Developing a test configuration for the Eico 666/667 Tube Testers by Mike at MDBVentures.com 11 Mar 2011 (revised 31 Mar 2019) http://www.MDBVentures.com - Great prices on great tubes! The 666/667 Eico tube testers are very nice as far as the electronics goes, although with a few caveats... see Eico 666/667 Testers: Buying, Restoring and Calibrating the Eico 666 Tube Tester http://www.fourwater.com/files/restr666.txt Unfortunately the quality of the electronics did not carry through to the documentation. The tube setup charts tend to have less than optimal setups for testing, or in some cases outright errors. Some of the errors can destroy the tube being tested. Heater values should always be checked as the Eico chart has a few errors with the heater settings which can destroy the tube. If you will be using the 666/667 Eico tube tester to test tubes, you should get a copy of the GE Essential Characteristics manual for tubes. The RCA Receiving Tube Manual will work too, but it is not as complete. You can get reprints of the manuals from Antique Electronics Supply http://www.tubesandmore.com and other electronics suppliers. In the manual you will find the critical information needed to test the tubes. The most important parameters you will need to know are the filament voltage, the plate voltage, and the plate current. Usually the manual will also tell you the grid voltage to use for the listed plate current. You will also need to know the tube pin-out, which the manual will provide, although you will have to flip to the back of the manual to find the diagram referenced in the tube chart. The Eico Instruction manual provides a starting point for developing the tube test for tubes that the Eico tube setup chart doesn't cover. Unfortunately it doesn't expand enough on what is being done to test the tube. As a result, some of the tests that have been developed are less than optimal, or just plain incorrect. To develope a test for a tube, find the tube in the GE or RCA manual. The manual will list a number of characteristics. Determine the filament voltage for the tube and set the Filament knob on the Eico tester to the closest filament setting. Next find the tube pinout diagram for the tube and set the 1 through 9 (12 on the 667) and C levers. To do this, you will need to know which lever is used for which tube element. At the bottom of the levers on the 666 (above the levers on the 667) is a number. This number is the pin number on the tube that the lever controls. There are also numbers listed along the slots for the levers. These numbers refer to the tube elements that the the lever can be attached to. Note: If you are going to do this a lot, put a piece of Scotch tape next to the levers and mark down the information as a reminder. A single letter usually works fine as a reminder. 1 - common (ground) (C) 2 - filament (F) 3 - screen (S) 4 - plate (P) 5 - grid (G) 6 - no connect/signal diode (X) Note: If there is more than one section in the tube, you may have to develope separate test sequencies for each element. See the notes below on how to setup for testing these types of tubes. 1(C). This is used to connect one end of the filament to the filament transformer. It also is used to connect the cathode to the common ground. For tubes with a suppressor grid, the suppressor grid is also connected to ground via this setting. 2(F). This is used to connect the other end of the filament to the filament transformer. Warning: Be careful, some tubes have a center tap connection on the filament. If the tube you are testing has a center tap, then set that pin to position 6(X) (no connection). Do not set the center tap to the common ground (C) or filament (F) position. Alternately, you can set the filament center tap to common ground and the ends of the filament to the filament (F) position. If you do this, you must make sure that you set the filament voltage to 1/2 the heater voltage. WARNING: This only works for tubes with a true center tap connection (like the 12AX7). Do NOT do this for tubes that use a filament tap as a ballast connection, (like the 35W4). It is generally best to not test this way as it is easy to get things confused and end up blowing the heater. Also some tubes like the 35W4 have a tap on the heater, but it is not a center tap. The tap is used to drive a pilot lamp. Trying to use it as a center tap would burn out the heater very quickly. It is better to just get into the habit of using the full end to end heater connections and not use the tap on the tube heater when testing (remember to set the filament tap to position 6 - no connect). 3(S). This is connected to the screen grid of the tube (if it has one). In some low power tube tests, plate current and voltage needs to be limited. Some of the tests developed by Eico use the screen grid setting to do this. The screen grid selection will use the next lower voltage setting from what the "V" lever is set to. This is a perfectly acceptable solution to the problem, so don't worry if you see it in their charts. Sometimes though, using the screen setting for the plate is not the optimal way to test the tube and is normally only usable as an alternative with triodes. Also, the testing method described in the following section assumes that the tube plate is tied to the plate position, not the screen position. Tying the plate to the screen grid setting will give different results. 4(P). This is connected to the plate of the tube. If you are testing multiple sections in a tube, only connect one plate at a time. Connect the other plate to the No Connect (6) position. See the notes below for a special case exception to this. 5(G). This is connected to the control grid of the tube. If you are testing multiple sections you can connect the control grids of all the sections to the grid position, or you can set the non-test section grids to common ground (preferred). 6(X). This setting has several uses. It is used to disconnect elements in the tube that you do not want connected to one of the other settings (such as when testing tubes with several sections). It is also used to test signal diode sections of the tubes. When the "S" lever is set to selection 6, this setting is used for testing signal diode sections of the tubes. Note: When testing a signal diode section of a tube, the plate connections of any other sections in the tube should be connected to the common ground (1). "S" Lever settings: This lever controls the plate and screen voltage and the grid signal multiplier. 1. 45 volts plate - low grid signal (0volt to 5volts) - 15 volts screen 2. 90 volts plate - medium grid signal (0volt to 15volts) - 45 volts screen 3. 180 volts plate - medium grid signal (0volt to 15volts) - 90 volts screen 4. 180 volts plate - high grid signal (0volt to 45volts) - 90 volts screen "V" Lever settings: This lever controls the plate current selection. 1. 160ma(0) to 40ma(100) 2. 40ma(0) to 10ma(100) 3. 10ma(0) to 2.75ma(100) 4. 2.75ma(0) to 0.75ma(100) 5. 0.75ma(0) to 0.3ma(100) 6. Signal diode Note: As with the Pin selection control levers, if you do this alot, it can be handy to write down the voltage and current settings on a piece of Scotch tape next to the "S" and "V" levels as a reminder of the setting values. The plate control potentiometer works together with the "V" lever settings to set and read the plate current for the tube. The Eico instruction manual contains the charts needed to determine the plate current. The currents noted above are for when the plate control is set between 0 and 100. The Eico meter measures current. It is normally connected in series with the plate, but it can also be connected in series with the grids to measure current through the grid. The meter just measures the current of the connection it is attached to with the push buttons (1-9/12 and C). The meter is reasonably accurate at measuring the current. The meter measures the current by measuring the voltage across a shunt load resistance. The shunt load resistance is determined by the "S" lever setting and the Plate control. S1=3.57 ohms, S2=14.27 ohms, S3=56.97 ohms, S4=227.97 ohms, S5=910.97 ohms, S6=910.97 ohms. The meter itself has a 1000 ohm movement that reaches full scale at 200uA. In addition to the shunt resistance, there is a 3K ohm plate control that inserts a variable series resistance in the meter circuit which is used to read the plate current. To set the plate current for the tube test, read the plate current from the GE tube manual (2nd page) and set the lever and the plate control to the indicated value. If the tube current is more than the tester can be set for, just set it to the maximum current value. If you are using a different manual than the GE manual, use 2/3 of the maximum current if specified, or use the normal operating current if that is provided. Note: in some cases this won't work properly and you will have to set-up a different current setting. This is ok as long as the current is within the normal linear portion of the tube's operating range. Get the tube voltage for the test from the GE manual (2nd page). If you are using a different manual, set the voltage to 2/3 of the maximum plate voltage if specified, or the normal operating voltage if that is provided. Set the "V" lever to the voltage setting that is lower than the voltage determined from the manual. If the grid bias voltage is specified and it is higher than 15 volts and the plate voltage selection you are using is 180 volts, set the "V" lever to position 4. Note: the actual current flow through the tube is determined by the combined settings of both the plate and grid bias controls. The "S" and "V" levers and the plate control configure the circuit so that it will operate at the desired current range and so that the meter circuit will properly measure the current relative to the 100 percent reading on the meter. The actual current through the tube is set by the supplied voltage (determined by the "V" lever) and the grid bias control. While the "S" lever can have a minor effect on the plate current, it is minimal. The primary purpose of the "S" lever is to set the current range in which the meter will be reading (similar to the current knob setting on a VOM/DVM). The Eico testers use a short tail constant voltage supply. That just means they use a low resistance voltage source so that the voltage on the plate will be relatively constant regardless of the current. That means that the current through the tube is primarily controlled by the plate voltage supply (45 volts, 90 volts or 180 volts), and the setting of the grid bias which controls the resistance between the cathode and plate (and thus the current through the tube). Note: Some of the Eico chart tests set the "V" lever to position 4 when it should be at position 3. While you can test at position 4, the grid voltage is much higher (45 volts instead of 15 volts), which means that the sensitivity of the grid control is greatly reduced. Small changes of the control can have large changes in the test result. Don't use position 4 of the "V" lever unless you really need it (ie the grid requires a large voltage, 15V or higher, to operate). To start, set the grid control knob to 7, unless you are testing a rectifier tube in which case you should set the grid control knob to 0 and leave it there for the test. Note: A signal diode is tested differently than a rectifier tube (see the signal diode section below). Some Eico test configurations set the tube plate to the screen grid setting on the tester. This means that the plate will have the lower screen voltage applied to it, not the plate voltage. On some of the earlier tester setup charts, the test settings for some tubes such as the 12AX7 settings use the screen grid setting for the plate. while this may seem odd, it doesn't hurt anything. All it means is that the voltage applied to the plate will be less than if the plate selection on the tester was used. It doesn't hurt anything. That being said, in some cases it can be useful to use the screen setting to provide the plate voltage. For space charge tubes it is actually required because it is the only way to get the plate voltage low enough to properly test the tubes. The other situation is that it can be used to increase or reduce the control grid voltage selection since the V selector controls both the plate voltage and the control grid voltage used to test the tube. Although in most cases it doesn't really make any difference. Just keep in mind that if you use the screen grid position to provide the plate voltage that it will be less than the voltage selected on the V control. (See the "S" lever settings noted earlier.) I have seen more use of the screen setting in the Eico charts used for the plate than would normally seem reasonable. My guess is that possibly the original setting was the result of a typo. Whoever followed up with newer tests didn't understand that, so they kept using the same setup configuration thinking it was the way the test should be done. Most of those test settings could be redone to use a less confusing configuration, but they work ok as is, so it doesn't really hurt anything. The later charts I've seen have mostly corrected this to properly use the plate position. The main issue is that if position 4 of the "V" lever is used when the grid voltage is normally in the 0V->15V range, then the grid control is going to be less accurate (only 1/3 of the control range is being used). The only difference between position 3 and position 4 of the "V" lever is the grid voltage range provided by the Grid control (3=0V->15V, 4=0V->45V). All that being said, there is a reason to not use the screen grid setting on some tubes, especially power tubes (like the 6L6). They have a screen grid who's purpose is primarily to reduce the possibility of parasitic oscillation by providing a "screen" between the cathode and the plate. Tying both the screen and the plate to the the screen grid selection (or the plate connection for that matter) defeats purpose of the screen grid, and it can potentially allow the tube to go into parasitic oscillation. That has the potential to seriously disrupt the test results, causing a perfectly good tube to show as being bad. (see http://www.fourwater.com/files/eico666parasitics.txt for more information on this) Testing the tube: The ideal method is to first test the tube using a curve tracer and then plug the tube into the Eico tester and adjust the grid and plate controls to achieve the same relative result as indicated on the curve tracer. Since most people don't have a curve tracer or an engineer to interpret the results, other methods must be used to determine the tube test results. Note: Always adjust the Line control after you plug the tube into the tester. The tube's heater current drain can change the tester voltages. The purpose of the line adjust is to compensate for the change. (When the merit level is pulled to test power tubes, they can load the tester voltage down even more. The test settings are normally tweaked to adjust for this change). If you have access to a good calibrated tester (such as a Hickok tester), you can use it to test the tube to get a good reference. Place the tube in the Eico tester and adjust the grid control to get the same relative reading. In some cases you may have to tweak the plate control to get the desired reading. If you don't have access to a calibrated tester, then the next best thing is to get several new tubes (preferably from different manufactures) and average the test results. Typically new tubes should test between 100% and 120%. Although some tubes may show a wider range of test results. Adjust the grid control to get the desired test results on the new tubes, average the results, and use that as your test setting. Again, you may have to tweak the plate controls to get the desired test results. If you don't have access to a calibrated tester or known good condition new tubes, then it becomes a bit trickier. But all hope is not lost. You can get a pretty good idea of the tube condition. This requires that you have a copy of the GE Essential Characteristics tube manual. Set the plate current and voltage controls as noted previously. Now comes the tricky part. For most tubes, the grid control will control the signal presented to the tube control grid. The higher the setting, the higher the voltage. The GE manual states the grid voltage in DC, but the Eico tester places an AC signal on the grid. However, it works out that generally the AC voltage comes close to providing the same relative results as the DC voltage listed in the GE manual. The grid control is very approximately linear with the grid voltage being about 5VAC at the 100 setting when the "V" lever is set to 1 or about 15VAC at the 100 setting when the "V" lever is set to 2 or 3. The grid voltage is about 45VAC at the 100 setting when the "V" lever is set to 4. Determine the desired grid voltage from the GE manual and set the grid control knob to the apppropriate value. If you have a digital volt meter handy, set it to AC volts and connect it between the grid pin and the cathode pin (use one of the empty sockets). With the merit lever pulled, adjust the grid control to set the grid voltage to the desired value. The 666/667 tube tester meter will now read the test result. This is a very rough estimate, and only works with tubes that fall within the bounds of what the tester can easily handle. Power tubes, which can significantly load down the tester, may not give as accurate test results. Also keep in mind that the Eico tester does not present the same conditions as indicated in the GE manual, so don't expect the exact same results. Some tubes, such as pentagrid converts are particularly difficult to test and may not work with this method. Note: The grid voltage can drop on tubes that draw heavy grid current or on tubes that have high plate current when being tested. This can affect the grid voltage, which is why it is best to measure it rather than guessing based on the control setting. Also power tubes can drop the over-all voltage of the tester which can affect the filament voltage, the grid voltage and the plate voltage, which of course means that the test results will not be accurate. The test configurations must compensate for the voltage drops that will be encountered with these tubes. Because of the high variablity of the way tubes interact with the testers, the above method is not going to be highly accurate without access to a reference tube. However, it can at least tell you if the tube is not likely to function very well or is just plain dead. Note: For rectifiers, you don't need to adjust the grid control, just set the grid control knob to 0 and if needed, adjust the plate control to get the desired reading. While all the above description works, it isn't exactly how the Eico Tube Testers test the tubes internally. If you are interested in what goes on under the hood. See the file http://www.fourwater.com/files/how-eico666-works.txt Most notably the Eico tester uses AC supply voltages on the plate and the grid, not DC as the tube would normally be used in a circuit. In addition, the grid signal is supplied in the same phase as the plate (and screen) AC voltage source, which means it is positive relative to the cathode during the time when the tube is in the conducting portion of the waveform. While a tube is normally characterized to operate with a negative DC voltage on the grid, which is then modulated with an AC signal, the tube will work with a positive AC signal on the grid. The test will still tell you the emission results, and If the grid is able to control the conduction, but it is not the normal way a tube is used. Why you won't get exact readings on tube testers: The GE essential characteristics manual shows design center values. In almost all cases these will be settings that are not directly setable on the Eico. As an example, the 12AX7 design center parameters in the GE manual show a voltage of 100 volts and 250 volts. The closest settings the Eico has are 90 volts and 180 volts. That means for the given grid bias voltage (which sets the plate resistance), there will not be a direct match between the specified design center current and the measured current. Note: One thing that can be confusing is understanding the plate resistance value. The plate resistance stated in the GE essential characteristics manual is not a static plate resistance value. It is derived from a dynamic measurement of the tube current at the design center. That is why if you divide the stated plate voltage by the stated plate current, you do not get the stated plate resistance. eg for the 12AX7 : 100V/0.0005A=200000 yet the manual says 80000. So why the difference? The plate resistance is a dynamic plate resistance. That means it is determined by measuring the plate current at a given plate voltage, then changing the plate voltage and measuring the plate current again (DeltaV/DeltaA=PlateResistance). For more information on understanding tube measurements see the file: http://www.fourwater.com/files/mutualconductance.txt Multi-section tubes: The above testing assumes a single purpose tube. Tubes that have two or more sections require some additional action. When testing multi-section tubes, there are several ways you can approach the testing. The easiest is to set all the pins of the tube section not being tested to either common ground, or no connect (selection 6). This keeps the tube section out of the test. Some test configurations will pre-set some of the pins to the normal test position so as to minimize having to change the levers when testing the tube. When testing a large number of tubes this can be an advantage. In this configuration, the plate is usually connected to selection 6 (no connect) so that it is not involved in the test. For even faster testing, some tube testing configurations even set the plate controls for both plates to the plate (or screen) selection. The down side to this method is that if the tube sections are significantly different, it can significantly affect the tube test results. Thus where this is normally only used is with tubes that have identical sections being tested. The Coletronics tube setup chart for the Eico testers normally keeps the tests separate, while the early Eico factory test setup chart tends to combine the settings. One feature that the Eico test setups like to use when testing twin triodes is to set all the levers to the full test setup, including the plates. The test button for the associated plate is then depressed and the merit lever pulled to test that section. When done, the other plate test button is pushed and the merit lever is again pulled to test that section. This makes the testing go faster because there is very little that needs to be changed to test both sections. The down side to this is that both sections interact to some extent during the test. So if one section has significantly lower emission than the other section, it will pull the test reading down giving a false reading for the better section. Still, this is a nice way to do some quick pre-screening of the tubes. As long as the two sections test close together, the test results will be reasonable. It is only when the two sections have significantly different emissions that the test results will be worse and if that is the case, you may want to discard the tube anyway. Signal diodes: The signal diode testing on the Eico is handled in a special way. Normally signal diodes operate in the microamp range while most other tube elements operate in the milliamp range. Since the normal test selections of the tester don't operate at the desired microamp range, position 6 of the "S" lever is used to set the tube tester for this special case. For this test the plate of the diode is set to selection 6 which has special circuits to operate the signal diode's special current testing. In this test, the plate supply voltage is disconnected and instead the Grid supply voltage is applied to the plate via a 24K Ohm resistor (R15). This places a much lower plate voltage on the diode and thus much lower current flow through the diode. The plate control is still adjusted to get the desired test results. Normally the grid control knob is set to 7 for this test rather than 0 as it would be for a rectifier tube test. Although you may notice that tubes that have the signal diode combined with a triode or other element in the tube may have the grid control knob left at the setting for the triode test. Since the signal diode does not have a grid, the grid control has minimal effect on the test and can normally be left at that position. This reduces the amount of changing that you have to do when testing the various sections of the tube. However, the controls will sometimes have slight effects on the test results anyway, especially with signal diodes that operate at low currents. Usually this difference is minimal though. If you have a copy of the Eico Instruction manual for the tester, you may notice that the procedure description given here is not exactly the same as the one given in the manual. The Eico manual relies on a known good tube being tested with known test results to be able to create the tube test setup. If you don't have a known good tube, the method described in the Eico manual is not much help. The method described above will come closer to the desired results if you don't have a good test reference to start with. Tube Grid Elements: It can sometimes be confusing to know which grid is which in a tube. As a general rule, the grid closest to the cathode is the control grid. This is the grid that is primarily used to control the tube operation. (For the exception, see the Space Charge tubes section below.) The second grid is normally the screen grid. This is used to help accelerate the electrons towards the plate thus making the tube more efficent. The suppressor grid is normally the grid closet to the plate. This grid is normally tied to ground, and often is internally tied to the cathode to achieve this without requiring an extra pin connection in the tube. The suppressor grid helps to stop electrons that are near the plate from going back to the control and screen grids. The suppressor helps to give the tube a wider range of operation. There are also special tubes that have additional grids for special operations. Pentagrid converters have more grids to allow mixing of signals and are used most commonly in frequency converters. They are normally found in radios used to convert the RF signals into the intermediate stage signal that is feed to the IF filters that are used to separate out the desired signal from all the RF picked up at the antenna. Pentagrid converters are difficult to test because of the construction and the nature of the way they work. There are many different ways used to test them, but many of the tests tend to indicated perfectly good tubes as being bad. I've seen a number of brand new tubes being shown as bad by some of the testers. The reality is that for most tubes, the primary failure is the emission of the tube going down (worn out), or the heater not working. All the other aspects of the tube tests are to find physical damage or physical failure (as an example, a grid connection coming loose. If a tube used to work and it was not subject to being dropped, then normally the only things that will change is the emission (how many electrons the tube can emit), and the heater (it either works or it doesn't). The other thing that can happen is the tube can become gassy. A gassy tube will usually cause a reduction in tube emission because the gas interfers with the tube operation. Because of this, doing a simple emission test on a pentagrid converter can often be far more informative and accurate than trying to do a full test on the tube given the difficulty of getting accurate test results with a pentagrid converter. A simple emissions test is done by tying all the grids to the plate and test it as a rectifier diode. Although sometimes tying the control grid to the cathode works better. Space charge tubes: Space charge tubes are tubes that use very low voltages (typically 12 volts) on the plate and heater. These tubes were designed for use in cars with 12 volt batteries. Because the voltage on the plate is so low, there is usually not enough voltage for the tube to operate properly. To help compensate for this, on some of the tubes, the first grid on the tube is a special accelerator grid and is usually tied to +12 volts. This helps to pull more electrons from the cathode, however many of them end up leaving via the space charge accelerator grid, so these tubes are not very efficent and have a very small dynamic range. When testing space charge tubes, there are several things to consider. Since the tester doesn't have a 12 volt plate selection, you must use the screen selection for the plate to keep the voltage down (The "V" lever is set to position 1). This will cause the Eico tester to use the 15 volt screen grid voltage on the plate, which is close enough for a space charge tube test. Set the space charge accelerator grid to the screen selection as well. The rest of the set up goes as previously described. Note: Some of the low voltage tubes such as the 12U7 don't use the extra space charge grid to pull electrons from the cathode. They simply operate with what little current is available from the cathode. Usually these tubes have higher filament current requirements to boost the number of electrons boiled off the cathode. Special tubes: Some tubes don't have the normal pin connections. This is especially true of transmitter tubes which often require special construction. You can still test these tubes by using jumper wires to connect the appropriate socket pin on the tester to the element connection on the tube. Be careful though. Many transmitter tubes require large amounts of heater current and plate current. It is not recommended that you try to draw more than about 5 amps from the Eico heater circuit or it may damage the transformer. Also since many transmitter tubes operate at large plate currents, the maximum plate current of 160mA on the Eico may not be enough to adequately test the tube. see the file http://www.fourwater.com/files/spacecharge.txt if you are interested in learning more about how space charge tubes work. Subminiature tubes: Extreme care must be used with these tubes. Many of them normally operate with only microamps of plate current and some require very low voltages (space charge operation). Operating the tube at the plate current levels normally used in the tube test circuits can damage the subminiature tubes in short order. If you must test these tubes, always set the plate lever to the screen selection to minimize the current and voltage to the tube and make sure that the V and S levers are set properly before pulling the merit lever. Keep the amount of time that the test lever is pulled to a minimum. Only use it long enough to get a valid reading. Also see the companion files: http://www.fourwater.com/files/restr666.txt http://www.fourwater.com/files/666-667-mod.png http://www.fourwater.com/files/eicotesting.txt http://www.fourwater.com/files/eico666-667-repair.txt http://www.fourwater.com/files/mutualconductance.txt http://www.fourwater.com/files/eico666meter-power-notes.txt http://www.fourwater.com/files/eico666tester-meter-check.txt http://www.fourwater.com/files/how-eico666-works.txt http://www.fourwater.com/files/eico666parasitics.txt http://www.fourwater.com/files/testertypes.txt http://www.fourwater.com/tubeinfo.htm Visit us at http://www.MDBVentures.com - Great prices on great tubes!