Some Very Good Mods For The HW-10x/SB-10x

The modifications below are mods to the HW-101 which any HW-101 owner will want to consider. Everyone will not want to make all these mods, depending on your operating preferences. These are categorized into 3 levels, simple, moderate, and major.

Simple Modifications

I recommend that everyone complete these two simple modifications. They are easily accomplished and they greatly improve the operation and convenience in using these nice old rigs.

Conversion to Low Impedance Stereo Headphones and Speaker Muting

The original audio circuit in the HW-101 had some very nagging features. First of all, high impedance headphones were required. These were produced in large numbers during World War II, but it is difficult to find them and the audio quality is not great. Low impedance headphones are the modern standard and available in many styles and at low cost. Secondly, the speaker remains active (at somewhat lower level) when headphones are plugged in. This is totally unacceptable particularly if your station is near other people who are doing other things and do not want to hear your QSO. Late at night when I am working CW (which I do 99% of the time) I do not want to disturb the sleeping XYL. The simple modification described here will fix both of these problems.

The circuit diagram for the modification is shown below.

The only parts required are a special 1/4 inch stereo phone jack with switching, two 100 ohm 1/4 watt resistors and about 3 feet of stranded wire. The special stereo jack may be ordered from the Order Page. The modification is accomplished in the following steps:

1) Disconnect and remove the old 1/4 inch mono phone jack.

2) Place the 2 100 ohm resistors and jumper wire on the new switched 1/4 inch stereo phone jack and solder all connections except the one where to two resistor come together.

3) Mount the stereo jack on the front panel of the HW-101.

4) Unsolder the green wire coming from the audio output transformer and one end of the 100 ohm resistor from the RCA (8 ohm speaker) jack on the right rear of the chassis and then reconnect the wire to the free end of the resistor.

5) Solder a 2 foot length of stranded wire to the RCA jack output and run this wire to the stereo jack and connect it to the pole of the unused switch.

6) Connect and solder a 2 foot length of wire to the junction of the green wire and the 100 ohm resistor and run this wire to the stereo jack and connect it to the normally closed side of the same switch.

7) Connect the White-Violet-Violet wire (which was disconnected from the old jack) to the stereo jack terminal where the two 100 ohm resistors come together.

Sidetone Volume Control

The original sidetone in the HW-101 is far too loud for anything short of a demonstration in a large auditorium. Even if you are strictly a SSB operator, you will want to add this modification for those occasions when you want to use the CW mode for tuneup. The only part required is a 200 K ohm audio potentiometer and a little wire. I purchase one with a knurled 1/4 inch shaft about 1/2 inches long. Mount the pot on the rear panel so that you can change the sidetone volume without opening up the cabinet. You could cut a hole, but I chose to remove the RCA jack labeled "SPARE" and mount it in that hole. The schematic for this mod is shown below.

The following steps will complete this modification:

1) Heat the end of R326 (1 Meg ohm) which is farther from V15 and pull that end of the resistor free from the audio circuit board.

2) Solder a 6 inch length of wire to the free end of the resistor and pass the free end of the wire through a nearby hole in the chassis.

3) Solder a 6 inch length of wire to the point from which the end of R326 was removed and pass the free end of the wire through a nearby hole in the chassis.

4) Solder a length of wire to some convenient ground location and connect the other end to the appropriate lug of the recently mounted pot .

5) Solder the other two wires to the appropriate lugs on the pot.

Conversion of the original AVC to AGC (Eliminates S-Meter Zero Drift)

The AVC circuit in the original design has a number of flaws, the most troubling of which is the S-Meter drifting. Also, since the AVC voltage remains relatively small, and therefore it has very limited gain control authority and is not effective in keeping the audio level at a constant level as signal levels varied. The modification described here eliminates these weaknesses. The S-Meter no longer drifts and the audio level rises very little as the signal level goes from very small to very large. It has the advantages of being very inexpensive to build, requiring little modification to the original circuit and being easy to install and calibrate. In fact, the original AVC circuit is not disturbed at all. The AGC circuit which is added operates along side the old AVC, effectively overriding it by providing more gain authority. The ZERO pot for the original circuit becomes an S-Meter calibration control. The only modification is a simple modification of metering circuit.

The circuit diagram for the AGC system is shown below. A supply voltage of about -15 volts is obtained by rectifying the 12.6 VAC filament voltage, only a few milliamps are required. You may recognize the circuit containing Q1, Q2, R1, R2 and R3 as a differencing amplifier. D1 and D2 provide a voltage reference of about 1.3 volts. The voltage output from the amplifier is integrated in C2, a tantalum capacitor. The resulting is the AGC voltage which is applied to grids of the IF tubes.

The parts required are very inexpensive and easy to acquire. The circuit can be constructed in a number of ways. I constructed mine on a small piece of Veroboard which was secured to the wiring harness of the HW-101. The steps to be followed to complete the modification are shown above.

Simple Fix For A Heterodyne Oscillator That Will Not Oscillate

The HW-100/101 and SB-100/101/102 transceivers have 8 heterodyne crystals, one for each of the 8 band segments. The schematic diagram below is an excerpt from the HW-101 manual. What often happens is that as the crystals age, the oscillator refuses to oscillate in one of the band positions, resulting in both the receiver and transmitter being inoperative on the particular band. In my case the 15m position showed up dead. When it first happened, I was able to get it going by going through the first step of the alignment procedure in which the heterodyne oscillator coils are adjusted. Then it happened a few more times with the same result. Finally I was unable to get it oscillate at all in the 15m position. What was I to do? I ordered another crystal and the coil, Y504, and coil L604. These were installed, but it still refused to oscillate. What now? New 12AT7, still no good. Searching online provided no direct help.

I decided that I could construct an oscillator using the Y504 crystal and inject its output into the grid of V19A. This amounts to removing Y504 from the board, using it in an oscillator and injecting it back into the place from which the crystal was removed. I have used CMOS gates/inverters for constructing oscillators of several types, including crystal oscillators. They are very nice to use and have the characteristic that they oscillate very reliably, in fact, they will always oscillate when sufficient feedback is provided. The frequencies involved require high-speed CMOS gates of inverters. I chose the 74HCU04, which is a 14-pin device containing 6 high-speed CMOS inverters. Seems a shame, but only 1 of the 6 inverters is used. The circuit for the new oscillator is shown below. It is quite simple. It is powered by 5 VDC which is derived from the 12.6 VAC filament supply. There are only 3 connections that must be made to the transceiver, 12.6 VAC and two connections, ground and oscillator output, to the place from which Y504 was removed.

L1 is required because Y504 is a third overtone crystal. Without L1 in the circuit the oscillator will oscillate at its fundamental frequency, 9.965 MHz. I wound about 10 turns of enamel wire around a pencil and removed the pencil before putting it in the circuit. Its value is not particularly critical, but its value can be changed to peak up the output. If you should have need to use this circuit for other bands, the value would change accordingly. For the 80m and 40m bands the crystal is probably a fundamental (but I am not certain), meaning that L1 would not be required. You should have no trouble constructing and getting this to work, but if you do, email me and I will be glad to consult.

Moderate-Difficulty Modifications

The following modifications are not simple, but can be accomplished by those with average or better skills in a relatively short period of time and with little change to the original HW-101. Most of the modifications proposed here use of the 12.6 VAC filament supply because it is a good source of power and it has ample capacity for the needs of any and all proposed mods. But is is important that the circuit remain balanced so that the tubes with 6.3 volt filaments are supplied with proper voltage. Therefore before doing any modifications to the circuit an understanding of how the circuit is designed and what must be done to keep it operating right is necessary. The filament supply is 12.6 Volts AC, but some of the tubes require 6.3 volts ans some required 12.6 volts. Those which require 12.6 volts are simply connected between the 12.6 supply line and ground an are this unaffected by any of these modifications. The circuit contains both tube filaments and these two #47 pilot lamps which operate from 6.3 volts. The circuit is designed so that, in the event of failure of a tube filament or lamp, the voltage across any of the other filaments or lamps would not be greatly upset until the faulty ones are replaced. The diagram below shows the original filament/lamp circuit diagram, drawn in such a way that it is more easily understood.

In the circuit above, the filaments are shown as resistances but their values are shown as conductances, the inverse of resistance. The total conductance in a parallel circuit is the sum of the individual component conductances. Notice that there are two groups of components connected in parallel and the two groups are connected in series. When the circuit is properly balanced, each of the two groups will have the same voltage(about 6.3 volts) across all its components. Another way of saying this is that the conductance of the two groups is equal and the two groups draw the same current. With this arrangement, one element burning out will not drastically upset this balance. Whatever change is made to this circuit, the conductances of these two groups should remain roughly the same.

LED Lighting and a Cooling Fan

The old rig with all those tubes inside can get really warm, especially if the air in the shack is not moving. A cooling fan can make a lot of difference and there is plenty of excess power availablein the filament supply to power a small fan. Fans that are readily available at good prices run on 12 Volts DC. You can buy them for $10 or so. They draw about 150 mA. If we take out the #47 lamps, we can replace must draw 300 mA, the amount drawn by the two #47's, from the filament supply in order to keep the circuit balanced. As it turns out we can build a doubler from the terminals where the #47's are removed and regulate it to 12 Volts DC. When this is applied to a fan drawing 150 mA, the current drawn from the 6.3 VAC at the input will be approximately 300 mA, just what we need. The new circuit is shown below. One to three birght white LEDs are also shown in the schematic. These will provide as much light as the #47's. They draw only about 25 mA, which is negligible. After this mod, my circuit was very well balanced with the top and bottom group voltages being 6.5 and 6.6 volts. So there you have it, a fan which keeps the rig cooler and cooler lighting. Select an 80 mm fan which draws 150 mA. Be sure that the noise level is less than 20 dBA. You can mount the 12 volt power supply with doubler on a small piece of veroboard and mount the fan by putting machine screws through the slots in the rear panel.

T/R Relay Replacement

The HW and SB series of transceivers have two 4-pole double-throw relays which provide the switching necessary for transitioning between the receive and transmit modes. The reasons, as I see it, that one might want to change to an alternate solution are:

-One or both relays have failed (Suitable replacements are available)

-Desire for faster, quieter operation

-Combination of the above

For me, being a CW man, I wanted QSK operation. Considering the number of things required to be switched, the best I was counting on was to get performance like that of the slow QSK mode of my old Ten Tec Corsair II. My friend Cap Allen, W0CCA, put me onto the "DS series relays".. These switch in 4 milliseconds (mS). At 25 WPM, a dot lasts 48 mS. I have not measured the speed of the original relays, but I do know that I lost the entire first dit with the unmodified HW-101, meaning that it takes something on the order of 50 ms for the old relay contacts to close. Using the DS relays, there will be at least a 10x reduction in switching time and there should be no perceptible shortening of the first dot or dash. The relays come in many different configurations and voltages. I decided on the 24 volt, DPDT version, We need 3 of the DPDT relays (6 switches) since only 6 of the 8 available switches in the original relays are used. There were a number of reasons for doing this, but you can decide for yourself. I derived the 24 volt source from the 12.6 VAC filament voltage, using a doubler. The block diagram for my solution is shown below.

Click the link below to download a document which will provide detailed information about how I completed this project.

"HW-101 QSK - T/R Relay Replacement".