The General Electric T1000C Stereo Receiver

My parents had one of these units on the side table in the dining room. My father put up an FM antenna outside on the roof so he could listen to more stations. In the early 1960s, there were not as many around as there are today. Our house was on the wrong side of a hill for the NYC stations, although Peekskill seemed to come in just fine. What is fascinating to me is the timing and cost. These stereos were made in 1963, not long after the Zenith/General Electric FM stereo system was adopted and first broadcast on WGFM (now WRVE) in Schenectady, NY (June 1, 1961). Not every FM station rushed out to install the new system.

General Electric T1000C Stereo receiver marketing

For a bit of a reference, $180.00 in 1963 is worth $1,868.65 in 2025. At that time, my father was an installer/repairman for New York Telephone. My mother was not working and six of us lived under one roof. That was quite a bit of money for an AM/FM radio.

The radio was normally tuned to 100.7 WHUD, which initially went stereo in 1972. Other stations that could be received: WGFM, WROW-FM, WSPK, WEOK-FM (now WPDH), and WGFH (later WINE-FM now WRKI).

General Electric T1000C stereo (Walnut cabinet)

I purchased this one on eBay for $70.00. It turns on (in fact, it did not turn off), there is a hum, the pots are scratchy, etc. However, if I tune it to one of the local AM stations, I can hear music under the loud 60-cycle hum. In other words, it works! So, I spent time fixing all the defects and enjoyed some nostalgia. According to this date stamp, the wood enclosure was made in January 1963. I would think the rest of the unit was made about the same time, which means this is one of the first radios in this model. This may have been manufactured in Bridgeport, CT, or Syracuse, NY. The serial number is missing from the back of the chassis.

Crushed capacitor

The main source of the hum appears to be this capacitor, which clearly has seen better days.

The on/off problem was the selector switch, which stuck in the on position because it was gunked up with dried-up lubricating oil and dust. I cleaned it with denatured alcohol and DeOxit.

Production date; January 1963.

The parts list included about $15 worth of capacitors, $1 for a new rectifier diode, a $7 telescoping FM antenna from Amazon, and $6.32 for two PLT 12 6.3 volt miniature lamps for dial light.

All of the tubes look to be the original GE units. After the recap, I turned it on and there was nice sounding AM, but no FM. The FM RF section has a triple triode (V2) which is the AFC, 10.7 MHz Oscillator, and mixer. This tube was loose in its socket and needed to be reseated. After that, everything worked.

GE T1000C chasis

All of the pots were scratchy. I cleaned them with DeOxit and worked them back and forth many times. After a while, they all are working.

FM Stereo receiver MPX decoder block diagram

I found a Sam’s Photofact (basic service manual) on this set. What is very interesting is the schematic for the multiplex receiver. This section decodes the L+R/L-R signals and produces the stereo audio. Unlike modern FM stereo receivers, in which the broadband multiplex signal is fed into one side of a chip and the discrete L/R signal comes out of the other side, the signal path through the various processing stages can be followed.

GE T1000 MPX decoder schematic diagram

The broadband MPX signal comes from the IF stage via wire #27. The signal is amplified by V6. The L+R (20 Hz to 15 KHz) or mono signal goes through a low-pass filter L17/C40; the 3dB cutoff should be around 16-17 KHz. The L-R and 19 KHz pilot goes to wire 34, thence through a high-pass filter C37/L16/C38; the cutoff should be 20 KHz or so. The L-R and 19 KHz pilot are Amplitude Modulated subcarriers on the FM signal. Wire 38 routes the MPX signal to V6 which recreates the 38 KHz subcarrier by doubling the 19 KHz pilot. This is filtered by a bandpass filter C13/L14. The L-R and the 38 KHz subcarrier are sent to the product detector.

Diode product detectors X4 and X5 (1N541) demodulate the lower sideband (23 – 37.98 KHz) and the upper side band (38.02 – 53 KHz) respectively. Those signals are summed in the matrix subassembly K4 with the L+R. Mathematically, the results are:

The Left and Right audio is then sent to the first audio stage V7 through a deemphasis network. If no 19 KHz pilot is detected, no 38 KHz carrier is recreated and this stage remains silent. In other words, you have to find an FM station in mono first, then flip it to stereo to see if there is enough signal to decode the L-R. One of the limitations of the first generation of FM stereo receivers. Newer versions of this set have a stereo light, or “Stereo Eye” so the listener knows when stereo reception is possible.

The front of the cabinet is covered with glass, which I cleaned with soapy water. The glass has part of the gold leaf trim rubbed off. I think this radio got a lot of use.

I let the knobs soak in soapy water overnight then cleaned them off with an old toothbrush. I believe that this radio was once in a smoking environment, based on the amount of yellow, gooey substance covering everything. I ended up disassembling the entire unit to clean it. I used a paintbrush and the shop vac to get all of the dust out of the cabinet.

General Electric T1000C disassembled

The speakers and speaker cones are in good condition. The speaker cabinets needed a little work; in both cabinets, the fronts (the part that is seen when both speakers are “closed”) were popping off. I had to glue a bit of wood back together and fix the metal holding brackets. The cloth on the speaker side is a little faded.

General Electric T1000C restoration complete

The wood finish is in good shape with a few scratches and dings. I decided to use Howard Restore-A-Finish. This is not the same as stripping and refinishing but rather repairing the existing finish. There was a water ring on top of the cabinet, which was removed with the Restore-A-Finish and light use of steel wool.

Three power supply capacitors, held down by a ty-base glued to the chassis

Reassembly went about as expected. I glued these tie bases to hold up the new capacitors.

The receiver is fairly sensitive and the dial is accurate. There is an alignment procedure in the repair manual, but I think everything is working as it should. I have spent enough time trying to fix things that are already working to know that for a 1963 tube receiver, this is good enough. Perfection, as they say, is the enemy of everything else.

So, how does it sound? Pretty darn good, as it turns out. I am working on a brief YouTube video with some religious music (I’ll post it when it is done). On the FM side, I can get WAMK, WBPM, WKXP, WJUX, WDST, WPDH, WFSO, and WPDA clearly with the whip antenna on the radio. AM, I hear WGHQ and WJIP.

I can hear the old man now, humming along to his favorite tune…

The Bext TFC2K broadband FM antenna

FM Broadband antennas are a compromise because they generally have less gain than tuned antennas, are more complicated, and take up more space. However, this antenna has none of those issues. The gain and radiation pattern appears to be almost the same as a tuned three-bay FM antenna.

We are finishing up an antenna project in Pittsfield, MA, this week.

Proposed W277CJ 60 dBu contour
Remnants of Shively 6812 4-bay antenna

The project involved replacing a Shively 6812 tuned to 95.9 MHz (WBEC-FM) with the TFC2K so that the W277CJ 103.3 MHz (WUPE) translator located on the roof of the 14-story Holiday Inn on West Street could be moved to the studio location. In this case, having the translator in-house will save significant rent. The new antenna will continue to serve as a backup facility for WBEC-FM when the main site is off the air for whatever reason.

Single bay, Bext TFC2K antenna

The input power per bay is based on the antenna’s input connector. In this case, each bay has a 7-16 DIN connector and the power divider is a 7/8 inch EIA flange. Thus the maximum input power for this setup is 5.5 KW. The licensed output for both facilities is far below that.

3 Bays leg mounted on the tower

According to the manual, this antenna should be spaced at 0.85 wavelength, which is frequency-dependent. I chose a frequency halfway between the two (103.3 – 95.9)/2+95.9 = 99.5 MHz. The formula from the Bext general antenna manual is:

D = (300/F) x 0.85

Where
D = the distance between center of radiating elements (booms)
F - Frequency in Mhz.

Thus, D = (300/99.5) x 0.85 = 2.56 meters (or 8′ 5″)

As this is a series excited AM tower, some type of broadband isolation coil is needed to cross the base insulator. This one is simply a large coil of 7/8 inch coax, likely with a capacitor across the outer conductor to create a resonant LC network.

To me, it looks like a water heater. Since the ground is frozen solid, we made a temporary stand. We will have to come back in the spring to create a permanent stand or perhaps a unistrut mount to the wall of the ATU building.

Kintronic ISO-88P-78EIA-4C

In the rack room, the transmitters are combined into a Bext FDCSDC2 star point combiner.

Antenna combiner

Broadband sweep shows a good match across the entire FM band. I will be interested to see how it performs with respect to the Shively single bay 6812 on the roof of the hotel (103.3 W277CJ).

Return loss, Bext TFC2K 3 bay FM antenna

The return loss looks good on both 95.9 and 103.3 MHz. The interference noted in the sweep is from local FM stations including the main transmitter for 95.9 MHz.

Weather related broadcast issues

Ice accumulation

We just finished our 3rd annual February ice storm. It is becoming somewhat of a tradition in these parts. After shoveling the driveway this morning, I sat down to enjoy my nice hot coffee. While doing that, I figured I would check some of the transmitters to see how things were going. That is when I noticed this:

The reflected power is much higher than normal indicating potential issues with the antenna deicers. I knew something was wrong after a quick call to the Burk, which stated the deicers were on. A quick double-check showed that the reflected power had increased by another 75 watts, so a nice drive to the transmitter site was in order.

Road to the transmitter site

Indeed, the controller had turned on the antenna deicers.

Antenna deicer controller

Using a clamp on amp meter, I saw almost no current on either leg of the 240-volt circuit. In the meantime, the backup antenna had 2 amps on each leg, which is normal. Then I noticed this:

Antenna deicer relays

The relay on the right shows signs of overheating.

I moved the Main Antenna circuit over to the aux antenna relay to get things going again. The current on each leg of the main circuit is 4.2 amps. Over the next 45 minutes, the reflected power returned to normal.

Other transmitter sites to the north in Albany have had similar issues. Unfortunately, those antennas do not have heaters or radomes. Thus, the only remedy is reducing power until the transmitter stays on.

I also noticed that when there is an antenna problem, the station does not sound as good as it normally does because of the bandwidth restrictions adding distortion in the frequency domain.

Baltimore Public Radio gets a new FLX20

Another liquid-cooled GatesAir transmitter installation was completed for WYPR, Baltimore, MD. This unit replaced a Continental 816 which had a long life.

The area around the transmitter site is not the best neighborhood. The building was formerly the WBAL-TV transmitter site and was built circa 1947.

WYPR transmitter site
Original four-legged Blaw Knox self-supporting tower from the WBAL-TV days
Pump station and HTF tubing
Wall-mounted heat exchanger
FLX-20 transmitter

The transmitter room is a little tight, so it was difficult to get a good shot of the front of the unit.

The station is running HD Radio with -10 dBc.

This site has a strange 3-phase delta AC power configuration. The middle leg is at ground potential, and the other legs all measure 240 VAC to ground. I’ve never encountered that before. This is known as a corner grounded Delta, which gets rid of the high leg associated with most closed 3-phase delta systems.

Ultimately, all the leg-to-leg measurements are 240 volts, so the power supplies are satisfied. With these transmitters, the phase rotation does not matter because there are no actual 3-phase loads in the transmitter.

Inventory control, Home Depot Reisterstown Road, Baltimore, MD

The sign says “Free Tools.”