EB5AGV's T-4XB repair notes

Keep them glowing!

Here you have some notes I wrote about the T-4XB transmitter repair.


Some time ago I sent a message about a T-4XB I had just received. It was supossedly unworking due to a bad PTO, but the seller was not sure as he had it as a spare for his working T-4XB and had not checked it. I would like to thank Gerald for thinking that this nice transmitter would be better working than stored as a spare.

Well, let's go on the repair process!.

First, as usual, I dismounted and cleaned the unit. The copper chassis was, as it seems usual in lots of B-lines, in bad shape and dirty. I used a new (for me) product to clean it: 'Tarni-Shield', a 3M metal cleaner, which also protects the metal surface once cleaned. It worked very well. Of course, there are still some dark spots here and there, but the overall looking is very good. Where the cleaning results were austounding was in the aluminum bars (which connect front panel knobs with variable capacitors) and also in the tube shields. They are now shinning like new! :-)

I cleaned also all the electrical contacts and tube sockets with 'Contact Cleaner', a CRC brand 'zero residue' cleaner.

Then, once all again at place, came the visual check. It is always very useful, as you will see... Looking at the PA cage, I found a burned choke (RFC3) and two cracked resistors (R31 and R33)...Could this be due to a bad PTO?... Not too probable, I thought. So, as the PA had some kind of trouble in his past history, I checked all the components around, and found that R26, R30 and R32 had changed its nominal value. I also found two colder joints, in PC1 and in an S4-rear trimmer. Once all of this was fixed, I checked resistances in the PA tube sockets, and found that all were on specs. By the way, I removed the 6JB6As and changed them for NIB Sylvania tubes, as the original tubes were of different brand and, as you know, this is not a good idea!.

The power supply I was going to use, was the same I used with the TR-4C I recently repaired, so I knew it was good, and no special care was taken on this unit.

So, after some more resistance checks, I decided it was time for the smoke test... Once done, all tubes lighted but, after a while, I heard one annoying sound: sparks!!!. I didn't feel comfortable with the sound, so the transmitter was fastly powered down. But I couldn't found the reason for the sparks, nor any burned part or the classic ozone odor... Hmmmm... Was I imagining things?. So, I powered it again... After a minute or so, I got again the sparky sound!. This time, I hold my breath and looked carefully... to find that the 0A2 was blinking!. So, what I heard was inside a tube... I was very happy to have an small stock of tubes, which included some 0A2s, so I changed it and, this time, the transmitter didn't generate any more the sound. Well, time for the voltage check!. More or less (as I checked them in SSB mode instead of the recommended TUNE mode), the voltages were fine.

Next obvious step was to check if I could load the transmitter on the dummy load... But things couldn't be so easy!. Nothing. Nada. Less than QRP :-)!.


About a month later, after some other BA and non-BA works, I put the T-4XB again on the workbench. My test equipment setup was as follows: TEK 475 oscilloscope linked to an HP-5328A counter, Fluke 77 DVM, HP-200CD generator, GM2883 HF signal generator and Bird 43 wattmeter. Another very useful item was a GC 'precision alignment tools' set.

I put the schematic on the desk, and read carefully the 'Theory of Operation' section. It is, IMHO, the best way to start a repair. You learn really how the things work inside those nice small boxes. Well, a fast check revealed that the carrier oscillator, the PTO and the crystall oscillator were working and at their nominal frequencies. Nice!!!. So the PTO was not the culprit, as I suspected... So, why I didn't get any power when I put the transmitter in TUNE position?. Looking at the schematic, I found that the signal used to tune the transmitter is generated by an R-C oscillator built around V9a. And my unit didn't oscillate... I connected the HP-200CD oscillator to the V9a grid circuit to be sure that the triode circuit worked. Once all resistors in the feedback loop checked, I thought that the culprit shoul be one capacitor (C126, C127 or C128). So I got new ones and changed them... Bingo!!!. Now I had a nice looking 910 Hz sinusoid when in tune position.

Checking the power output in 7200 KHz, I got about 5 watts. Whow, the little Drake was starting to live again :-)!. Evidently, what the T-4XB needed badly was a complete aligment. Somebody had played with ALL of the adjustable parts... Then I was very lucky to find, while connecting the oscope probe to a T-3 pin, another cracked resistor, R61 (150 ohm), which still was measuring about 120 Ohm but needed change.

After a careful alignment, following closely the manual instructions, the T-4XB is now working at full nominal power in every band (of course, more power in 80 and less in 10 meters). As usual, the 6JB6A needed neutralization to work properly. An small hint on this: neutralize in 20 meters, then do a better alignment in 15 and then go to 10. You can do it directly on 10, but it is best to be near of the final adjustment, as in 10 meters the neutralization is very sensitive to the variable capacitor position.

I have checked the generated audio (with a Shure 444 microphone connected) and it is really good, as usual with these transmitters.

Well, that's all!. Please, if you need any help with the alignment procedures, let me know, as I'm getting used to them and hope I could be of some help.


I forgot to tell that this unit is the same which had the trouble with the plate current reading wrong. There is a resistor, R38, which is in series with the meter. This resistor is a 'factory selected' one. Well, in my unit I had a 150 Ohm resistor. With that resistor, I got 70mA when the real current was 105mA. I checked R36, which is a 2.7 Ohm cathode resistor used to measure the plate current and, as it was a mixture of different resistors (really ugly looking!), I changed it for a 5% brand new 2.7 Ohm resistor. But the meter, as expected, was still reading wrong. So I decided to change R38. I needed a 17 Ohm resistor to get accurate readings!. I wonder what happened to the meter to behave that way, but now it is very close to the real current, both in standby (70mA) as in working condition (I have checked it and, when it reads 180mA, real current is 182mA; not bad at all ;-)! ).


Go back to Drake 4-line Page