Entries from December 2009 ↓

Drake TR3

I dont need a new radio but sometimes it is nearly impossible to ignore a good deal. The TR3 was advertized as a parts radio. The asking price was $25 and the seller was local. No shipping. I could go pick it up.

So I did.

Parts radios dont work. Moreover, some parts in parts radios dont work either. The trick is to only use those parts that do work. Sometimes having a parts radio is not an advantage because the part you need to repair the ‘real’ radio is also bad in the parts radio.

In my case I wanted to repair the parts radio. Not that I needed another radio but a working radio is better than one that is broken. Besides there is little to compare with the aquisition of a new toy to inspire one to action. A working TR3 would be a real asset in that it provides 300 watts PEP. At that power level a linear might not be needed.

This TR3 was missing a couple of 12AX7s, all three finals, and an audio output transformer. I had saved the audio output transformer from the R4C when I converted it to solid state audio. That transformer works just fine in the TR3. A couple of 12AX7s were salvaged from an old strip chart recorder amp and we had noise coming from the speaker attached to the TR3.

Trouble was the noise did not change in pitch regardless of how much the tuning knob was turned. I finally discovered that the linear mechanism of the VFO had been moved past the stops and was no longer engaged in the tuning screw. Perhaps that was done to gain access to one of the screws that held the old output transformer. I got it tracking again after an hour of fiddling. It did not appear to require any alignment. The receiver was picking up plenty of signals.

12JB6s are still available at around $30 each. So much for inexpensive sweep tube finals. I never did care for sweep tube finals. Every Drake transmitter I ever had has been converted to use 6146s. The conversion is a simple matter of changing out the tube sockets, setting the bias level and neutralizing the finals. Current maket conditions have the 6146s priced around $30 each as well. So if you are missing the finals and dont have any spares, it would make sense to spend the money on 12JB6s and avoid the extra work in modifying to 6146s. In my case I had a supply of 614s so it made sense to use them.

The TR3 uses a 12 volt filament system. Tubes with 6 volt filaments are wired in series to work off 12 volts. Careful attention is paid to make sure that series connected tubes have equal current requirements so that all tubes get the filament power they need. You cant hook three 6 volt filament tubes in series to work off 12 volts. My solution to this problem was to run the three 6146 finals off 6 volts. I did that by reassigning the speaker lead to the power connector to carry the 6.3 vots AC. Not sure I like the idea of a speaker lead coming out of the power connector. Seems like an accident looking for a place to happen. A speaker can still be conneced through the phones jack on the side of the radio. That is another sore point. Side mounted phone jacks and controls are fine if you dont have any other equipment but they get in the way of everything else you want to set adjacent to the TR3.

More knowlegeable readers might point out that there are 6146 style tubes that have 12 volt filaments. The objective here was to economize by using parts on hand. After spending all of $25 for the radio I was not interested in spending another $100 to make it operational.

I did end up spending another $25 on paint, resistors, capacitors, and a decent soldering iron to convert and restore the TR3 to an operational condition. I already had an AC3 powersupply to run the T4XC. I used it to check out the TR3 but ended up building a special supply for the TR3. When you do specials like that you can design for optimum preformance. Drake only supplied 650 volts to the plates of the 12JB6s. 6146s are prefectly happy with 700 volts. I ended up running the 6146s at 750 volts. That is a little over 400 watts PEP. This little rig puts out a respectable signal on 75 meter SSB without need to run a linear.

The paint I used was Rustoleum semi gloss black. I bought two cans but only needed one can to do the job. Sanded all the scratches out with a power sander. Grey primer. The primer is not really needed but it sands better than the finished coat. You want to sand the primer to get a smooth surface. Then three very light coats of semi gloss black and the cabinet looks like new. Rustoleum semi gloss black is not an exact match for the original Drake color but we are not doing a touch up job. We repaint the entire cabinet so the color matches perfectly.

New resistors and capacitors make the conversion easier. While it is possible to salvage the resistors and caps from the old sockets, new parts are only 12cents each and are easier to use.

12JB6s only have one cathode pin. 6146s have three cathode pins. I suppose using one of the three would suffice but we ended up using all three. The cathode resistor used by Drake is 15 ohms. We used three 45 ohm 1 watt metal film resitors per tube and tied the other ends of the resistors to a #12 solid copper wire bus to connect all cathodes together and to the RF choke at the far end of the final compartment.

Screen leads were bypasse to ground with .005 ceramic 1KV caps. All three cathode pins were bypassed with 450 pf 1KV ceramic caps.

The original screen resistors were 68 ohm half watt carbon. One of them was open. I replaced all three with 68 ohm 1 watt metal film.

The new tube sockets I used did not require any modification to the tube socket holes. Make sure you use octal sockets that fit a one inch diameter hole. I did not even try to use the captured nuts that Drake used to mount the novar sockets. 4-40 hardware works fine. Use locknuts or paint the treads with nail polish to keep them form coming loose. When selecting octal tube sockets make sure they have a decent mounting ring. You will want to solder to this ring to terminate the ground side of the bypass capacitors. You also need to solder the ring at one point (or more) to the copper plated chassis to make sure you get really good electrical connections.

Mount and solder as many parts as possible to the tube sockets before mounting the tube sockets in the radio.

Note that the novar socket mounting rings were sldered to chassis as well. Save the noval sockets. I have about a dozen noval sockets in the junk box. I am savng them in case I find a stash of sweep tubes. More importantly, they can be used to reverse the conversion to 6146s should the need every occur.

The original grid connections were done using a brass strip. That same brass strip was reinstalled by soldering it to pins 5 of the 6146 sockets. It is not as neat an installation as the original but it is perfectly functional.

So now for $50 dollars and a week of light part-time work we have a new radio. Hard to believe that this thing sold for $550 when it first came out in the early 1960s. Estimates have that $550 equal to $3700 in 2009 dollars.

This TR3 is 40 years old. I doubt that it had any tubes replaced in that time. Those that I replaced were to populate empty tube sockets. There is no reason to believe that this thing will not work for another 40 years. However by that time replacement tubes might no longer be available. They are scarce and expensive even today. So it might be a good idea to set asside a stash of spares for future use.

I would be happy to let this thing go for $300 complete with power supply today. Once I get attached to it that might change.

Hold on a minute. Lets see. I have about $100 in parts in this project including the power supply. I also have about 20 hours time invested in this thing. Twenty hours taken over $200 is $10 an hour. I would not be happy working for minimum wage. No, I would need at least $700 for that TR3 and I dont know of anyone willing to pay that. Guess I will just keep it. Looks like I am already attached to it. Might donate it to a worthy cause and claim $3700 for tax purposes. Its not my fault that the dollar lost its flavor since 1960.

A word about restoration. Some folk view restoration as a knob polishing excercise. They take great pains to clean and polish to make the rig look like a new penny. I have nothing against polishing and cleaning as long as it does not interfer with operation. If the radio does not work no amount of cleaning is going to make it useful. Junk is junk regardless of how clean it is.

This project resulted in a radio that is not spotless or shiney but it works as well or better than when it was new.

300 Ohm TV feedline

It used to be that every TV antenna used a 300 ohm feedline. It was 300 ohm twin lead. It was cheap and the lowest loss feedline you could get. The only problem it had was that it picked up noise. In an analog system noise shows up on the screen. To get rid of noise TV feedlines were changed to coaxial feed lines. These were shielded and pretty much eliminated noise.

Digital TV does not show noise on the screen. All it needs is a minumum signal level. Once that signal level is met you get a perfect picture. No noise, no ghosts, no aberrations, just a perfect picture.

So now I am going back to 300 ohm twinlead and dumping the coax. Less loss in signal from the antenna using 300 ohm twinlead and it is much less expensive too.