Tuesday, November 2, 2010

Improving the Heathkit HR-10B Receiver

UPDATE:  24 April 2012.  Just updated the schematic, which had an error (the source pin of the MPF102 oscillator should connect directly to T5 pin 2).  The new schematic is Rev. 2 (the old schematic was Rev. 1).

The Heathkit HR-10B is a 5-band, 7-tube amateur radio receiver manufactured from 1967 to 1975 and the companion to Heathkit's DX-60B transmitter. Essentially, the HR-10B design is the same as its predecessor, the HR-10 -- the only change seems to be that the top cover was painted with a wrinkly finish rather than the smooth finish of the original HR-10. It requires an external speaker or headphones.



I've always like the way the Heathkit HR-10 series receivers looked with their functional control layout and slide-rule dial. I picked this one up for a reasonable price, and I thought I'd give it a try. Powering it up, I immediately noticed a number of problems:
  • Slide Rule Dial tracking diverged greatly on 80 meters.
  • Receive frequency changed significantly as "RF Gain" was varied.
  • Couldn't use AGC (AVC) for SSB/CW modes.
  • Broad Selectivity.
  • Deaf on 15 and 10 meters.
From my internet research, these problems seemed to be common. Not a very good receiver, and it got me wondering...was there anything I could do to improve it?

Slide Rule Dial not tracking on 80 Meters.

If I calibrated the HF oscillator at the 3.5 MHz mark on the dial and then tuned up in frequency , I found that as I tuned towards 4 MHz, I would hear a 4 MHz signal at about the 3.96 MHz dial mark. In other words, as I tuned through the band the oscillator diverged significantly from the scale markings.

The tuning capacitor, when the dial is at 3.5 MHz, is at maximum capacitance. The fact that I'm receiving a 4 MHz signal at the 3.96 MHz dial tick tells me that the capacitance of the tuning capacitor has decreased too much as I rotated the dial.

One way to fix this is to add additional parallel capacitance to the 80-meter oscillator tank circuit so that , as the variable capacitor is tuned, the the overall "delta" in capacitance is reduced. I found that, for the amount of divergence that I was experiencing, paralleling C30/C66 with a 6 pf Silver Mica capacitor brought the dial into close enough calibration for my purposes. (There is a bit of divergence at around 3.6 MHz, but there's nothing I can do about that).


Receive Frequency shifts with changes in RF gain.

This seems to be a common problem with the HR-10 series receiver, and dynamic variation of the "125" plate-voltage line (i.e. the junction of R44, R43, and C56 in the schematic) seems to be the source of the problem. This voltage is generated by dropping the DC from the cathode of the rectifier through a series 1500 ohm, 10 watt resistor. Thus, because any change in RF gain changes plate current, plate voltage will also change because the the voltage drop through the 1500 ohm resistor has changed.

Unfortunately, as plate voltage varies, so does the frequency of the oscillator(s).

One way to fix the frequency shifting is to stabilize (regulate) the plate voltage. I added a series-string of four 5 watt Zener Diodes from the R44, R43, and C56 junction to ground (there's also a series 10-ohm resistor so that I can measure current, and thus power-dissipation, through the zeners). These diodes consist of three 33V, 5 watt diodes and one 18V diode, for a total voltage of 117 volts (prior to adding the zeners this node measured 144 VDC instead of the spec'd 125 volts, so there's headroom). Power dissipation in the 33V diodes measured to be about 1.2 watts apiece, so there's plenty of margin, dissipation-wise.

These diodes are shown at the top of the schematic below.

(Click on schematic to enlarge)

Note: Reference Designators in the schematic reference the original Heathkit parts. Parts without reference designators are new parts.


AGC (AVC) for SSB and CW

The HR-10B suffers from the standard problem with receivers designed pre-SSB: the AGC is worthless for SSB. Instead the user is advised to turn off the AVC, set the AF Gain to 3 o'clock (i.e. HIGH!), and then adjust the RF gain for an appropriate signal level.

In other words, there is no AGC for SSB or CW!

Before getting further into modifications, let's first take a step back and try to understand why this is...

Vintage receivers (prior to the days of product-detectors) typically used their diode-detector to detect both AM as well as SSB/CW detection. In SSB/CW mode this detector is driven by the output of a basic heterodyne mixer in which this output contains the BFO signal that it is driven with, as well as the beat products. These beat products form an envelope on the output waveform which is detected by the diode-detector. There are several problems with this method of demodulation for SSB and CW.

First, because the SSB or CW signal is demodulated with an envelope detector, the BFO signal must be quite a bit larger than the IF signal if there is to be minimal distortion on either CW or SSB. You can get an idea of why this is so by looking at the image below and comparing the envelopes of the two waveforms (Es = Eo and Es = 0.5Eo).


(Click on image to enlarge)
(Terman, F. E. Radio Engineers' Handbook, First Ed., McGraw-Hill Book Co., 1943, Page 567)

Imagine that Es is the IF signal representing a CW signal and Eo is the oscillator. If the amplitude of Es is significantly less than Eo, then the envelope on the resultant mixed waveform looks close to a sine-wave (look at the envelope of the Es = 0.5 Eo signal). And because this envelope is detected with the diode-detector, it will sound fairly undistorted.

But as the amplitude of Es approaches that of Eo, the envelope becomes much more distorted (look at the envelope of the Es = Eo signal), and thus the resultant detected output will be full of harmonics and sound grossly distorted.

So the IF signal must always be appreciably less than the BFO signal. But...this introduces another problem. Because AVC is also derived from the signal at the output of this mixer (which contains the BFO signal if its on), if the BFO is on this BFO component at the output of the mixer will swamp the AVC circuit and thus severely attenuate the receiver.

For this reason the operator manuals for older receivers state that, when receiving CW (or SSB) signals, the AVC should be turned OFF, the Audio Gain turned UP, and the RF Gain manually adjusted to provide a comfortable signal level. Not very convenient nor friendly to your ears when a very strong signal suddenly pops up nearby, and an excellent reason for adding a product detector and upgrading the AVC circuitry.

So...I decided to update the AGC circuit and at the same time add a product detector in lieu of the original CW detection scheme.

First thing I did was to replace V5, a triple-diode tube (6BJ7) with three 1N4148 diodes. (I had some DC voltage on the AVC line even with no input signal that I attributed to "leakage" in the 6BJ7 tube. Replacing the tube with diodes fixed this problem, and, of course, also lowered power dissipation).

For the AGC circuit I added a 1N4148 diode to change the AGC voltage-doubler configuration from a "Villard" circuit to a "Greinacher" circuit, which has better ripple characteristics. I increased the AGC decay time by paralleling a new 0.22 uF cap with the existing 0.05 uF cap (C29) and moving the location of the 1M resistor (R26) to increase the decay resistance from the original 2.2M ohms to 3.2M ohms.

With the 1M ohm resistor that had controlled the charge-rate moved, I replaced its function with a much smaller 9.1K ohm resistor (this value doesn't need to be exact -- in fact, you can probably get away with just using a jumper in lieu of this resistor).

I used an NE602 for the product detector -- my original goal being to use its on-chip oscillator for the BFO. Unfortunately, when I tried this (using the "stock" HR-10B BFO components) I found that the BFO frequency would "pull" with incoming signal strength (e.g. as RF Gain or AVC varied). I couldn't discover why this was happening, so I worked around it by replacing the on-chip oscillator function with a simple external oscillator using an MPF102 FET, and it worked much better.

If the BFO is on when in AM mode, you can hear it heterdyning with the carrier of the incoming signal, so it's necessary to turn the BFO off when receiving AM signals. To disable the oscillator the low-end of the oscillator tank circuit, T5, is removed from ground using a 2N7000 transistor. To turn the BFO on, this transistor must first be turned on to short pin 1 of T5 to ground.

While experimenting I ended up with quite a bit of attenuation at the input of the NE602 (the capacitive divider). I'm not sure if this much attenuation is needed; I added it because, during my testing I was experiencing some distortion issues and this seemed to help. However, I was making a number of changes around this time, and I could easily have over-compensated. Don't take these values as being the final word -- experiment!

The demodulated output from the NE602 drives two separate paths -- the audio path and the AGC path. I wanted to decouple the audio-path gain from the AGC-path gain (this is an audio-derived AGC circuit) just in case I needed different gains for the two paths. The op-amp inputs are fed via simple low-pass filters (to remove any residual RF from the output of the NE602). Gain of the audio path is about 37 dB, while gain of the AGC path is about 39 dB. This isn't much of a difference, and one could probably use the same op-amp to drive both paths.

The TL082 op-amp has a max power-supply rating of 35 volts (when powered with a single supply). I powered it with 30 volts to ensure that I'd have plenty of headroom when experimenting with gains -- the op-amps are biased at 15 volts, which give them about a +/- 12 volt swing (the TL082 output limits when within (roughly) 2-3 volts of either power-supply rail).

AGC gain is set to give me the same S-meter reading (roughly) when in either AM or SSB mode (BFO Off or On).

Audio gain is set to give the same audio ouput at the speaker (very roughly) when in either AM or SSB mode.

A relay is used to select between AM (no BFO) and SSB (BFO) modes. In AM mode, the HR-10B demodulation and AGC circuitry is the same as the "stock" receiver (with the exception of the changes to the voltage-doubler and RC time-constants described above). A 48V coil for the relay is used to minimize current drain (and thus power dissipation) -- it only draws 4 mA when on.

No changes were made to the Noise Limiter (ANL).

Here's a photo showing where and how I mounted the op-amps and the NE602. You can also see the string of zener diodes I added for oscillator stability near the top of the photo.


Note: Reference Designators in the schematic reference the original Heathkit parts. Parts without reference designators are new parts. And for many of these parts the value isn't critical -- I usually just pulled parts out of the junkbox that were in the ballpark of what I wanted.


Broad Selectivity

The HR-10B receiver has a two crystal crystal- lattice filter spec'd at 3 KHz down at 6 dB at an IF frequency of 1681 KHz.

Although 3 KHz might seem narrow, I've found that the skirts of the filter (on my receiver) are not very steep at all. This gentle roll-off of the filter skirts results in audio that is fairly broad, and, in fact, for AM reception I find that the receiver actually sounds pretty good.

I did find an article in Electric Radio regarding modification of the HR-10B crystal filter (as well as crystal filters in other receivers -- refer to the Electric Radio articles in the "Resource" section below). I decided not to attempt these mods at this time.


Deaf on 15 and 10 Meters

Lack of sensitivity on the high bands is a common complaint for this receiver, and mine is no different. I've poked around at this, and it looks like it's caused by a couple of things.

1. The RF Preamp (V1 and associated circuitry) has appreciably lower gain on 15 and 10 meters.

2. On 15 and 10 meters the HF Oscillator, rather than beating the incoming signal with the fundamental of the oscillator to get the IF frequency, instead beats the incoming signal with the second harmonic of the oscillator frequency. The amplitude of the second harmonic will always be less than that of the fundamental frequency, and, depending upon how the second harmonic is generated, the second harmonic might be significantly lower in amplitude.

Signal level at the output of the mixer is a function of the level of the input oscillator, so a lower-level oscillator signal will result in a lower-level output (and this can be exasperated if there's a square-law (or higher!) function in the mixing process.

I haven't yet looked into improving the performance on 15 and 10, given that there's sure to be a stability issue, too, given that we're using the second harmonic for the conversion. In other words, jitter or drift at the fundamental frequency means twice the jitter or drift at the second harmonic, and thus twice the degradation in stability!


Other Notes:

1. There's an optional Crystal Oscillator (HRA-10-1) which can be plugged in. This is a useful option!

2. Others have mentioned that alignment of the RF/Oscillator section can change when the bottom steel plate is reinstalled after completion of the alignment procedure. It has been recommend that holes be drilled in the bottom plate so that the receiver can be aligned with the plate in-place. I've not yet done this.


Resources:

Electric Radio Magazine Articles:
  • "Resurrection of a Heath HR-10B Receiver," Paschall, Issue 210, Nov. '06
  • "The Heathkit HR-10 Receiver," Hanlon, Issue 232, Sept. '08
  • "Heathkit HR-10 Receiver Update," Stock, Issue 234, Nov. '08
  • "Modifying Heathkit Crystal Filters," Stock, Issue 231, Aug. '08
Search Heath listserve archives here.

HR-10B Schematic here.

HR-10B Modifications here.


Standard Caveats

There might be mistakes. I cannot guarantee that everything is accurate. Use at your own risk!

The HR-10B has high voltages -- use caution whenever working on it!

Tuesday, October 26, 2010

WRL Duo-Bander 84

I came upon a pair of these radios (plus an AC power supply) a few months ago while visiting a ham-radio store that specialized in older, used gear. It's a Duo-Bander 84, manufactured by WRL (World Radio Laboratories, Inc.) sometime in the late 60's.
   
(Click on image to enlarge)

The Duo-bander 84 is designed to be a sideband-only rig for 75 and 40 meters, and thus coverage on 75 meters is 3.8 to 4.0 MHz, and coverage on 40 is 7.1 to 7.3 MHz. 

Tuning-up the transceiver's Transmitter is a single knob operation (rather than the usual peak-the-grid, dip-the-plate, and adjust-loading). Simply insert carrier into the Transmit signal by un-nulling the carrier balance (using the "NULL" control), and then adjust the "TUNE" knob for maximum output on the meter. (The "NULL" control (carrier-balance) will then need to be re-nulled). 

The PA consists of a pair of 6HF5 sweep-tubes. With about 700 volts on the plates (the radio requires a separate power supply, by the way), I find that my peak power out is a bit more than 100 watts.

(A copy of the manual can be found via the link in the Resources section, below.)
   
Top view (click on image to enlarge)

Bottom View (click on image to enlarge)


Bottom view of a later version.


Note the new holes in the chassis under the IF board, and the crystal filter is sans case. I don't know if these changes were done in the factory, or were made later in the field by an end-user. 

Notes: 

1. Adjusting PA bias. Per the manual, one should set the PA bias so that, when the carrier is properly nulled and there is no voice-excitation (and the meter switch on the back panel is set to "Bias"), the meter needle, when transmitting, is on the "bias" calibration mark on the meter scale. 

Unfortunately, one of my two radios had a broken meter. How should I then set its PA bias? 

Well, I wasn't too sure how accurate the meter was on the rig that was operating properly (and note, the meters are not calibrated in mA), so I didn't want to use it as a calibration reference. I did a bit more research and discovered that the Swan 350 also used a pair of 6HF5 tubes in its PA. Their manual is a bit more explicit, and they state that the PA bias should be adjusted to 50 mA idle current when in Transmit mode.  

I decided that if 50 mA is good enough for the Swan, it's probably good for the Duo-Bander, too! So I verified that the PA cathode resistance to ground was 2.5 ohms (there are four 10 ohm resistors in parallel), and adjusted the bias-pot on the back panel so that the PA cathode-to-ground voltage was around 0.125 volts. 

And with the PA Idle Bias adjusted for 50 mA, the meter needle sits at the "Bias" marking on the meter faceplate (when the meter is in BIAS mode)! 

2. No ALC circuit. The Duo-bander 84 does not have an ALC circuit to limit voice-peaks. To prevent excessive overdriving, I prefer, while monitor the output RF waveform with a 'scope, to adjust the mic's gain until the peaks are just at the peak-power out (this point will become evident as you adjust the mic-gain past this point -- the RF peaks will not get any higher, and you'll see more flat-topping). 

I've thought about adding an ALC circuit, but decided that it wasn't worth the effort. For those who are interested in experimenting, a good place to start would be to look at the schematics for the Swan 350, the Galaxy V Mark II, as well as other radios (Heathkit HW-12A, Galaxy GT-550, etc.) that use sweep tubes in their finals. You'll see an ALC circuit that's common to all of these radios and which consists of a pair of diodes used as a negative peak detector to generate a negative ALC voltage based upon the PA grid voltage. 

3. Distortion on transmit audio

Both of my radios exhibited significant distortion on their transmit audio when I was first testing them. On both, I traced the cause back to a bad C9 capacitor. This is a 2 uF, 50V cap that acts as an AC ground for the collector-load of Q4. The balanced-modulator (Q6 and Q7) requires that the audio-drive to it (from Q4) consist of two signals 180 degrees out-of-phase and of equal amplitude. Thus the phase-splitter's (Q4) emitter and collector loads should be identical. If C9 is not a good AC ground for R11, then the amplitudes will not be equal, and there can also be a phase difference between the two that is not equal to 180 degrees. 

On both radios I replaced their C9 caps with 4.7uF, 63V axial electrolytic caps that I had in my junk box, and the distortion problems greatly improved. (Note: it's OK to use a 4.7uF to replace the 2uF in this application. Larger value caps provide a "stiffer" AC ground for the audio signals, due to their lower impedance at audio frequencies). 

Here's C9 in the schematic:
   
(Click on image to enlarge)

4. Excessive Transmit Audio Low Frequency Roll-off. 

Per the alignment instructions, the filter passband can be "shifted" in frequency by adjusting the Carrier Crystal frequency with trimmer C39 (mounted next to the carrier crystal at the back of the radio). I found that, even with the crystal adjusted to shift the filter passband as close to the carrier as I could, I still had excessive roll-off in the low-frequency audio, so much so that it seemed as though the low-frequency cut-off was around 500 Hz. 

Poking around the audio path with a 'scope, I discovered that there was excessive low-frequency roll-off occurring just after the mic-jack coupling capacitor C7 (0.01 uF). I paralleled C7 (0.01 uF) with a 0.1 uF cap (or you could simply replace C7 with a 0.1 uF cap), and this removed the excessive low-frequency roll-off.
   
Bandwidth is now 300 - 3000 Hz. 

5. Carrier Crystal Oscillator stops oscillating in Transmit. 

While I was trying to adjust the Carrier Crystal oscillator frequency to shift the filter passband so that the low-frequency cut-off was around 300 Hz rather than 500 Hz, I discovered that, as I rotated trimmer C39 towards one of its limits, the oscillator would stop oscillating when I was transmitting (but there was no problem in Receive mode). 

As an experiment, I paralleled C15 (150 pf silver mica cap) with a 100 pf silver mica cap, and this seems to have cured the problem. My reasoning for adding this cap is: the oscillator would stop oscillating as its frequency was lowered, and so I assumed this meant that the trimmer cap was approaching its maximum capacitance. I decided to add capacitance across C15 (that is, to the fixed-cap side of the voltage divider formed by C15 and C39) in order to return the capacitance ratio between these two caps back to a value where the oscillator still oscillated. It seems to have worked, but I can't say that this is an optimal solution. Consider it a band-aid which fixed the problem for this particular transceiver.
   
(Note that you can easily add this capacitor by simply mounting it across the coax (from the Carrier oscillator) connected to the two pins at the back of the printed-circuit board, very close to the trimmer cap C39.) 

6. Receive Distortion due to AGC: 

While operating this radio I noticed that, for some signals, there was some subtle distortion on the receive audio. This problem seemed to manifest itself with stations who were using wide-band audio (e.g. lots of low frequencies). Fortunately, most signals I copied didn't seem to have this problem. 

But the distortion was noticeable enough on a couple of stations with whom I talk regularly, and so I decided to look into it. 

I noticed that, per the schematic, there are actually two AGC lines: an "AVC RF" line (to the grid of V6) and an "AVC IF" line (to the grids of V3 & V4). 

The AVC IF signal has a very fast decay time constant (essentially, the decay of C21, a 0.02 uF cap, is controlled by R22, a 33K ohm resistor, which is C21's discharge path into C22, a much larger 0.22 uF cap). 

 AVC RF, on the other hand, has a much slower discharge -- C22's decay is controlled by R23 in series with R24.
   
(Click on image to enlarge)

Looking at the AGC (or AVC, if you prefer) signals with a 'scope, I noticed a potential problem with the AVC IF line. If you look at the top photo below, you'll notice there are a lot of "spikes" on its waveform (as the AGC goes more negative, there is more attenuation). These spikes are voice peaks at C21, and they quickly decay via R22. 

One often sees this sort of AGC response in older receivers. I believe this fast-decay AGC is to limit the gain of fast transients (e.g. static crashes?) yet not have them affect the longer-delay AGC. Unfortunately, in my experience, these sorts of spikes can "modulate" the received signal and produce perceptible receive audio distortion, and often the receive audio will sound better if one can eliminate this sort of AGC modulation. (Of course, recognize that, in doing so, there can be a trade-off with limiting the gain of fast-transient noise). 

For the WRL Duo-Bander 84, one way to clean up this distortion is to add more capacitance to C21 -- that is, make it larger. But rather than add another cap, one can simply short-out R22, the 33K ohm resistor between C22 and C21. This means that the AVC IF line's decay time constant becomes the same as the AVC RF line's time constant, and you can see the result on the AVC IF line in the lower photo, below.
   
(One issue with this sort of mod, though, in which the capacitance is greatly increased, is that the attack time will slow down (because you're charging more capacitance). The Duo-Bander's Receive AGC is audio-derived, and therefore, as with many other receivers with audio-derived AGC, you can get an audible attack "pop" at the start of strong signals. In theory, the mod I've made should exacerbate this type of popping, but I haven't noticed much difference, if any, in attack pops with or without this mod.) 

The mod can be made easily without removing the PCB: To short-out R22, jumper the top of C22 to the top of R37, as shown in the photo below.

 
Again, I want to stress -- the original distortion that I was hearing was very subtle and only manifested itself on a couple of signals that I listened to regularly. For the most part, all other receive signals sounded fine. So this modification is certainly not necessary, and, if you do try it, it's quite possible that you won't notice any difference on the majority, if not all, of the signals you copy. 

. Other problems... Here are some of the other problems I found in my two Duo-Bander radios:
  • Very little gain on receive. I traced this to R46 (47K, 1W plate-load for V7b) reading infinite ohms. Replaced with a 47K, 2W resistor from my junk-box.
  • Very low TX audio. I traced this to a faulty R9 (270K, 1/2 watt) in the Mic Preamp circuit (it read infinite ohms with an ohmmeter). Replaced with same value resistor.
8. Measuring resistance per the Manual's Resistance Chart

While trouble-shooting the rig I discovered that a number of resistances listed in the Resistance Chart were reading infinite ohms, despite their non-infinite values in the table. 

These were typically plate resistors (or other resistors connected to the +400V B+ supply), as well as, for example, cathode-resistors in the TX path (e.g. V7 cathode). The reason why they were reading infinite with my ohmmeter was that they had no path to ground. (Note that the manual specifies its resistance tables in the chart to be "resistance to ground"). 

To make an accurate comparison of all resistances, per the chart, I'd recommend that you actually measure across the specified resistors, or you could ground the appropriate relay pins (e.g. pins 2 and 6 for Receive and Transmit B+, and pin 7 for TX cathodes) and then make your resistance measurements (in either case, though, be sure that all power is first disconnected from the radio -- remove the connector from the power supply to the Transceiver's P1 Power Plug!). 

9. Power Supply: The manual states that the power supply requirements are:
  • HV: 800VDC @ 400 mA
  • Low B+: 325/375 VDC @ 200 mA
  • Neg: -100 VDC @ 30 mA
  • 12VDC @ 200 mA
  • 12VAC (or DC) @ 5 A
Per one of the links in the Resource section below, three power supplies were available for the Duo-Bander 84:
  • Deluxe 400 Watt AC supply AC384A $89.95;
  • 400 Watt DC Supply DC384A $99.95;
  • 250 Watt AC Power supply AC48A $49.95.
Dale, W4OP, adds this note:

"I thought it might be nice to have the WRL supply. What I ended up buying on eBay was a WRL Duo-Power 300. I only found out what it was after arrival. While it has a nice  metal WRL label on it, there is no mention of a model #. W9RAN came to the rescue, identifying the supply. It is a compact AC supply but has  two clasps that allow for  a vibrator add on supply for mobile operation.
Sure didn’t look that big in the photos. Recapped it and it runs as it should. You might add it to your list of power supplies for the Duo-Bander."

In addition, other Galaxy power supplies allegedly can be used, too (I've not personally verified this, though), per this post on the Electric Radio Forum:

 
12/05/02 08:47 PM | 0 Good Guy Alerts Vote Edit Reply | WRL Galaxy Duo-Bander 84 power supply info wtd:
n3ibx New Member 1 Post 0 Good Guy Alerts Washington Crossing N3IBX Ignore User Hello All, Would anyone know if the AC supply from a Galaxy 5 MkII will work with the Duo-Bander 84? Any assistance will be appreciated. Mod-U-Later, Joe Cro N3IBX
Joseph Cro

12/06/02 07:38 PM | 0 Good Guy Alerts Vote Edit Reply | Galaxy Power Supply
Stu New Member 1 Post 0 Good Guy Alerts Elmira NY K4BOV Ignore User Yes Joe,

Any Galaxy power supply except the PSA-300 (came with the Galaxy 300) will run the Duo-bander without modification. The PSA-300 will certainly run a Duo-bander; but, the 12 pin Jones plug must be rewired to conform with the later Galaxy pin configuration.

If you want to send me the serial number of your Duo-Bander, I may have some service bulletins for your particular series. Might want to give me the serial number of the power supply as well. They all work as I indicated; but, some are more capable than others.

Stu/K4BOV


10. Some other radios using 6HF5 tubes in their finals:
  • Swan 350 and Swan 400
  • Drake 2NT
  • Galaxy 300, Galaxy V, and Galaxy V Mark II
  • Galaxy 2000 (Amplifier)
  • Hallicrafters HT-46 and SR-400

Resources:
 

1. Specifications and more information on the Duo-Bander 84 can be found here and here. (Admittedly these are both sketchy). 

2. An Instruction Manual (in PDF format, and including alignment instructions and schematic) for the Duo-Bander 84 can be found here


Standard Caveat: 

 Of course, I may have made a mistake, so use my suggestions at your own risk! Also, this radio uses high-voltages that can kill you. Always use caution when working on a radio of this type.

Saturday, August 21, 2010

813 AM Transmitter Accessory: External PTT Control

My 813-based AM transmitter (see posts here, here, and here) sits about 8 feet behind me in the shack. I wanted to put the TX/RX antenna relay near my operating position (rather than at the transmitter which would have meant running an extra length of coax over to it), and I also wanted a conveniently-placed TRANSMIT switch next to my operating position so that I could easily put the transmitter into XMIT mode.

I also wanted to use my Flex-5000 transceiver as the receiver because it has, to my ears, a very good Synchronous AM detector and sounds great on AM. But I needed a way to "Mute" the 5000 via external control (sure, I could use my PC's mouse each time and click on the Console's "MUT" button, but what a pain -- I wanted automatic control). Unfortunately, the 5000 lacks an input that can be used as an external mute control for the receiver. But all was not lost ...

On the 5000's back panel there is an RCA connector for an external PTT input. I normally use this to place the 5000 into XMIT mode. But why not have it serve a dual purpose? That is, why not let it act as either a PTT control (its normal function) or a Receiver MUTE control?

I modified the Flex Console code so that the 5000's PTT RCA jack can be used for either of these two functions. And to select which function this RCA jack would perform, I added a new "button" to the Flex Console.

Let's start first with the circuitry. Here's the schematic:

(Click on image to enlarge)
Design notes:
  1. Everything came out of my junkbox, which is why you'll see a 26 volt relay (the Antenna Relay) mixed with 12 V relays, and powered from a 24VAC transformer. I used what was at hand.
  2. J7 connects to the PTT RCA on the 5000's back panel, and can either be used to place the 5000 into XMIT (via a pushbutton attached to J6), or as a 5000 RECEIVER MUTE control when I'm transmitting with the 813 AM transmitter.
  3. J5 connects to the first set of contacts in the 813 Transmitter Sequencer (JP5 pins 1 & 2, per page 3 of the 813 AM Transmitter schematic). This keys the Antenna Relay as well as the 5000 Receiver Mute.
  4. The military Antenna Relay internally shorts the unused RF port, as I've shown in the schematic.
  5. Switch SW2 is a toggle switch and, when ON, it places the AM Transmitter into XMIT mode.
  6. Originally relay K3 did not exist and switch SW2 connected directly to the AM Transmitter's PTT input (via jack J4). I quickly discovered, though, that if I toggled SW2 ON to transmit and if I'd forgotten to turn on power to my External PTT circuitry (that is, the circuit above), the Antenna Relay wouldn't switch-over during XMIT, and I'd transmit into a short. Not good! So relay K3 was added as a safety interlock. If power isn't ON, relay K3 is OFF, and switch SW2 cannot place the AM Transmitter into XMIT mode. SW2 can only act as a PTT switch if power is ON.
  7. Normally I use the AM transmitter with my Flex-5000 as the receiver, but sometimes I use my venerable Collins 75A-4.  Relay K4 mutes the 75A-4 during TX.

Here's the finished box (shown in Receive mode):



And here's a snapshot showing the new "button" I added to the Flex Console.


If this button is depressed (as it's shown above), the 5000's PTT RCA acts as a Receiver MUTE input. The software (modified by me) MUTES the 5000 and this button will turn RED whenever the PTT RCA is shorted -- that is, whenever the AM Transmitter is transmitting.

If this button is not depressed, the PTT RCA operates normally, that is, as an input for an external PTT control. In this mode, my external PTT push-button puts the Flex-5000 into transmit.


Caveats!

Standard warning applies: I may have made mistakes when writing this post or in my design. I cannot guarantee that everything is correct. Use at your own risk.

Wednesday, August 4, 2010

AM Transmitter, 813 Style, Part 3 (Everything Else!)

This post (Part 3 in a three-part series) describes the 813 transmitter's power supplies, the control circuitry, and the meters. (Part 1 is here and Part 2 is here). 

 First, let's start with the High Voltage Supply schematic:
   
(Click on schematic to enlarge)
Notes regarding the above schematic:

1. The high-voltage supply is a standard full-wave rectifier into a capacitive-input filter. The caps are rated at 350 volts each, and so eight caps in series handle the high-voltage (overall the capacitance is 275 uF at 2800 volts). The high-voltage is equally divided across each cap with a 50K ohm resistor across each. 

 2. Additionally, there's a bleeder-resistor chain totaling roughly 40K ohms that can be connected across the entire capacitance bank to more quickly bleed off the high voltage whenever power is turned OFF. When the transmitter power is ON, this bleeder resistance is disconnected, so that power isn't wasted. 

 3. There is a 25 ohm power resistor to limit input-current surge into the transformer primary when the HV supply is first turned on. This resistor is then shorted-out when the HV reaches a certain voltage (refer to the Control circuitry schematic below). 

 4. A set of relay contacts is in series with the HV transformer primary. These contacts serve two purposes. First, when power is first turned (to the low-voltage supplies and to the filaments of the 813s), there's a delay of about 1 minute before the HV will turn on via these contacts. Second, if there's an over-current condition in either the PA or the Modulator, this relay opens to remove AC power from the HV supply. The STARTUP neon lamp signals when either of these two conditions is true (that is, when there's no AC across the HV transformer primary). 

5. The HV transformer (I picked up in a trade) has multiple output taps and its input can be wired for either 120 or 240 VAC. I'm using its 3000V output taps and its input taps are wired for 240 VAC (but actually connected to 120 VAC). This gives me, with the capacitive-input filter, an idle plate voltage of about 2300 VDC. 

 6. Five 1KV diodes series-connected are used in each side of the full-wave rectifier. They don't have parallel resistors or caps to "equalize" their voltages because I'd read (some time ago), that modern diodes don't need this sort of protection. But I could be wrong -- you may want to add these. 

 7. Also, an unloaded HV of 2300 VDC implies that each diode string is seeing about 4600 VDC PRV (because the voltage across the capacitance string should be equal to the peak-voltage, given that it's a capacitive-input filter). This is getting quite close to the 5KV rating of the string; for a bit more margin I would recommend adding another diode in series on each side of the full-wave rectifier, so that each string has 6, not 5, diodes. (Under load (i.e. transmitting), the PRV seen by the diode strings drops.) 


Here is the schematic for the Low Voltage Supplies:

(Click on schematic to enlarge)
This page of the schematics shows the PA grid-bias power supply, the PA screen power supply, the modulation transformer, and a voltage monitoring port.

Notes regarding the above schematic: 

1. The PA grid-bias power supply uses series-pass regulation. When the transmitter is idle, this voltage is about -240 VDC. While transmitting, the zener-diode string voltage-reference is connected to the PA Power Return, and the grid-bias voltage goes to about -170 VDC. Power resistors R93-R96 are used to share power dissipation with transistors Q6 and Q7 so that the transistors don't dissipate the brunt of the power when they are sourcing current. Depending upon the desired grid-bias voltage, different resistors in the R93 - R96 chain should be used (and the others shorted-out). For a grid-bias voltage of about -170 VDC, I've shorted-out R93 and R94. So, assuming a 25 mA total grid current, this means that about 39 volts will be dropped across R95 & R96 (about 1 watt, total dissipation) and 30 volts across Q6 and Q7. Assuming they share current equally, Q6 and Q7 will each dissipate about 0.4 watts. 

 2. There is a 40 VDC supply used for powering the Control circuitry as well as various relays. 

 3. The Screen-voltage supply is another bridge-rectifier circuit with capacitive-input filter. Its output voltage can be adjusted using the Variac (T3) connected to the primary of the transformer. Low-voltage windings of the transformer (T4) provide 7 VAC (roughly) to power various lamps (rather than loading-down the 40 VDC supply with their power requirements). 

4. Screen-voltage can be switched between either 0 volts or 300-400 VDC with switch S10. Zero volts results in a lower power output from the PA and is useful when tuning the transmitter. 

5. There are a pair of banana jacks on the front panel of the power supply so that internal voltages can be monitored with an external DVM or scope. Many of these voltages are scaled down to keep them under 50 VDC (for safety reasons). Voltage measurements are:
  1. PA Control Grid Voltage (Tube 1) /10
  2. PA Control Grid Voltage (Tube 2) /10
  3. HV/100
  4. PA Power Return (PA Plate current * 10 ohms)
  5. Modulator Power Return (Mod Plate current * 10 ohms)
  6. 24 VDC
  7. PA Screen Voltage /10

Here is the schematic for the Control Circuitry:

(Click on schematic to enlarge)
This page contains the Sequencer and its associated relays, as well as control circuitry to:  One-minute time-delay, HV Transformer input-current surge protection, and HV Power Fault detection. 

Notes regarding the above schematic: 

 1. The sequencer is based upon a W2DRZ design and uses a bi-directional shift register to "nest" the relays such that the first one ON is the last one OFF. The clock-rate at which the sequencer marches from one relay to the next is controlled by the potentiometer R35. 

2. U5D prevents the HV supply from being turned on for about 60 seconds after the low-voltage power supplies are turned on (e.g. the filaments for the 813s, whose filament transformers are on the same AC line as the low-voltage supplies) by not driving the HV_FAULT signal low during that time. (I don't know if one actually needs this type of protection for 813 tubes, but it was simple to add, and so I thought, "Why not?"). 

 3. If either the PA or the Modulator plate-current exceeds 600 mA (over-current), either U5A or U5B will detect this and release their relay, which then turns OFF the relay whose contacts are in series with the HV transformer primary, removing 120VAC from this transformer. Also, the appropriate FAULT lamp (PA FAULT or MOD FAULT) illuminates when either of these fault conditions occur. If an over-current fault is triggered, this fault condition latches ON and 120 VAC to the HV transformer primary cannot be reconnected until the fault is first cleared by pressing the CLR FAULTS button on the front panel. In other words, 120 VAC is only applied to the HV transformer if: 60-second-warmup-finished AND no-PA-overcurrent AND no-MOD-overcurrent. If any of these three conditions is false, then the relay that connects the primary of the HV transformer to 120 VAC will not turn ON. 

4. When AC power is first applied to the HV transformer primary, surge current into the primary is limited by a 25 ohm power resistor in series with the primary. However, as the HV approaches its operating level, this resistor needs to be shorted-out so that there is no current-limiting into the transformer. U5C detects when the HV voltage reaches an appropriate point and drives a relay to short-out the resistor. (Currently, this trip-point is when the HV supply reaches 1000 VDC). 

5. The /BLANKING BNC connector, J22, isn't used. I'd originally planned to connect a Tektronix 604 Display Monitor (essentially just an X-Y CRT) to use as a trapezoidal waveform monitor, and the /BLANKING signal would blank the CRT when in Receive mode, but I scrapped this idea when I discovered that I'd need to move my "Audio Sample" port (I describe this issue in more detail in the "Modulator Deck" post below). 

 6. Most of the lamps (except for the Fault lamps and the Startup lamp) are 6 volt lamps run from about 7VAC (series-resistors bring the voltage across the lamps down to about 6 volts). The 7 VAC comes from some low-voltage windings on the Screen-voltage transformer, which I use for the lamps so that I wouldn't unnecessarily burden the 40VDC supply with the lamps' current requirements. (The down-side of using these windings is that the brightness of these bulbs will vary depending upon the setting of the Variac used to set the Screen Voltage, but I'm willing to accept this compromise.) 

 The Fault lamps are 28 volt lamps because the signals that drive them are also used to "latch" the fault condition, and my latching design requires a DC "high" signal be fed back into the ULN2003A. I use the 40VDC supply for this, dropped-down appropriately with resistors (~28 VDC for the lamps, and ~20 VDC for the two inputs of the ULN2003A). 


 And here is the schematic for the meters:
 
(Click on schematic to enlarge)
The meters should be self-explanatory. I used whatever I had at hand that had the styling I wanted and that also had scales that did not need to be redrawn. For example, I used 0-50 mA meters to measure current in the 0-500 mA range. 

 The "Modulator Plate Current" meter can also be used to monitor RF Current (0-5 Amps, so that output power can be monitored). This feature is described more fully in the Addendum section below... 


 And finally, here's the Wiring Diagram showing the interconnection between the decks of the transmitter and with external equipment:
   
(Click on image to enlarge)

Here are some photos of the construction:


First, start with a THICK piece of sheet-metal and then...

...add some wood bracing.

Due to space constraints in the area designated for power supply circuitry, I mounted the HV bleeder resistors, AC surge protection resistor, and associated relays on top of the transformer:

 

(Heavy Metal!)


Building the PA Screen and Grid supplies, and what will be the sequencer.


The back plate, showing all of the I/O connectors.


Wiring it up!


With Modulation Transformer mounted.


After mounting the heavy bits, a homemade dolly helps moving it around the shack.


Meters, left to right: Mod Ip, PA Ig, PA Iscrn, PA Ip, HV


PA Grid Current, 0-50 Miles-Per-Hour!


The finished transmitter, in its rack and on-the-air!


Miscellaneous Notes on the Power Supply: 

1. I wanted a Power Supply deck that I could slide in and out of the rack in case I needed to work on it, which is why I made the floor dolly and why I also added three handles to the power supply (one handle on the back of the plate to let me tilt the power supply deck from the dolly onto the bottom of the rack, and the two in front to let me push it into, or pull it out of, the rack). 

Note: the two heavy transformers were installed after I'd put the deck on the dolly! 


Addendum...RF Current Measurement:

With the transmitter buttoned-up in its rack and the rack in the corner of the room, there was no good way to measure RF Output power during tune-up of the transmitter and Matchbox, given my shack configuration. After trying to tune up the transmitter a few times, I discovered I really wanted a meter right at the transmitter that I could watch while adjusting, say, the transmitter's LOADING control. 

Because the Modulator Plate Current meter (the left-hand meter of the five) is of limited usefulness, I decided to make it a dual-function meter. That is, why not add a switch and some circuitry so that it could read either Modulator Plate Current (0-500 mA) or RF Output Current (0-5A)? Converting from RF Current to power is a simple conversion: assuming the tuner is tuned for an SWR near 1:1, then Pout = I*I*50. For example, a current of 2.24 amps equals 250 watts. 

In my junk box I had an old Heathkit HM-2102 SWR meter from which I'd previously pulled out the meter (to replace a blown meter in an SB-220 amplifier: see previous posts) -- its RF box with its two SO-239 connectors would be perfect for the RF current transformer and rectification circuitry. 

The circuit is straightforward and very similar to a design in recent ARRL handbooks (in their "Station Setup and Accessory Projects" chapter) -- see the schematic for the Meter Panel, above. I used 150 ohms in lieu of 50 ohms at the transformer's secondary so that I'd have enough voltage to drive the meter, and the 1:40 turns-ratio means that this load of 150 ohms is equivalent to an insertion of an additional 0.1 ohms in series with the RF line. In other words, loading effects are negligible. 

 Calibration was done against an LP-100 power meter at both 50 watts (1 A RF current) and 100 watts (1.41 A RF Current) into 50 ohms.


RF Current Sampler


2.3 Amps equals 264 watts.

How does it sound on the air? 

 You can listen to a clip of the 813 Transmitter on W6THW's website here. It's the track labeled "K6JCA 813 RIG (AM)". 

 (The rig was putting out about 300 watts, carrier power. Mic is a Heil PR-40 run through a Beringer 802 Mixer/EQ box, which feeds the Johnson Ranger's microphone input.) References: Sequencer Designs: RF Current Sampler:
  • Refer to the High-Power Directional Coupler (that immediately follows the "Tandem Match" project) in recent ARRL Handbooks (e.g. page 22.42 in the 1997 edition of the ARRL Handbook).
Caveats! Standard warnings apply: First, I may have made mistakes when writing this post or in my design. I cannot guarantee everything is correct. Second (and most importantly), this design uses high voltages that can kill you. Be cautious and BEWARE!

AM Transmitter, 813 Style, Part 2 (Modulator Deck)

This post (Part 2 in a series of three parts) describes the modulator stage of my 813-based 75-meter AM transmitter. (Part 1 is here and Part 3 is here). 

The modulator uses a pair of 813s wired as triodes and in a push-pull configuration. It's driven by an external audio driver (in this case, the audio from a Johnson Ranger). The modulation transformer is mounted on the Power Supply Deck, and can be found in the schematics posted in Part 3 of this design.
   
(Click on image to enlarge)

Here are some pictures.
   

Using a scrap chassis from the junkbox. Just a few extra holes!


Notes on the Modulator Deck 

1. John Staples' "Electric Radio" article (issue #57, January, 1994) mentioned the use of a small amount of negative voltage (around -1.5 volts) as grid bias for the modulator tubes. I had originally designed a DC supply into the Modulator Deck to provide some small amount of negative voltage for biasing these grids, but, during testing, I discovered that it wasn't very "stiff" and would fluctuate significantly with voice peaks. 

A quick test revealed, though, that 0 volts bias actually worked pretty well, and it had the added advantage that I could eliminate all of the bias supply components! 

I left the ability to add a negative bias supply externally, though. An external bias supply can be connected between pins 6 and 7 of the Octal Jack on the back panel of the Modulator Chassis. If no external bias supply is used, then the bias must be set to 0 volts by shorting out these two pins.  

(Note: with 1800 VDC plate voltage and 0 volts grid bias, idle plate current is about 23 mA through each 813 (connected as triodes), which equates to about 45 watts plate dissipation per tube. 

2. I'm not sure if the suppressor grids of the 813s should be connected to ground or connected to the other two grids in the tube when using the 813's as triodes. I followed W6BM's example (and for which he plotted his curves) and connected all three grids together, as mentioned in his article and per his schematics which he kindly sent to me (these were not published in Electric Radio). 

But I noticed that K1JJ (see the K7JEB website below), as well as W7XXX in his Electric Radio article (ER # 125), connected the suppressor grids to ground. The 1959 Edition of the Radio Handbook discusses "Zero Bias Tetrode Modulators" (section 30-8) and also shows a pair of 813's in push-pull with the suppressor grids grounded and the control grid and screen grid connected together. And W7XXX, in his Electric Radio article, alludes to potential stability issues (and the modulator becoming an unwelcome generator of RF), but unfortunately he doesn't provide any references. 

(Digging further, I found a discussion on the AMFONE forum (here), and some mention of stability and hi-mu versus low-mu configurations, but again, nothing that I, as an engineer, would consider definitive.) 

So -- is there anything wrong with connecting the suppressor grid to the other grids? Frankly, I don't know, and my research hasn't yet revealed any adequate explanations as to why this approach might be bad. But at least John, W6BM, plotted his tube characteristics using this connection configuration: so there is some data associated with it, and I decided to follow his approach. So far there hasn't been a problem... 

3. When I first designed the Modulator Deck, I didn't have resistors in series with the grids of the two 813s, and I discovered during testing that modulator plate current would suddenly sky-rocket, tripping my over-current protection circuitry in the power supply. Adding 100 ohm resistors in series with the grids of each tube calmed them down. (Note: W6BM used 56 ohm resistors, but I didn't have this value in my junkbox, and 100 ohms seems to work fine.) 

4. The 2K ohm resistors across each grid are supposed to provide a more constant load for the Ranger driving this Modulator Deck, per John, W6BM. John used 1.6K ohm resistors, but I had 2K's in the junkbox, so in they went instead. Are they really needed? I don't know -- a distortion test made with and without these resistors would answer that question. 

5. I had originally added the transformer-coupled "audio sample" so that I could do trapezoidal monitoring of the transmitter's performance, and my thought was that sampling the audio prior to the modulator would be best, because then I could see if there were any non-linearities introduced by the modulator/mod-transformer. Well, it was a nice idea in concept, but it doesn't work in practice. Because of the time-delay through the tubes, you really must sample the audio at the output of the modulation transformer. Otherwise, you get a "phase distorted" trapezoid. 

Old ARRL handbooks have photographs showing this type of distortion on the trapezoid waveform. For example, from the "Amplitude Modulation" chapter of the 1955 ARRL handbook:
   
Anyway -- rather than add another audio sampling circuit at the modulation transformer (and its high voltages), I decided that it would be sufficient to just monitor the RF itself using my already-existing "RF Sample" port on the PA Deck, and simply adjust the audio gain by look for "zero-lining" on the RF waveform -- after all, this was how I monitored the performance of my other AM transmitters, and it seems to work well. So my "Audio Sample" port on the Modulator Deck really isn't needed. 

6. I added some 1 ohm resistors and test points (i.e. feed-thru caps) to allow me to measure the cathode current of each 813 independently, as well as overall grid current. Measuring each tube's cathode current allowed me to easily find a matched-pair of 813s. 

7. For experimenting with grid-bias voltages, an external DC supply could be connected between pins 6 and 7 on the octal plug. For 0-volts grid bias, these two pins should be shorted together. 

8. Although I grounded the bases of the 813 tubes in the PA Deck (using fingerstock), I did not ground the bases of the 813s in the modulator deck. 

How does it sound on the air? You can listen to a clip of the 813 Transmitter on W6THW's website here. It's the track labeled "K6JCA 813 RIG (AM)". (The rig was putting out about 300 watts, carrier power. Mic is a Heil PR-40 run through a Beringer 802 Mixer/EQ box, which feeds the Johnson Ranger's microphone input.) 

References 

 Articles:
  • "A Modern One Kilowatt AM Transmitter," W6BM, Electric Radio, #15, July, 1990
  • "813 Triodes as Modulators," W6BM, Electric Radio, #57, January, 1994
  • "Triple X 813 Homebrew Transmitter, Part Two," W7XXX, Electric Radio, #125, Sept., 1999
  • "Zero Bias Tetrode Modulators," Radio Handbook, 1959 Edition, Editors and Engineers, page 662
  • "Checking Transmitter Performance," ARRL Handbook, 1955 Edition, ARRL, pages 271-273
Websites, Modulator with 813s as Triodes:

Caveats!
 

 Standard warnings apply: First, I may have made mistakes when writing this post or in my design. I cannot guarantee everything is correct. Second (and most importantly), this design uses high voltages that can kill you. Be cautious and BEWARE!

Tuesday, June 1, 2010

AM Transmitter, 813 Style, Part 1 (PA Deck)

(This is the first part of a three-part series. Parts 2 and 3 can be found here and here.)

[Note: I've changed the circuit slightly from my original publication in this post. Refer to the 19 August 2010 Addendum below.] 

Some time ago I toured the shack of a friend, W7MS (Mike), in Reno, Nevada. I was very impressed by his collection of boatanchor equipment, but I was especially wowed by his RCA BTA-250M Broadcast Transmitter that he'd converted to 75 meter operation. 

RCA's BTA-250M was designed to generate 250 watts carrier output using a pair of 813s modulated by another pair of 813s. Mike had done a great job of restoring his radio, and the four 813s, lit up side-by-side, were beautiful. 

After I left I began thinking...I had a box full of 813s up in the attic. I wonder if... 

Well, skipping ahead...about half a year later I finished constructing my 75 meter AM transmitter. Like the BTA-250M that inspired it, it too uses four 813 tubes: two in the PA and two in the modulator. It's designed to be driven by an external audio and RF source, and I use a Johnson Ranger to drive mine (identical to how John Staples, W6BM, drives his 813 transmitter, as described in Electric Radio, issue 15). 

 My transmitter generates output carrier RF power in the range of 200-350 watts. Here's the schematic of the PA Deck :

 
(Click on image to enlarge)

Notes on the schematic: 

1. A large part of the design is based upon the 80-meter 813 amplifier described in "One-band Kilowatt Amplifiers," which can be found in the 1961 - 1968 editions of the ARRL Handbook. I designed a different pi-network using the equation in the Wingfield equations (reference recent ARRL Handbooks). 

2. Per the original "One-band Kilowatt Amplifiers" article, the amplifier doesn't require any neutralization on 80 meters, so none was added. 

3. There's a two-pole, three-throw rotary switch that's used to select screen-grid current monitoring (either the left tube, the right tube, or both tubes together). Monitoring screen current independently allows (allegedly) for tube matching. 

4. Originally, I didn't have parasitic suppressors in the plate circuits of the 813s, but, when I first started testing the deck, I was seeing a lot of high-frequency stuff on the 'scope I'd connected to the "RF Sample" Output BNC, and I thought that this might be parasitics, so I added the two plate suppressors. They changed nothing, and I later discovered (using my spectrum analyzer) that the high frequency crud was all harmonically related to the fundamental -- that is, it's the natural byproduct of a Class-C amplifier, and that there were no parasitic oscillations. I decided to leave the plate suppressors in (out of laziness), rather than remove them, but, per the "One-band Kilowatt Amplifiers" article, they shouldn't be needed on 80 meters. 

5. Given the high-frequency harmonic components that I was seeing on my spectrum analyzer (up to and beyond 200 MHz), I built a metal cage around the entire PA deck to minimize unwanted EMI radiation. 

6. The input network is the same as the one described in "One-band Kilowatt Amplifiers." C38 was changed from 0.001uF to 0.01uf to give a bit stiffer connection of the input network to ground (because there's no neutralization required, this capacitor doesn't need to remain 0.001uF that was used in the original article). 

7. The pi-network's inductor is a three-inch long piece of air-inductor stock that I had in my junkbox (2.5" diameter, 6.7 tpi, 12 gauge wire). This length gives a max inductance of about 15 uH, but I tap it at around 10.8 uH. 

8. To design the Pi-Network I first calculated the load that I needed to present to the plate using the equations for Class C RF Power Amplifiers found in the RF Vacuum Tube Amplifiers section of older editions of the "Radio Handbook," published by Editors and Engineers. (For my calculation I used 350 watts out (carrier) at a B+ level of 1650 VDC. This gave me a Plate Load (RL) of about 2600 ohms.) 

Then, given this load and the desired Q (Q should be in the range of 10 - 20; I chose 12), I used the Wingfield equations from the ARRL Handbook to calculate Pi-Network components. I put all of these equations into an Excel spreadsheet to allow easy manipulation and experimentation "on paper." 

(Note: equation nomenclature changed in later editions of the Editors and Engineers "Radio Handbook" from that used in earlier editions, and I believe an error crept into the text. The best way to determine if an edition is in error is to compare the variable being solved-for in the description of the Class-C calculation steps (particularly steps 6 and 7) against the variable being solved for in the same steps of the "Sample Calculation" that follows this description. For example, the 18th edition of the book, the terms ebmin and epmin are swapped between their use in the description of the equations and their use in the "Sample Calculation" which follows the description. If you're putting your equations into something like Excel, watch out, or you'll have a problem! I fixed this by assuming the terms ebmin and epmin were correctly used in the "Sample Calculation," and so I swapped them instead when they were first mentioned in the prior description of the calculations.) 

19 August 2010 ADDENDUM: 

I noticed that, while transmitting, the RF power output would slowly increase from 250 watts (my initial setting) to 300 watts over a period of about 3 minutes of continuous transmitting. And this effect would reoccur after I let the transmitter idle for awhile (i.e. cooling down) and then began transmitting again. 

In other words, it acted suspiciously as though the Pi-network's "loading" setting was changing with heat (its capacitance decreasing). So I pulled the RF Deck out of the rack for some bench testing. 

If heating were an issue, as a quick test I transmitted for a few minutes, then powered-down (letting the HV decay to 0 volts!) and felt various components in the RF deck. Most felt OK, temperature-wise, but one capacitor, a 500 pf, 20 KV cap that I had placed in parallel with the "LOADING" variable-capacitor, was suspiciously warm. Hmmm...could this be the culprit? 

I was using two sections (out of three) of the loading variable-cap (both sections connected in parallel). I wired in the third section of this cap (giving me 1800 pf max instead of 1200 pf) and removed the fixed 500 pf HV cap that was in parallel with the loading cap. 

Powered back up, tuned the transmitter for 250 watts carrier output power, and after three minutes...it was still 250 watts! Problem fixed! Apparently the cap was lossy and, with the tank-circuit currents, it was heating-up and changing its capacitance. 

I was a bit concerned that, with the three sections of the variable-cap wired in parallel, adjusting the LOADING control for a desired power might be a bit touchy because the capacitance might change too quickly as I turned the knob, but it's actually quite acceptable (admittedly, I have BIG KNOBS on my controls, which help when making fine adjustments). In this new circuit configuration, LOADING adjusts power from a min of about 180 watts to a max of about 370 watts RF output (carrier only, as measured using a Bird 50 ohm dummy load). I typically run the power at 300 watts carrier output. 

The PA Deck schematic page (above) is now labeled "Rev. 2", to differentiate it from the original Rev. 1. The changes incorporated into Rev. 2, are:
  1. Delete C81 (500pf, 20KV fixed cap).
  2. Change C30 from a 1200 pf max variable cap to an 1800 pf max variable cap.
(Note: Other schematic pages are still at Rev. 1).
   
The culprit!

Fixed cap removed and 3 sections of variable-cap wired in parallel.

And here's the finished transmitter, up and running!

Some additional photographs showing construction of the PA Deck...
   


Building an RF "cage" around the PA using scrap sheet metal I purchased and had cut-to-size at a local metals recycling place.

 
I grounded the metal base of each 813 in the PA section at two different spots for each tube using flexible "fingerstock." (The fingerstock flexes out of the way whenever a tube is inserted or removed). I don't know if this is necessary, but I recall reading about it somewhere (can't recall where, though, at the moment).
    
The angled piece of black material between the tubes and the front panel is actually a rectangle of PCB material that I painted black and stuck into the PA Deck to deflect the fan's air up and out through the screen material on top of the case. 

 The finished PA Deck:
 
Other Notes: 

1. The plate voltage when not transmitting is about 2300 volts DC, but it sags down to around 1800 at 250 watts out. This sag is probably due to the transformer itself coupled with the capacitor-input filter (rather than choke input filter) that I'd decided to go with (hey, I already had the caps in the junkbox). Modifying the supply to a choke-input filter may give me more output power, but honestly, it's more work than I think it's worth, so I'm leaving it as it is. (Note to self, though: next time, do a load test on the transformer and filter before installing everything!) 

2. Screen voltage is about 350 volts idle and drops to 300 volts when transmitting. 

3. At 290 watts out, HV reads about 1775 VDC, Plate current is 230 mA, Screen-grid current is 51 mA, and grid current about 20 mA. (Therefore efficiency is about 71 percent). 

4. When running at a Pout of about 290 watts, plate voltage of 1775 VDC, ebmin of about 300 volts (assumed), and efficiency of 70% (plus other assumptions per the "Radio Handbook" equations) these numbers work out to a plate load of about 4000 ohms. For an inductance of 10.8 uH in the pi-network, Q (given a 50 ohm load) calculates to be about 17, so we're in the ballpark of a Q between 10 and 20. 

5. Note the following pi-network Q relationships (using the Wingfield equations):
  • As output power increases (by changing loading capacitance), for a given value of pi-network inductance, pi-network Q will decrease.
  • As frequency decreases, for a given value of pi-network inductance, pi-network Q will increase.
How does it sound on the air? 

You can listen to a clip of the 813 Transmitter on W6THW's website here. It's the track labeled "K6JCA 813 RIG (AM)". (The rig was putting out about 300 watts, carrier power. Mic is a Heil PR-40 run through a Beringer 802 Mixer/EQ box, which feeds the Johnson Ranger's microphone input.)

References: 

 Articles:
  • "A Modern One Kilowatt AM Transmitter," W6BM, Electric Radio, #15, July, 1990
  • "813 Triodes as Modulators," W6BM, Electric Radio, #57, January, 1994
  • "An AM Kilowatt Using 813s 1989 Style," WA4KCY. Electric Radio, #5, September, 1989
  • "One-band Kilowatt Amplifiers," ARRL Handbook, 1961 - 1968 Editions, ARRL
  • "Class-C Amplifier Calculations," Radio Handbook, Editors and Engineers, 18th Edition (1970) [See note in text above re: error in equations.] Or one could use an earlier edition of this book, such at the 15th edition (pages 153-156) which doesn't have this error.
  • "Tank Output Circuits,"ARRL Handbook, 1997 edition, ARRL, pages 13.5 - 13.9 (Describes the Wingfield pi-network equations.)
Websites, Transmitters Websites, RCA BTA-250M Manual Websites, 813 Data Sequencer Designs:

Caveats!
 

Standard warnings apply: First, I may have made mistakes when writing this post or in my design. I cannot guarantee everything is correct. Second (and most importantly), this design uses high voltages that can kill you. Be cautious and BEWARE!