Restoration of Heathkit SB-200 (Part 1)

Over the past month I was rebuilding my Heathkit SB-200. When I originally bought the amp on eBay, there were a few small issues here and there - aging components, burnt HV choke, etc. The amp was actually a SB-201 converted to SB-200. It had the 10 meters band added with the original band-switch, tank coil and input network but no other mods. I wanted to update the amplifier and install all of the Harbach mods - HV power supply board, soft-key, soft-start, fan and relay. When I started the upgrades, I quickly realized that I can easily do a complete rebuild by replacing most of the components with modern versions. I stripped the amplifier down to the bare chassis and practically started from scratch, using the original Heathkit manual.

This is the Harbach High-Voltage Power Supply board. It is an improved version of the original HV board with modern computer-grade capacitors and high-voltage diodes. The equalizing resistors are of higher value and lower wattage. The board is a drop-in replacement. It also includes a string of 1% resistors for the voltage divider in the HV meter circuit.

Update: I upgraded the 6 Harbach PS capacitors (180uF/450V) with even bigger 390 uF/450V. This made the power supply "stiffer" and now the amplifier performs better than the stock version. The most obvious improvement is the lower HV drop under full load and a tad higher output power. Such an upgrade (installing high capacity filter caps) should be done only if the Soft-Start mod is already installed. The high-capacity filter caps bank draws much higher initial current (when charged for the first time during a power-on). If the amplifier doesn't have a step-start circuit, this in-rush current surge could trip the internal and/or external breakers (especially if the amp is powered with 120V) and also puts under strain the transformer/rectifier circuit.

This picture shows the multi-meter, bypass capacitors on the filament voltage line used for the backlight lamp and the new rectifying circuit for the front panel status LEDs. New, larger solder lug terminal strip (with more positions, from Radio Shack) is installed to accommodate the extra components of the low-voltage rectifying circuit - a diode, bypass capacitor and electrolytic filter capacitor. I am using half of the center-tapped filament circuit to supply the LEDs with about 3.5V DC . The half-winding is the same half used to power the Soft-Key module.
Two paralleled diodes (A-K, K-A) are installed right across the meter's terminals for protection. In the event of a high-voltage glitch the diodes will clamp the maximum voltage across the meter's coil, saving it from damage.

The underside of the amplifier. All wires were replaced with silver-plated/Teflon insulated solid wire. The coaxial lines were also replaced with silver/Teflon MIL coax cable. Almost all components were replaced with the modern type version - the new resistors are metal-film and ceramic-composite type. All capacitors and diodes were also replaced (most of the parts I ordered from Mouser. I also bought some components on eBay and from West Florida Components). The meter switch was carefully cleaned and lubed. Both UHF connectors (SO-239, input and output) were replaced with new silver/Teflon bulkhead type - the old ones were worn out and in pretty bad shape. For the KEY and ALC line connectors on the back panel I installed new female bulkhead BNCs. RCAs will work fine too but they don't have a locking mechanism.

The Harbach mods - Soft-Start (left) and Soft-Key (right). I installed the Soft-Start mod by gluing a piece of Plexiglas (with High-temperature RTV silicone) to the chassis and attached the PCB with cable ties through special L-shaped holes in the Plexiglas. Note that while this method worked just fine for me, Harbach recommends different location and way of installation. Harbach sells two versions of the Soft-start mod - for 240V and 120V AC mains. Each version is using different value power resistors. The best way to go is to order the "240V version" plus an extra set of the 20 Ohm resistors even if you are using 120V AC. For 120V AC mains, the extra set is soldered in parallel to the existing resistors. This makes changing the operational voltage more convenient. The directional coupler of the SWR/power metering circuit is seen near the top of the image. The coupler assembly was removed to allow access to the input coaxial cable and for replacement of the RF connectors. I cleaned and re-installed the coupler assembly after replacing the connectors and cable. One should be very careful with the germanium detector diodes (small glass body) - they are brittle and can be overheated during soldering, not to mention that they are also hard to find nowadays. The thin coaxial cables between the coupler's detector diodes and the front multi-meter rotary-switch were replaced with small diameter silver-teflon type coaxial cable. The power cord was changed to a heavy gauge one (AWG #12). Stress-relief is provided by a metal cable clamp. Visible at top left is a ferrite toroid (mix 43) with a few turns of the small diameter coaxial KEY line. This should take care of stray RF in the key circuits.

The new fan motor from Harbach and the new Tx/Rx relay. It turned out that the new relay had lower coil resistance compared to the original relay. During Tx mode, the relay coil is part of a voltage divider for the tube's operating negative grid bias. This lower resistance in the new relay coil had to be corrected by changing the value of the coil's paralleling resistor. I did some voltage divider calculations and shared the results with Jeff from Harbach. He adopted the mod and is selling now a new resistor together with the relay upgrade kit to fix this issue. Test measurements showed that the corrected circuit is providing the needed -2V grid bias in Tx mode.

This picture shows the high-energy "Glitch" resistor on the right. This resistor is encased in Kapton foil and mounted on teflon stand-offs. The high-voltage red wire going to the RF deck was salvaged from the power supply of a microwave oven - it is rated for 10kV. At the top of the picture are visible components of the front panel status LEDs as well as the current limiting resistors for the LEDs (located in teflon tubing). Also visible is the back side of the Operate-Stand-by switch. I used a bi-color LED to show the amplifier's operational status - Green when the amp is in stand-by mode and Red when in Operate mode. A blue LED is used to indicate the status of the KEY signal. Extra bypass caps are installed at the Stand-by switch line and LEDs. The switch is DPDT type - one pole of the switch is used to interrupt the KEY line (between the KEY jack and the Soft-Key module) and the other pole is switching the bi-color status LED.

Picture of the glitch resistor, main wire harness, some of the front panel LEDs circuit, the multi-meter switch and potentiometer. Also visible is the HV diode string (3 diodes in series) used to protect the meter during a HV "glitch" event. I added an extra RF bypass capacitor between ground and the output of the high-voltage supply. Teflon stand-offs are used to raise the glitch resistor above the chassis and away from other components/wires to prevent arcing. Kapton tape is applied as extra insulation around the high-voltage line.

Palstar DL1500 dummy load

Today I received my 1.5 kW dummy load made by Palstar. The model I ordered is DL1500.

This dummy load seems to be very well made and it is very light too! I needed something smaller than those heavy oil-filled loads made by Bird.

The resistive element in the dummy load is a single (custom-made for Palstar) ceramic-composite resistor. The resistor is fairly large, runing along the length of the enclosure and it is clamped between the SO-239 connector and the bottom part of the aluminum enclosure. DC resistance at room temperature is about 53.7 ohm (the meter is calibrated). This agrees with the factory specifications of 54 Ohm +0/-5%. Most dummy loads measure higher-than-nominal resistance when cold and the resistance drops with increase of the temperature. At working temperature of 70-90 degrees Celsius the resistance is exactly 50 ohms. Further heating will drop the resistance a little below 50 ohms. We are talking DC resistance but judging by the construction, the reactive component should be minimal at HF.

There is a metal hyperbolic cone around the ceramic resistor. The cone is mounted on aluminum stand-offs and it is used to correct the reactance component of the impedance at higher frequencies (especially on 6 meter). The enclosure is made from perforated anodized aluminum sheet. The RF connector is of silver/teflon type. The resistive element is held in place by two stainless steel clamps.

My MFJ-259B shows resistance of 52 ohms and reactance of X = 4 ohms for SWR 1.1:1 at 54 Mhz. At 14 Mhz the reactance drops to 2 ohms and R is exactly 53 ohms. Considering that the MFJ259B is not the most accurate instrument in the world - it should be just OK. This load is rated up to 500 Mhz but I am planing use it only on HF. I think the real characteristics are even better than what 259B shows! I'll take more accurate measurements with my Telepost LP-100 soon. At first glance DL1500 seems to be an excellent product and I am happy with it. A note of caution - this dummy load can take 1.5 kW power but only for up to 10 seconds and 100W for up to 10 minutes! Modifying the load for forced air cooling (installing a fan or two) will improve these numbers but breaks must be made between the intervals of high-power usage!

Let the blogging begin ...

At last, I decided to create my own blog. Seems that blogging is all of the hype nowadays. Every now and then I need to post something on the web and a blog might be a convenient form to do so! Only the future will show if I am going to get "hooked". This blog is not going to be different from any other blog out there - it will serve my interests and the posts here will be focused on my hobbies and fun activities. Enjoy!