experimental operation in this new segment of the AM broadcast band and I decided that with the lack of activity in this frequency range that it was a good time to put up a "MedFER" (Medium Frequency Experimental Radio) beacon.
Figure 1:
The "Balanced Modulator" (Baseband) version of the PSK31
transmitter/exciter.  Built to test a concept, it has a few flaws,
but it did work.
Click on the image for a larger version.

The balanced modulator method

Upon investigating various methods of producing a PSK31 signal I experimented with the generation of a bipolar baseband signal that could be applied directly to a balanced mixer.  While this method worked well it had the problem than it required that all following stages be linear.

A diagram of the prototype of that transmitter may be seen in Figure 1.  For this transmitter a crystal-controlled oscillator is constructed using two transistors (Q1, Q2) and the output is buffered by U3, a 74HC00 quad NAND gate.  The frequency used for this circuit was unimportant as it was a "proof of concept" and I (think that I) used a 4.9152 MHz crystal which, although not in any amateur band, still allowed an "across the room" reception with a short length of wire as an antenna.  Following the first U3 NAND buffer the remaining sections are used to provide a two phase signal with the output split 180 degrees which fed a very simple balanced modulator consisting of just two diodes, a few capacitors and some resistors.

To provide modulation a PIC16C84 is used to provide a 32-step staircase modulation using PWM techniques.  This PWM output, done using "bit-bang" software uses a frequency of 1 kHz which is exactly 32 times that of PSK31's 31.25 Hz baseband frequency, and is then filtered with a two stage R/C low-pass filter network consisting first of a 4.7k resistor and 0.1uF capacitor followed by a second stage with a much higher impedance consisting of a 150k resistor and 0.033uF capacitor providing around 3dB of roll-off at the 31.25Hz baseband frequency and about 40dB of attenuation at the 1 kHz PWM rate and an acceptable amount of Inter-Symbol Interference ("ISI").  The result of this filtering is that the vast majority of the 1kHz energy is removed, leaving a pretty clean 31.25 Hz baseband signal.

Figure 2:
Phase diagram of balanced modulator
circuit in Figure 1.  The propagation
delay of the gates result in a rather
imprecise 180 degree phase shift
causing the upside-down "Vee"
in the phase diagram.
The filtered PWM output is then buffered and split into two signals, one of them inverted, using several op-amp sections and these two signals are applied differentially via simple R/C networks across the two diodes:  If the baseband signal from the PWM output were to go "positive" (e.g. above the mid-supply voltage)  the other side would go "negative" and turn on one diode, but it if were to swing the other way the other diode - fed with the RF signal that was 180 degrees out of phase with the first - would be turned on.  The end result is a fairly nice, linear BPSK envelope and baseband waveform when viewed on a receiver with an oscilloscope.

While it worked to prove a concept, this signal has a few shortcomings.  First, the RF signal from the oscillator and buffer is not likely to have a precise 50% duty cycle which means that a bit more RF energy would be available in one phase than the other, resulting in a somewhat "lopsided" BPSK amplitude envelope that only minimally affects demodulation and overall signal quality.  The other problem has to do with a NAND gate being used to provide the 180 degree phase shift (e.g. signal inversion) in that the addition of the inverting gate adds a few 10s of nanoseconds of propagation delay.  While this doesn't sound like much, it does amount to a significant number of degrees of phase even at low HF frequencies and the end result is that the "Phase Diagram" (see Figure 2) is slightly distorted and produces the inverted "vee" pattern.

While I could have gotten this method to work (e.g. used a bandpass/lowpass filter to get a nice, clean sine wave and a transformer to get the 180 degree phase shift) it does hove a down side:  Again, all subsequent stages would need to be linear.  While not a great technical problem it did mean that for the MedFER transmitter, which has a 100 milliwatt DC input power limit according to FCC rules, a linear final amplifier would have at best around 70% efficiency which would mean that I'd lose a bit more than 1dB of signal over an amplifier that was 100% efficient.  While this may not sound like much I figured that I could do better with a more efficient amplifier scheme.

Comment:
This "baseband" PSK31 signal produced using the differential op amp scheme noted above was successfully applied experimentally to some "digital only" radios such as the Small Wonder Labs "PSK" series.  This was accomplished by "lifting" the balanced modulate above DC ground via capacitive RF coupling and applying the modulation differentially to the diode ring mixer and shifting the carrier oscillator to move this "DC" signal into the crystal filter's baseband.  The pages linked near the end of this article provide details on this modification.

The Amplitude Modulator Method

Having proven the ability to produce a reasonable quality PSK31 waveform with a lowly PIC I decided to try a different approach:  Apply high-level modulation to the output amplifier stage.  What's more, this amplifier stage need not be linear at all:  It could be a conventional Class C stage which could boost the efficiency to something around 80%, but I decided on going a step farther and use a Class-E amplifier.

Figure 3:
Diagram of the "AM" version of the transmitter using separate amplitude
and phase modulation paths, allowing a non-linear but highly efficient
Class-E output amplifier to be used.  The capacitor, diode and resistor
on the gate of Q1, the output transistor, are used to prevent the FET
from being stuck "on" and shorting out the power supply should
the RF drive disappear for any reason and the output of the NAND
gate driving it be left in a "high" state.
Click on the image for a larger version.
I first became aware of the Class-E amplifier more than a decade earlier when my friend Mark, WB7CAK, designed one for his LowFER (Low Frequency Experimental Radio) beacon that operated in the 160-190 kHz "experimenter's" band authorized by §217 of FCC part 15.  As with MedFER operation, the input power was also limited - 1 watt in this case.  After a bit of number crunching and fiddling on the workbench Mark came up with a simple circuit and a few basic equations that described how such an amplifier could be built and published an article in the Western Update - a small publication tailored mostly for LowFER.  Because this publication may be difficult to find I have reproduced it with permission from the author and it may be found here:  (Link).

While the maths behind the derivation of the operation of a Class-E amplifier can be somewhat involved, the concept is quite simple:  When the drive signal to the transistor - typically a power MOSFET at LowFER frequencies - goes low, the transistor shuts off and it does this quickly (e.g. driven "hard") so that transistor spends as little time as possible "partially" conducting in between "on" and "off".  When the transistor turns off the voltage on the drain rises, being pulled up by the choke in the circuit, but then falls again due the "ringing" of a resonant circuit on the output tank.  Because this tank circuit is tuned appropriately, precisely at the time that the drain voltage hits zero the output transistor is switched back on.  The result of these two events is that the FET is either completely on or off which means that little or no power is dissipated in it and when the FET is (quickly!) turned back on, it does so just when the voltage swings to zero, practically eliminates any losses that would occur at that instant due to the intrinsic resistance of the FET absorbing the current and from other losses of components from tank circuit being "shorted out" when voltage is present.

Figure 4:
The constructed MedFER beacon transmitter, built on the bottom
of a weather resistant outdoor enclosure to be mounted at the base
of the antenna.
The result of all of this is an RF amplifier that (exclusive of the drive signal) is demonstrably capable of 95%-98% efficiency!  In the MedFER and LowFER world this means that with our power level being limited on the input, we will have, for all practical purposes, all of our input  power at our disposal rather than, say, 70-80% of it as would be the case with almost any other amplifier type.

The obvious problem with a Class-E amplifier is that the drive signal must be a fast rising/falling square-shaped wave that slams the transistor on and off which means that amplitude modulation of that drive signal is not easily managed if efficiency is to be maintained.  What one can do is to modulate the power supply feeding the amplifier instead.

Remembering that a PSK31 signal consists of two parts - the amplitude modulation and the phase shift - we can split these two signals in the modulator.  The first part, amplitude modulation,  may be done by modulating the supply voltage of the output amplifier stage.  The second part, phase modulation, may also be done early in the path of the drive signal simply by flipping the phase of the RF signal under computer control.  In order to keep the signal "clean" all we really need to do is to time the flipping of the phase with the amplitude being brought to zero so that we don't transmit the broadband "click" that would otherwise occur when we did this abrupt phase shift.  The schematic of this transmitter is depicted in Figure 3.

Figure 5:
The phase diagram of the signal
produced by the "Amplitude
Modulator" MedFER PSK31
beacon transmitter.  The phase
shift is precise and the intermodulation
products are well within the tolernaces
dictated by good operating practice.
In this circuit the frequency-determining crystal oscillator operates at four times the transmitter frequency, or around 6.8 MHz in the case of the MedFER transmitter.  During construction it was observed that at around 1.7 MHz it was was easier to achieve Class-E operation at this power level with a drive waveform that had a 25% duty cycle so a 74HC4017 counter was used, wired as a divide-by-four giving two 25% duty cycle outputs, 180 degrees apart.  To select which of these signals were to be used a simple MUX and driver was constructed using four NAND gates, this time being designed so that the same amount of propagation delay would occur during either phase to eliminate the upside-down "Vee" seen in Figure 2.

The PWM signal was generated using simple R/C filtering in the same way as it was for the balanced modulator circuit, but this time op amps were used to set the offset and gain (or "span") so that the baseband waveform could be precisely adjusted in both amplitude and so that when the baseband signal went to zero, the output power from the Class-E circuit would as well, compensating for the voltage offset of the series modulating transistor, emitter-follower Q4.  The output transistor, Q3, is a low-power MOSFET wired into a simple L/C "tank" circuit that is tuned to result in the coincidence of the zero crossing of the drain voltage and the transistor being turned back on by the 25% duty cycle drive signal.  Multiple taps are provided on the tank coil making it easy to set both the output power and match it appropriately to the load presented by the resistance seen at the loading coil.
Figure 6:
Loading coil used to match the transmitter output to the
feedpoint impedance.  This coil is wound using 3/8"
copper tubing and uses a variometer inside the coil
to provide a low-loss means of adjusting the inductance.

For modulation the PIC produces a semi-sine waveform that looks very similar to one "cycle" on the double-frequency output of a full-wave diode rectifier and when this waveform amplitude is taken to "zero" another output of the PIC causes a phase switch to occur.  It is in this way that the BPSK modulation is broken into two parts - the phase change and the modulation envelope - and we are able to use a highly efficient, non-linear amplifier for the output.

After constructing this I later learned that a similar scheme was applied to amateur satellites (starting with OSCAR 7) that included linear transponders.  In order conserve precious power, the linear transponders were constructed using the "HELAPS" (High Efficiency Linear Amplifier using Parametric Synthesis) system where the amplitude and phase components of multiple signals in the satellite's passband were converted into their phase an amplitude components allowing both energy-saving class-C RF amplifiers and DC-DC switching converters to be used, the end result being a faithful, amplified reproduction of the input signal with a lower power budget that would have otherwise been required. This system was proposed by Dr. Karl Meinzer, DJ4ZC, and you can read about it on the AMSAT.DL web site here - link.

Where is it now?

This beacon was mounted in its enclosure on the roof of my house in 1999 and a rather large loading coil (see Figure 6) was to match its output impedance to the top-hatted 3 meter vertical antenna  - and it is there to this day.  While not regularly used, it still works, provided that the tuning of the loading coil and variometer is checked before operation.  Since the beacon was constructed, more broadcast stations have taken to the air in the "new" AM segment, but its operating frequency - nominally 1704.965 kHz - is just below the top edge of the band and as far away from QRM as is possible.

In the past the BPSK31 signal from this beacon has been copied during the daylight hours at a distance of 75 air miles (approx. 120km) and it had been copied in various places in the western U.S. at night.  This beacon has since been modified to so that it may be externally on-off keyed so that "QRSS" (low-speed Morse with multi-second "dit" lengths) could be sent in addition to PSK31 allowing even greater distances to be spanned under more diverse conditions.
 
I haven't done much with the code for this transmitter other than add a few features when it was ported to the (then) newer PIC16F84.  Needless to say, there are more modern devices available that contain hardware that would have simplified the design such as that to generate a much higher frequency and higher resolution PWM signal and perhaps one day I'll investigate their use.

For more information on this and related projects - including schematics, various applications, more pictures and some source code, visit the "CT Medfer Beacon" web page - link and related pages linked from there.

[End]

This page stolen from "ka7oei.blogspot.com".

Saturday, December 31, 2016

A simple push-pull audio amplifier using russian rod tubes and power transformers

As one sometimes does, I was perusing EvilBay a while back and saw some ex-USSR sub-miniature pentode tubes for sale.  In looking up the part number - 1Ж18Б, which is usually translated to "1J18B" (or perhaps "1Zh18B") I was intrigued as they were not "normal" tubes.

Many years ago I'd read about the type of tube that is now often referred to as a "Gammatron" - a "gridless" amplifier tube of the 1920s, so-designed to get around patents that included what would seem to be fundamental aspects of any tube such as the control grid.  Instead of a grid, the "third" control element was located near the "cathode" and "anode" - or even a pair of anodes.  As you might expect the effective gain of this type of tube was rather low and despite its working, it really didn't catch on.  It was the similarity between the description of the "Gammatron" and these "rod" tubes that interested me.
Figure 1:
A close-up of a 1J18B tube.  Note that the internals are a collection of rods
rather than "conventional" grids and plates.
Click on the image for a larger version.

Some information on the "Gammatron" tube - not to be confused with the later-used "Gammatron" product name - may be found at:
  • The Radio Museum - link.
  • The N6JV virtual tube museum - link.

In reading about these peculiar "rod" tubes I became intrigued, particularly after reading some threads about these tubes on the "radicalvalves" web site (link here) and the "radiomuseum" site (that link here).  Since they were pretty cheap I ordered some from a seller located in the former Soviet Union.

This past holiday week I managed to get a bit of spare time and decided to kludge together a simple circuit with some of these tubes which are pentodes with the suppressor grid internally connected to one side of the filament.  The first circuit was a simple, single-ended amplifier with one of these tubes wired as a triode.  Encouraged that it (kind of) worked I decided to put together a simple push-pull amplifier for more power.
Figure 2:
Diagram of the push-pull amplifier using 1J18B tubes wired as triodes.  On T1, a single 5 volt winding is
the audio input and the series 120 volt primaries, wired as if for a 240 volt connection, is used as a center-tapped winding
for the 180 degree split to feed the two tubes.  The speaker is connected to the "115" and "125" volt taps of T2.
No serious attempts were made to maximize performance.
Click on the image for a larger version.

Figure 1 (above) depicts the electrical diagram of the amplifier that was literally constructed on the workbench using a lot of clip leads and "floating" components as shown in the pictures.  Because this was a quick "lash-up" I used components that I had kicking around with no real attempt whatsoever to obtain maximum performance.

The audio source for this was my old NexBlack audio player, designed to drive only a standard pair of 32 ohm headphones.  To get some voltage gain and to obtain the 180 degree phase split to provide differential drive to the pair of tubes I fed the audio into one of T1's 5 volt secondaries with the grids connected to the dual 120 volt primaries in series, using the middle as the center-tap to which a "bias cell", a single 1.5 volt AAA cell, was connected to provide some negative voltage.

Even though T1 was a simple split-bobbin dual primary, dual secondary power transformer, it worked reasonably well in the role of audio transformer.  With the 5 volt to 240 volt secondary and primaries, the turns ratio was approximately 1:48 implying a possible impedance transformation of 2304-fold across the entire "secondary".  In this application the actual impedance is not important as it was only the "voltage gain" and the 180 degree phase split that was sought.  In the configuration depicted in the Figure 2 there was more than enough drive available from the audio player to drive the tubes' grids into both cut-off and saturation.

Both V1 and V2 were wired in "triode" configuration with the screen (G2) tied to the plate supply and the audio and operating bias being applied to the first grid.  Because these tubes' filament voltage is specified to be in the range of 0.9 to 1.2 volts, a 4.7 ohm series resistor, R1, was used to drop the filament voltage from NiMH cell B2 to a "safe" value of about a volt.  The plate voltage was provided by five 9-volt batteries in series with a bench supply to yield around 60 volts - the recommended voltage for this particular tube.
Figure 3:
The amplifier, wired up and scattered across the workbench.  The audio
player and T1 are along the left edge, the tubes are in the middle and
the output transformer, speaker and batteries that make up the
plate supply are seen to the right.
Click on the image for a larger version.

In the same spirit as T1 the output transformer was also one designed for AC mains use rather than an audio transformer.  In trying a number of different transformers that could be wired with a center-tap on the highest-voltage winding - including the same type as used for T1- I observed that the highest audio output power was obtained when I used the plate voltage transformer that I'd wound for a (yet to be described) audio amplifier that I'm constructing.  (For an article about the construction of this transformer follow this link).

For T2 this transformer was used "backwards" with the 982 (unloaded) volt center-tapped secondary being connected to the tubes' plates in push-pull configuration.  With a tone generator being used as the audio source I experimented with the various taps and winding combinations and found that the best speaker drive was obtained across the "ten volts" of the 115 and 125 volt taps of the primary.  Based on this configuration the calculated turns ratio is therefore around (982/10) = 98:1 implying an impedance transformation of 9604:1.  With the 8 ohm speaker, the total impedance across the entire winding is therefore calculated to be approximately 77k, or around 19k between the center-tap and each end.  In rummaging around I noted that this particular transformer appeared to have the largest turns ratio of any that I had on-hand!

Perhaps due to the "open" construction and flying leads and/or the lack of any swamping/terminating resistance on the grid side of T1 I noted on the oscilloscope some high-frequency oscillation on the audio output which was easily quashed with the addition of 100pF capacitors C1 and C2 on the grids of the tubes.  The addition of C3 as a power supply bypass had a very minor affect, slightly improving the amplifier performance as well - such as it was!
Figure 4:
A close up of the two tubes, flying leads, C1 and C2 and filament
battery B2 in the background.
Click on the image for a larger version.

In initial testing bias cell B1 was omitted resulting in a quiescent current of around 6 milliamps with 60 volts on the plates.  Adding this cell  to provide a bit of negative bias lowered this current to around 2.5 milliamps while also improving the output power capability somewhat.  Increasing this bias to about -3 volts (two cells in series) resulted in lower audio output and a noticeable amount of crossover distortion indicating that too much of each audio cycle was occurring where the tube's linearity suffered and/or it was in cut-off.

The audio output power was a whopping 250 milliwatts or so at 1 kHz and approximately 10% distortion while the saturated (clipping) output power was around 550 milliwatts.  Referenced to 1 kHz, the -3dB end-to-end frequency response was approximately 90Hz to 12kHz with a broad 3 dB peak around 6 kHz.  On the "full-range" 6"(15cm) speaker that was used for testing this amount of power was more than loud enough to be heard everywhere in the room and sounded quite good with both speech and music.  If I had used a higher-power "rod" tube like the 1J37B or 1P24B and adjusted the impedance accordingly I could have gotten significantly more output power from this circuit.

While the overall frequency response performance could have been improved somewhat with more appropriate termination of transformer T1, one cannot reasonably expect the use of transformers intended for 50/60 Hz mains frequencies to provide the the best frequency response and flatness - particularly with the high plate impedances of the output.  Having said this, it is worth noting that power transformers such as that used for T1 may not only be used as a driving transformer but it could have also been used as an output transformer in a push-pull configuration, albeit with a lower impedance and audio output power for these particular tubes.  While the performance may not be ideal, these power transformers worked surprisingly well and their price, variety and availability make them suitable candidates for a wide variety of applications!

After satisfying my immediate curiosity about these tubes for the moment I un-clipped the flying leads, unsoldered the capacitors and resistors and put the parts away.  Some time in the future I'll put together a few more "fun" projects using these interesting tubes.

[End]

This page stolen from "ka7oei.blogspot.com".