Eico 666 / 667 Tube Tester
First seen about 1956, this Eico ‘Dynamic Conductance’ tester is a very popular model to find at ham radio swap meets, and a ‘cult’ favorite. It came as a kit, or factory wired for an extra $40. Where most tube testers were housed in a vinyl-wrapped wood box, Eico liked to use an 18-gauge steel ‘chassis’. Some folks feel steel enclosures allow a heat build-up inside, especially with a large 83 rectifier and a 5Y3 rectifier giving off quite a glow. Add to that the fact that there is no ventilation or cooling fan in any tube tester I’ve ever seen, and it does makes some sense. Eico avoids this situation all together by not using any rectifier tubes. While there is a simple unregulated power supply, the sensitive meter movement was combined with multiple scales, helping for more accurate readings. I suppose to really avoid potential obsolescence, the tester had an extra power supply and further dedicated meter scales for testing those new-fangled transistors! I own, and have seen, more than one Eico 666, and no one (including myself) has ever actually used the transistor test function of this unit.
Advertisement extols the many virtues of the Eico 666.
The advertisement above is taken from the December, 1956 issue of Radio & Television News, and extols the many virtues of the Eico 666 tube tester.
- The leakage test uses the meter movement to measure actual leakage resistance, up to 20Meg! This is far superior to the typical neon bulb indicator, which is usually limited to about 1Meg worth of leakage resistance.
- You can specifically measure the heater-cathode leakage, and get the ‘real-world’ measurement.
- You select from multiple ranges of test voltages. A 12AX7 is not tested at the same plate voltage as a 6L6!
If there is any ‘problem’ to using the Eico 666, it is as follows. The roll chart has an unusually high number of errors, or changes from one edition to the next. The roll charts are designated from about 666-03 to 666-09. The designation can be seen written at the very extremes of the chart. Be careful; do not run the chart right off of the rollers trying to see which ‘version’ you have! Here, in a nutshell, are what the levers do in setting up an Eico 666. The ten levers (each corresponding to the successive tube pin number) have six positions to chose from, and they connect the tube pins thusly.
The ‘S’ lever selects the ‘coarse’ meter sensitivity shunt resistance, which is set further with the ‘Plate’ control. The ‘V’ lever selects the appropriate combination of plate, screen, and grid voltages. Let us assume we are testing a common 6SN7 tube. First, here is the pin connections for a 6SN7.
Now, here is how the various Eico 666 tube charts want to you to set up the test for the first triode.
Here is what I notice. First, the heater connections are pretty easy to get ‘right’. And, Eico seems to have it ‘right’ for successive roll charts. From the 666-03 chart, the grid connection seems OK, but the plate is actually fed from the screen supply. Over at the 666-08 chart, the plate connection is ‘corrected’. All roll charts suggest the cathode gets grounded, which is OK. However, for roll charts 666-03 through 666-06, the two 6SN7 triodes get connected together in parallel! In 666-08, the section not being tested get grounded, as does any unused pin connection (lever 9 and 10 in this case). Also note how the voltages are increased in the 666-05 compared to the 666-03. They are lowered to a more reasonable level for the 666-08, where the bias voltage and meter shunt vary considerably from previous roll charts. Things get really weird for ‘common’ guitar amplifier tubes like the 12AT7 and the 12AX7. Let’s start with the 12AT7, shall we? Here is how triode section ‘one’ is set up.
The ‘Coletronics’ settings are courtesy of a published book by the aforementioned company. Coletronics compiled updated tube tester data for many companies, under the manufacturer’s auspices. Since these settings are so radically different, I decided to check the plate current through this poor little 12AT7, using each tube chart.
What you may notice is that the 666-09 chart setting really lays a beating on the poor 12AT7! According to many receiving tube manuals, a 12AT7 is rated for 2.5-watts of plate dissipation. By checking your typical RCA Receiving Tube Manual, we can see a 12AT7 with a plate voltage of 100VDC is expected to have 3.7mA worth of plate current. At higher plate potentials, the current shoots up to 10mA, but remember that higher plate voltages lowers the internal plate resistance of the tube, and the logarithmic increase in plate current is expected and makes perfect sense. The 666-09 chart makes even the weakest 12AT7 appear ‘Good’, while the 666-04 seems to run the tube more realistically. Let’s move along to the 12AX7, shall we? The first triode section is set up as follows.
This is a very frightening discovery. At tube chart 666-04, the triode halves are connected in parallel. Someone woke up for 666-09 and the Coletronics publication, but the extreme disparity in voltages and bias settings makes me believe the person responsible for these settings was still quite hung over. Let’s compare plate currents using the various tube charts.
According to a typical RCA Receiving Tube Manual, the 12AX7 has a plate dissipation of 1.2-watts. With a plate voltage of 100VDC, you are expected to have approximately 0.5mA worth of plate current. The Coletronics settings puts approximately -1VDC worth of bias and 100VAC on the plate of the12AX7 to achieve the 5.53mA plate current. No wonder the warning of not to test the poor tube for any longer than necessary is issued (see below)! What makes the mystery truly perplexing is the fact that many tube charts have you alter the voltages, bias, and meter shunt for the two halves of the 12AT7 and 12AX7! Why? Some roll chart settings appear to be grossly ‘incorrect’. Here is a sampling of how Eico will ‘suggest’ you set up the ubiquitous 6V6 on their 666 tube tester. I have recently added measured plate current in the far right column (seen in red).
This is almost scary how ‘bad’ Eico screws up these settings. Now, no one will lose an eye, nor will anyone ruin a perfectly good 6V6 using any Eico roll chart to test a 6V6 (or many other tubes). I still get an uneasy feeling when I realize the 666-05 roll chart wants me to connect my 6V6 as a triode, and the test current is really excessive. While over at the 666-09 settings, my 6V6 has far less bias and shunting of the meter than the 666-03 settings. Does this promote higher readings? Try it for yourself. The last thing to note is not on any roll chart, but seen in the owner’s manual. It clearly states a cautionary note;
(To avoid tube damage) when taking a meter reading, do not hold the merit button longer than necessary to acquire that reading.
This does not instill confidence, but is essentially a friendly reminder of something we should know already. The test voltages used are usually more reasonable in the later roll charts (but not always), and you can always check the test current through any tube to determine if the test parameters are reasonable. As I mentioned before, in a 6SQ7 the triode is rated for a maximum plate current of 1.1mA, and some tube testers can inflict over ten times that amount. This is why some testers have a similar warning about walking away from your tube tester with the poor tube being tested right to death. One of these days, I will have my own Eico 666 tube chart up here with a select few guitar amplifier tubes. The charts I supply will be thoroughly tested more than once, and will be ‘right’. For now, the few tubes listed above should help many of us. Another thing I always added to my Eico 666 ‘wish list’ was the capability of testing for gas or grid emission. Well, here is a tip from test equipment guru Norm Leal;
Here is a good, easy way to check for gas on an Eico 666. Set up the tester with grid control at “7”(zero volts bias). Open the grid circuit by moving the lever number ‘5’ to position ‘6’. The meter reading should drop. If meter reading goes up the tube is gassy.
Sounds simple enough. The results may not be as accurate as with a typical B&K tester, but this is still a very useful tip. Pass it along to other Eico 666 owners, please.
Finally, be sure to check out the page devoted to the Jackson 648. Here, I will share my opinions on whether or not the ‘Dynamic’ principle is any ‘better’ than a plain-Jane emission tester. I believe it is, but many other ‘authorities’ disagree. I also believe the Eico ‘Dynamic’ circuit is a little better than the Jackson ‘Dynamic’ circuit; Eico uses a filtered, negative DC grid bias supply. But, I am getting ahead of myself. Do check out that link right above, OK?