This is another out of the ordinary post, this covers the repair of a Triplett 3212 Tube Analyser. There are more than one version of this, some have the roll chart as part of the case, this one has the roll chart separate from the case. I chose to post this as I could only find some examples of the tester on the web, but nothing of the schematic. While there are some similar Triplett models with simialr schematics, the resistor values are different. I did look through “BAMA”. and those models were not close.
A tip of the coffee mug to Triplett. They are still around making test equipment and were kind enough to provide the schematic for the 3212, with fast turnaround of my request too. And again, many thanks to them.
This model has a number of examples on the web, but no schematic that is not without charge. So the schematic with a little more legible values is here, and many thanks again to Triplett for this:
I pulled this out a few days ago to evaluate and repair the meter. It had seen a little rough handling while the glass face of the meter had been broken- Suprisingly, the only issue was a lowly spider web strand which kept the needle from moving back to the “zero” resting point, but I did not know this until I had the meter opened up to check it out. The glass piece was easy enough to replace, although I was not sure that I would have clearance with “double strength” glass initially as the original was single strength. Single strength is barely 1/16 of an inch (about 1.5mm) thick, and double strength is about 3/32 inch thick (about 3mm). The glass is held in by 6 screws and washers, and additionally made dust proof by a sealant. The sealant scraped away from the Bakelite/Phenolic face easily, but did require some care. I bedded the new piece of glass with simple gasket sealercompound, and reinstalled the mounting screws in an alternate pattern to keep pressure even on the glass and to keep it from shifting too much within the recess for the glass. For similicity of assembly, the adjusting screw was lined up so vertical, and the slot that it rides in was also lined up vertical prior to final assembly, so once assembled, when the “Zero calibration” screw was turned, the needle moved until set to correct “Zero” location.
I did not take any action shots of the glass replacement/meter repair as this post was an after thought.
When initially tested the unit after the glass replacement, Line voltage was not able to be calibrated, it measured consisitently at 40%, so I opened the back again and checked all resistors to make sure none had opened. There is a wirewound resistor inside of the meter movement itself, but that is there so the meter itself is calibrated as a 1 Milli Ampere/100 millivolt movement (1 MilliAmp/100millivolts brings the needle to full deflection) and any deflection of the movement indicates that internal resistor is just fine when adequate current or potential is applied to the lugs.
The 2 carbon composition resistors in series that I point out in the rear view image were well out of tollerance for the circuit once I was able to compare to the Triplett provided schematic, so the pair was replaced with a metal film resistor of sufficient wattage. Other updates are the replacement of the copper oxide rectifier with a silicon rectifier (1N4007), and replacement of the old paper cap with a polyester capacitor of the same 0.1 microfarad capacitance at 630 WVDC. The copper oxide rectifier is safer than selenium, but still has significant reverse bias leakage, the silicon diode has much less leakage, so a significant upgrade to performance. This is the little black thimble sized item mounted with a #6 screw on the left side of the turret board indicated in the image- the hole is indicated, the copper oxide rectifier has been removed.
When I first put this tester into service after buying it at an estate sale, it would not calibrate at all, and not even respond, the original heavy duty line voltage potentiometer had corroded open, and the wirewound originally in the position where the new film resistor was installed (75k/R12) was replaced by those 2 composition resistors (pointed out as “not going to anything” in the image), and was this time replaced with the 75K film resistor, because the composition resistors were too far out of tollerance. The original R8-5K resistor was also well out of tollerance and had been replaced at that time when I first worked on it back in the 1990’s. The power cord was upgraded at that time to a grounded cord for safety reasons. The top cap wire was also replaced along with a better “Gator clip” instead of the bent wire clip.
Once it was updated for the 21st Century: this time around with “in tollerance” parts, it did not calibrate to line voltage. After reflowing many of the solder joints I could reach, I flipped it over again to test it, and no real improvement. Then I remebered that the needle on a meter face that large could be interfered with by just the static on the face of the meter. So I rubbed a dry paper towel over the face and it disipated the static charge enough to allow the needle to now pass the 40% level and be set to line level correctly. One last thing I will do is to check each of the heater/filament voltage settings to make sure they are close to what they are supposed to be. With the line level now being able to be accurately set, they should measure up quite well.
Would it have made a difference if I had deduced the static on the glass issue first? No; everything else that was done this time of repair I would be replacing anyway because the parts were out of specification or were not entirely reliable, therfore the order of their replacement/correction is irrelevent. I only make mention of it here because the issue of static on the face of these meters is not always an issue for people, but when it is- it is one of those less common things can burn a lot of time just because it is not obvious.
With a layer of mold scrubbed off it with paper towels and bleach water, it looks presentable now. I think what caught my eye originally was the large meter. And the markings are pretty much the same color they were when the meter was on the assembly line. Apart from some rust from where it had been stored between the 1950’s up to the time of that sale where I bought it, it is not in all that bad of condition.
I may have a similar knob for the #10 test switch, otherwise I may just leave it for now. The Roll charts sometimes comes loose on the ends due to the celluloid tape that was used originally, Celluloid tape will have certainly failed by now. I just used the classic translucent clear tape to re-anchor the ends, and it has worked just fine for several years since I initially put it in service in the 1990’s.
The settings for the testing of tubes are the filament/heater voltage, the circuit type, the load, and 10 switches that are set to one of three positions for the basic testing. This unit differs from the other Triplett models that may still be found. And this is an “Emission” type tester like many of the similar marked models. Some people get dismissive of them, but: 1- it tests tubes in the manner the RMA settled on, if even for a “Go-No Go” test. While it is not a perfect tester, once you get used to using it, you can tell if tubes are going to be well enough matched when in a “Push-Pull” circuit, or in balanced preamp stages such as 12AX7’s where each half of the tube is used for each stereo channel to provide acceptable results. For matching pairs or quads, a tester like the Triplett 3444 (any version) or a TV 7, or most of the EICO testers are a better choice than simple emission testers, especially if you plan on selling used tubes, or for close matching tubes via characteristics for best results.
And for 2- it has the settings for many tubes that are not even listed on other tube charts, such as the single letter tubes, and old ones such as an “00A”/”100A”, and one of the few that can test pencil tubes.
Even the best testers out there should not be used as a final arbiter of merit when used for personal reasons, because some radio chassis will still perform admirably with tubes that test low- on the order of “30” on an emission type tester- I actually have one of those quirky radios. It is one that was likely made for an electronics class as the cabinet had a rather generic look to it and no name on chassis or cabinet, or dial- but it worked very well with several tubes that would not work in other radios. I found out that it could work well with weak tubes a little by accident as I was swapping the known good tubes for weak tubes in a few other radios and just plugged the weak tubes back into that chassis so I knew which tubes went into which sockets, and one day I decided to just try it with the known weak tubes, and it worked just fine.
So if you have a tester, and you have numerous tubes that test low- test them in circuit and see how they sound. Some tubes like 6SQ7/12SQ7, 6A8, 6A7, 1A7 notoriously test low. You may have a fault developing in your tester, or you may have had a knob incorrectly set when testing (there have been a couple examples of the wrong settings being published in roll charts even) or the tubes just test low, but work fine in circuit. Let the circuit be the final judge on merit. Also double check to see that the tester is applying the correct heater/filament voltage to the tubes under test. While you may rely on the tester’s built in calibration, it never hurts to check to see if the heater’filament voltage is correct- it may be low, or in some cases, may be high if you have trouble getting repeatability in tests. More about this in relation to my own 3212 tester at the bottom.****
If you happen to be restoring an old TV set, you may find the best results with used tubes is to put the higher reading tubes (when you have several of the same type in a set) in the earlier stages. Such as if you had 5 of type 6GH8A, the tubes that tested better should go in the circuits such as 1st RF, 1st IF, Sync Amp, Quadrature, Color Killer, rather than the later stages because those first stages are critical to a good signal reception. The same is true for radios- if you have 3- 6SK7 or 6K7 for example, for best results, the highest tested tube would go into the 1st IF, or if the radio had a tuned RF stage that used that tube, it would go there. the lowest tested tube would go into 2nd or third IF. This makes the reciever more sensitive to signals than if weak tubes are in those locations.
Some last notes- sometimes a tube will test low, and sometimes this is a tube heater or filament that needs to be heated for a short while above the normal operating temperature to boil off impurities, and this can be done by increasing filament or heater voltage with no load on the tube. At the higher filament or heater setting, the meter will read higher when under test, but there are times when the general merit once the tube has cooled down; has improved to an acceptable level. It usually does not take too long, nor does it require significant increases in the heater potential. This is essentially what CRT rejuvinators did/do, and the plug in units that were sold as “Picture Tube Brighteners” usually increased the heater voltage of the kinescope (CRT) and those increased the emission of the old tube enough to be kept in service. They did work suprisingly well too.
****While trying to get the unit to calibrate for checking filament/heater votlages, it would not adjust correctly even after static neutralization, so I chose to try an alternate calibration method, and it actually worked very well. The main thing is to calibrate the tester to have correct heater/filament voltages, the other voltages will fall where they were desingned. Although replacing a Selenium or Copper Oxide Rectifier reduces in circuit resistance of rectifiction, this leads to slighly higher voltages for the “B” voltage for testing, but this is within a reasonable amount of variation, so i snot a significant concern fo emission type testers such as this one. While it may seem a cavalier approach, it really is not because the tester calibration step is supposed to establish a consistent reference point for the windings of the transformer(s) which are wound with specific numbers of turns and taps for filament voltages and test voltages of the unit. Most importantly calibrated in Emission type testers is/are the heater/filament voltages of the tubes under test.
If your alternate calibration is correct, with just the meter taking measurements, you should be able to set the switch to each position and see 25 volts at “25”, 0.75 volts at “0.75”, etc.
The steps for this alternate method is to select a tube that you have tested on other testers to know it’s relative merit. Then set this tester accordingly without installing the tube just yet. (Normally you would set the switches to the correct positions before inserting tube into test socket in normal testing as well to prevent burning out tube elements or causing other problems that could damage the tube to be tested). Only instead of a tube, you are placing meter test leads at the pins where the heater/filament would be. You need to be checking with a good AC voltmeter across the heater terminals. For my calibration test, the tube I chose to set for was a 12SK7. This tube has heater locations at- pin 2 and pin 7. So I set the number 2 switch to “up” position and switches 5 and 7 to the “Down” position as indicated in the roll chart. This is the important part, otherwise there is minimal voltage to measure at those heater/filament locations.
The calibration knob had been fully clockwise when trying to set calibration the original way, but it would not reliably set the meter needle to correct position, and was always low. However when I set up to measure the heater voltage, it was too excessive, so I adjusted the “Line” Control” down to a point where the measured heater voltage was correct for the filament voltage knob position, and that point was actually fully counter clockwise. Since I had replaced the original rheostat, there was no true “Off” position, but in that fully counter clockwise position (with the replacement potentiometer, actually being a 250 Ohm unit instead of 175 Ohms of the original); all the test voltages for the heaters, measured across pins 2 and 7 of the octal socket measured correctly. I tested my reference tubes and they were right in the range they were supposed to be, so from now on, when I calibrate this tester, I will use my alternate calibration procedure and not worry about the tester’s meter indication for calibration, but I will make sure the filament /heater voltages are correct before test. The ultimate issue may be something as obscure as the glass being double strength as opposed to single strength and thus interacting with the needle in a capacitive manner; but right now, I am not going to worry about that detail because the rest of the components of the tester checked out within specifications, and with the alternate calibration procedure the reference tubes checked out where they were supposed to, so if a tester has some issues with calibration, use the alternate method instead, as I will certainly be using the alternate method for this tester for reliability. Lastly, even though the line control potentiometer was slightly different, I had swapped that before this last calibration method because I was unsure of the condition of the initial replacement; and where it is in the circuit, (R13), it turned out to not be an issue in my case because of where it ultimately had the correct filament/heater votlages. This may sound repetitive, but I am just trying to “cover the bases” of possibilities with these testers.
The tangential blog.
The passive solar blog- outgrowth from some projects of mine.