A valve tester can be useful in repairing electronic devices. The I-177 is an American military tester used in early World War II.

The I-177 is a valve tester used extensively by the USA Signal Corps and the US Navy during World War II. It is a small portable tester, for measuring valve characteristics, such as short circuits, cathode emission, gas, and mutual conductance. The I-177 was designed by Hickok for the Signal Corps, and ruggedised for military field use.

Picture 1: Dynamic Mutual Conductance Tube Tester I-177

Picture 2: Carry Case

It weighs 15.75 pounds, and is 5.75 inches by 15.5 inches by 8.5 inches. It was used individually, or was included in the Test Set I-56-K, together with a Voltohmmeter I-166, a Test Unit I-176, in a CS-130 case. The I-177 was superceeded by the TV-7 valve tester.

Since this tester is American designed and manufactured, the manual (TH 11-2627) refers to it as a “tube tester”. The manual is dated 3 August 1944, and is not the initial issued manual, as this one also covers the models I-177, I-177-A and I-177-B. In addition, it also covers the Tube Socket Adapter MX-949/U.

The I-177 is in a metal case with a removable protective lid. The case is splash proof but not water proof. All the leads and wires are inside. The US Army version is green, and the US Navy version is painted grey. All the controls and sockets are on the front panel. The mains cord is inside the lid, next to a Tube Data book. A circuit is riveted to the inside of the case, underneath the tester.

There are 14 sockets on the front panel to accept the valves for testing. The sockets are:

  A     UX4    4           80
  B     UX5    5           807
  C     UX6    6           42
  D     UX7    7           59
  E     K8A    8   American Octal   6V6
  F     B8G    8   Loctal       7A8
  G     K8A    8   American Octal    VR150
  H     B7G    7   Miniature 7 pin     0A2
  K     B7G    7   Miniature 7 pin     6X4
  L     B7G    7   Miniature 7 pin     6AQ5
  M         5   Acorn       955
  N     B8G    8   Loctal        1201
  O          6    Bantam      HY125
  D   Centre of UX7   2   Pilot Lamp

Table 1: I-177 Valve Sockets

The valve sockets are for the valves in use during that period, and do not cover modern, English or European valves. Notice that there are two Octal sockets, two Loctal sockets and three 7 pin miniature sockets. This is because the filament connections on some valve designs use different pins for the filament connections. Rather than switch the filaments from pin to pin using a high current mechanical switch, they have duplicated the valve sockets, with the filaments hard wired to different pins. This also simplifies the switch settings with the A and B controls. Despite the duplicated sockets, only one valve can be tested at a time.

If you are testing many valves, you may wish to make or purchase a “socket saver” which will protect the socket in the tester. In the case of the Octal socket, this is merely an Octal male plug (perhaps and old valve base), with an Octal socket, wired pin for pin. You could make your own adapters the same way, for English Octal, or European side contact sockets. You would have to determine the correct settings for each valve to be tested, and add those characteristics to the data book.

I was asked to repair a faulty unit, which caused me to retrieve mine from under the bench, to be used for comparison. My example is an I-177 and the faulty unit was an I-177B. There were some differences, mainly due to improvements.

Picture 3: I-177B front

The most obvious external difference, is the change from black to brown sockets, and the more robust power cord. The I-177 uses a figure 8 lead, whereas the I-177B uses a normal round rubber cable, containing two wires. Internally there were many differences, mainly due to tropicalisation, as is denoted on the front panel by the MFP stamp. The manual has extensive instructions on how to tropicalise a tester, if it has not already been done.

Picture 4: I-177 Inside

Picture 5: I-177B Inside

The transformer is in a sealed can, but exposed in the original. The wiring is plastic but cotton covered in the original. There are no valve clamps, but there are in the original. This appears strange, because the valves are upside down, and may work their way out. However, when the case is closed, and it is carried by the handle, the valves are oriented the correct way, so any bumps will not make them come out.

The valve sockets are in a line across the top of the panel, so they do not interfere with the controls, and can be easily seen during testing. The N and the O sockets are in the middle of the panel and appear as an afterthought. The buttons used to test the valves are scattered around the front panel.

There are two flying leads, one for the top cap connection (labeled CAP) and a separate special lead for the Acorn valve. There is an additional lead than can be used for valves with the plate on the top cap, for example an 807 or 1625. The top cap is connected to pin 6 of the 7 pin socket (socket C).

Located at the left hand side is the POWER ON and OFF switch, with the FUSE next to it. The fuse is actually an incandescent light globe. The mains cord is fixed and enters through the top of the panel. This is for 110 volts AC only, and is fitted with a 2 pin American plug.

The I-177 meter is calibrated in actual MICROMHOS, with three scales, 3000, 6000, and 15000 micromhos. There is a three position switch, located at the bottom centre, to select the three micromho ranges. The unit of measurement for resistance is the OHM. The inverse of resistance, is the unit of measurement of conductance, called MHO (also called Siemens). There is also a REPLACE (red colour) and a GOOD (green colour) scale, with a question mark in between. Diodes have a separate marking, indicated as DIODES O.K. which starts at 800 micromhos.

Picture 6: Meter

Figure 1: Shorts Chart

At the left of the front panel is a function switch, which selects TUBE TEST. It also has 5 different positions to test for SHORT circuits between valve electrodes. There is a neon lamp which will light to indicate a short. There are two terminals labeled NOISE TEST, which can be connected to a radio receiver, so that any valve noise can be heard. The notes on Nolan Lees web page, warn that the shorts test should always be performed first, as any short circuit inside the valve may damage the meter, or the bias potentiometer. The chart shows which electrode combinations are tested.

The emission of the valve to be tested, will vary with the filament or heater voltage. The voltage must be set accurately, and this is achieved by setting the “LINE” (input) voltage to a calibration mark on the meter. Press the LINE TEST button and adjust the LINE ADJUSTMENT control until the meter needle rests on the LINE TEST mark.

The valve test characteristics are read from the data book, and set by five controls. There is an A and a B control at the bottom left, which are set to different positions for each valve. The correspondence of the 12 positions of each switch, to plate, grid or cathode is not at all obvious. The next control is the FIL control which sets the filament or heater voltage, and this covers many voltages from 1.1 volts to 117 volts. It also has a BLST (for ballast valves) position and an OFF position. To the right at the bottom, are two controls labeled L and R, also set from the data book, to a value between 0 and 82. The R control sets the valve BIAS. The L control adjusts the meter sensitivity reading for the plate current (called SHUNT on later testers).

There are seven push buttons, to perform the test on the valve. The AMPL TEST is used to test amplifier valves. There are four buttons for testing rectifiers, labeled DIODE TEST, RECTIFIER STD, TEST OZ4, and 117N7.

When testing small rectifier valves, use the DIODE button. Do not press AMPL or RECTIFIER STD or OZ4 buttons, as the high voltage may damage the valve cathode.

The 117N7 valve has a pentode and a half wave rectifier in the same envelope. The switch settings test the pentode section. Then the 117N7 and the RECTIFIER STD button are pressed to test the rectifier section.

When testing a valve for gas, set the controls from the data book, and push the GAS No.1 button. Set the meter to 500 micromhos (the manual says 100, but I think this is an error), using the R control. Still holding the GAS No.1 button, push the GAS No.2 button. The meter should only move one division, it will move more if the valve contains gas. The gas test does not apply to rectifiers, only to amplifiers.

For testing pilot lights or small globes, use the center socket of the 7 pin valve socket (socket D). Turn the FIL control to set the lamp voltage. The lamp should light at the correct brilliance.

There is not an accurate test for a ballast valve. The FIL switch is set to the BLST position, and the SHORT switch is used to check for continuity.

Gas voltage regulator valves and thyratrons can also be tested. The controls are set as per the data book. The RECTFIER STD button is pressed and the R control is adjusted to the data point in the book. The valve should strike normally, as visible through the top of the glass envelope.

For testing magic eye valves, the FIL switch is adjusted normally. The L, R, and MICROMHOS switches are not used. Set the A and B switches, then press the AMPL button and the eye should open. Change the B switch as per the data sheet, press the AMPL button again, and the eye should close.

All the controls and test sockets are on the front panel. Inside there is a power transformer, which supplies all the necessary testing voltages. There are two rectifier valves inside. An 83 mercury vapor rectifier supplies the plate and screen voltage. A 5Y3 rectifier valve supplies the grid bias. All the components are attached to the front panel, so when the screws are undone, the complete unit can be withdrawn from the case. The main indicators for a valve test, are the meter, and the SHORTS lamp. Additionally, the FUSE will indicate excessive current flow, when it glows. The NOISE terminals (when connected to an external device) will indicate any mechanical valve noise.

The I-177 valve tester only works on 110 volts AC, and can be accomodate LINE voltages from 105 to 130 volts AC. The power transformer actually has a 93 volt primary, which allows for a possible low line voltage, the fuse and the 200 ohm LINE ADJUSTMENT control. When the LINE button is pressed, the secondary is connected to one of the diodes in the 83 mercury vapor rectifier, and then through a dropping resistor to the meter, plus a shunt. The LINE ADJUSTMENT control is used to set to the meter to the calibration mark, thus assuring the test voltages are correct.

Figure 2: Simplified Line Test Circuit

The FUSE is an ordinary single filament automobile lamp, in a bayonet single contact (B15S) base. These are used on older cars and older motorcycles, so they are readily available. Much is said about “only” using a #81 fuse bulb, “do not substitute”, and people sell these, as well as “solid state” replacements. The correct lamp is 6 volt 10 watt (6 candle power). It will glow at turn on, then fade away. Testing a heavy current valve, like a 5Y3, will cause the lamp to just glow. Also a 6 volt 5 watt bulb (3 candle power) will work, but glow a lot more. The lamp has mains on it, so it is recessed into the panel, and only the glass envelope is accessible.

Picture 7: Testing a 6V6 (fuse lamp is glowing)

The SHORT switch applies AC voltage to the valve, which is connected to different pairs of electrodes. The SHORTS lamp will light if there is any connection, or leakage below 250 K ohms. The lamp connections also go to two terminals, so that noise can be listened to. The presence of microphonics can be checked by tapping the valve, but the manual warns that several types of valve may fail if tapped.

Figure 3: Simplified Shorts Test Circuit

Rectifiers are connected to the transformer secondary, and the meter. There are three buttons, DIODE TEST, RECTIFIER STD, and TEST OZ4, which basically increase the test voltage on the rectifier, from 19, to 167, to 312 volts AC. Pressing the wrong button can overload and damage diodes or some rectifiers. I checked this, and ruined half of a 6AL5 dual diode.

Figure 4: Simplified Rectifier Test Circuit

The manual makes a distinction between a “mutual conductance” test and a “quality” test. The quality test gives a meter reading using the RED/GREEN (REPLACE/GOOD) scale, with a question mark between them. This is a fast bad/good test. The mutual conductance test gives an actual reading in micromhos, which is useful for selecting matched valves.

Figure 5: Simplified Balanced Circuit (Left), Simplified Valve Testing Circuit (Right)

In the Left Figure 5, the two transformer windings, each produce a voltage that flows through alternate halves of the rectifier plates P1 and P2. This current flows through the load resistor RL, then the balancing potentiometer RM, and through the meter. When it is balanced, the meter does not deflect, as the currents are equal and opposite.

In the Right Figure 5, the valve under test, replaces RL. The valve passes a steady current, and the meter still does not deflect.

The voltage E is the bias. When an AC voltage is superimposed on E, the bias effectively becomes smaller, then greater, and so the valve passes more, then less current. The meter current is now unbalanced, and the meter shows this as a deflection. The proportional deflection depends on the amplification of the valve which is effectively the mutual conductance.

Figure 6: Simplified Conductance Test Circuit

The meter deflection is set by the MICROMHOS range switch, and the L control. The bias for the valve is set with the R control. An AC voltage of 4.7 volts RMS applied to the grid, effectively adding to the DC bias.

Figure 7: Simplified Gas Test Circuit

The gas test is only for amplifier valves. The valve is set up for a quality test, and a series grid resistor is introduced. If the valve is good, there will be no grid current flow, and the grid resistor will make little change in the plate current. If there is gas in the valve, then there will be grid current flow, and the grid resistor will reduce the bias. This will result in more plate current, and it will show as a plate current rise on the meter.

The owner reported that while he was testing valves, he paused for a meal break, while leaving it turned on, and when he returned the tester was misbehaving. The faulty tester would work for testing amplifier valves, but would make the meter hit full scale for other tests.

The tester was powered up slowly with a VARIAC, and left on for a while, to ensure that the mercury vapor rectifier was working properly. The rectifier valves inside appeared to be good. The voltages appeared normal. There was no obvious damage or evidence of shorts. The LINE voltage function appeared to work. I spent some time figuring how the circuit worked, but it was difficult due to the many valve socket interconnections, and the complicated switching. I printed several circuit copies and traced wires in red pen, trying to determine the function, and which component may have failed. I checked all the resistors and the few capacitors, but none were faulty. At this point, the problem seemed elusive. In actuality, it was a situation that occurs only occasionally, as it was a fault that could not be seen on the circuit!

Figure 8: I-177 circuit

As I had reached an impass, I decided to approach it from another direction, by performing a calibration. I followed the Rick Putnak procedure.

The article on the web page, first addresses the internal components, stressing the balancing of the valves, or fitting new ones, new resistors, new capacitors, and cleaning switches. The term “balancing the valves”, refers to the function of each diode within each double diode rectifier. The 83 should have both diodes fairly close in emission. The 5Y3 also should have both diodes fairly close in emission. I tested these it another tester.

The article then gives some standard test settings for the switches. The voltages should be measured with an older 1000 ohms per volt meter, but if one is not available then a modern digital multi-meter should be shunted with a resistor, to simulate the loading. For 5 volt measurements, use a 5K resistor, for 50 volts use a 50K resistor and for above this voltage use a 250K resistor.

The first procedure is to check the LINE calibration. I pressed the LINE TEST button and adjusted the LINE ADJUSTMENT control until the meter needle was on the LINE TEST mark. There should be now, plus 150 volts DC on the plate of the valve, with a + or – 2 volt tolerance. I plugged my simulated 1000 ohms/volt meter into pins 3 and 8 of the octal socket, and pressed the AMPL button. The I-177 showed 150 volts, which was correct, but the I-177B showed 154 volts, which was too high. I fitted a parallel resistor across the series resistor (part #75), until the meter reading was correct for the correct plate voltage.

The plate voltage should only be present when the AMPL button is pressed, as was verified on the working tester. However, the faulty tester had the voltage present all the time! This was traced back to the actual AMPL switch. The switch is a block of wafers, held together with screws. It should have been open circuit (megohms) between the normally open contacts (N/O), but it was actually 60 ohms. There was no obvious short or sloppy soldering visible. The switch was disassembled and examined. A carbon track waws discovered across the N/O contacts, inside the block. The track was removed and re-insulated. The fault was now gone, and the tester behaved normally.

Picture 8: AMPL Switch

Picture 9: Carbon Track

I decided to finish the calibration procedure on both testers.

I checked the screen voltage, between pin 8 and pin 4. It was correct, being 130 volts DC, with a + and – 2 volt tolerance.

The signal voltage was checked, between pin 8 and pin 5. On both testers, it was measured at 5.0 volts RMS. The manual states 4.7 VRMS but Putnak says 5.0 VRMS, each with a 0.25 volt tolerance. The voltage is derived from a winding on the transformer, so there was no easy way to adjust it. It was left unchanged.

I checked all the filament voltages on both testers and they were correct.

I measured the DC bias between pin 8 and pin 5. The bias voltage is checked at three positions of the R control, which has graduations from 0 to 82. At position 0, it should read 0 volts. At position 18 it should be -3.0 volts with a + and – 0.1 volt tolerance. At position 82 it should be -42 volts. The voltage on the I-177B was only -30 volts at full rotation (position 82). A 26K resistor was placed across the R control, which corrected the full rotation voltage to 42.5 volts.

I checked the balance of the L control. I connected a 10k ohm 10 watt resistor between pins 3 and 8, which simulates a valve. The meter ideally should read zero, for all positions of this control. The I-177 was about 2 small divisions out. The I-177B was good for 82 to 60, but the error got progressively worse from 60 to 0, but was not excessive. If this was required to be fixed, the control would need to be replaced or rewound. It is a dual 150 ohm per section, wire wound potentiometer, so replacing or rewinding it would be a problem. Both controls were left as they were.

The mutual conductance accuracy was checked. This requires a 10mA AC current to be connected between pin 3 and pin 8. A small transformer was wired up to a VARIAC. The secondary provided 30 volts AC and a 3.6 K resistor was inserted. The VARIAC was adjusted to ensure there was 10mA flowing, as measured with a multi meter. Both valve testers gave a 2000 Gm reading, which was correct, on all three MICROMHOS ranges.

To simulate a short circuit, a 250K resistor was connected between pin 5 and pin 8. The SHORTS lamp was alight for SHORT switch position 3. Other positions were tested by using to the SHORTS chart, and all corresponded.

To test for gas, the controls were set as per the data book for an amplifier (I used a 6V6). The L control was set at Gm (position 60), and the R control at 82. I pressed the GAS No.1 button, and readjusted the R control until the meter read 500 on the 3000 scale. While keeping the GAS No.1 button pressed, I pressed the GAS No.2 button. The needle moved only one division, indicating a good valve. I then connected a 1 megohm resistor between pin 5 and pin 7, which simulated a leak between the grid and heater. I pressed both buttons, and the meter needle moved four divisions, indicating gas in the valve. This was correct on both testers.

To check the rectifier calibration, I set the R control (bias) to zero. I measured the voltage between pin 3 and pin 8. I pressed the DIODE TEST button and there was 20 volts AC. I pressed the RECTIFIER STD and there was 170 volts AC. I pressed the TEST OZ4 and there was 310 volts AC. This was correct on both testers.

I then connected a 1N4007 silicon diode and a series 1 K resistor (5 watt) between pin 3 and pin 8 (cathode). I set the L control to 61 and pressed the DIODE TEST button. The meter deflected to DIODES OK. I then set the L control to 68 and pressed the RECTIFIER STD button. Again, the meter deflected to DIODES OK. I changed the resistor to a 10 K (5 watt) value. I set the L control to 74 and pressed the TEST OZ4 button. The meter deflected to DIODES OK. This was correct on both testers.

Both testers were now calibrated.

When new American valves were introduced, extra base configurations were required, so an MX-949/U adapter box was produced, which provided 9 additional sockets.

 X1           7     Septar       832
 X2           7     Acorn       6F4
 X3      8A     8     American Octal    6B8
 X4      8G     8     Loctal        7E5
 X5           8     Subminiature     CK506
 X6           8     Subminiature     5636
 X7           7     Subminiature     5702
 X8      7G     7     Miniature 7 pin     3S4
 X9      9A     9     Miniature 9 pin     12AT7

Table 2: MX-949/U Valve Sockets

This is housed in a similar metal box. It has a flying lead which plugs into the I-177 octal socket. There is no electronics in this box at all, it is merely to provide extra sockets. There is a slight change in the design philosophy, in that the sockets are not hard wired to the flying lead. There is a patch panel, so that any pin from the I-177 octal socket, can be connected to any pin on the MX-949 sockets. This allows for the newer valves, and also for any valves which may be introduced after.

Picture 10: Tube Socket Adapter MX-949/U

Figure 9: MX-949 circuit

Additionally, if you wish, you could decide not use the two octal sockets on the I-177, and just use the octal socket on the MX-949, and patch it accordingly. Similarly for the other duplicated sockets on the I-177 tester. There is another benefit, in that a meter may be inserted in any lead to monitor current, if desired. You can substitute another filament supply, if the valve requires more current or a different voltage that the I-177 can supply. You can also substitute a different signal voltage and frequency if desired.

This is a newer valve tester than the I-177 and it has the inclusion of more modern valves, the deletion of older valves, and the simplification of controls. It follows the same lines as the I-177 but the layout that is more convenient. The valve sockets are still across the top, but the duplicated octal, loctal, and B7G sockets, and the Bantam socket have been removed. The sockets from the MX-949 have been included, the B9A, and the two subminiature sockets, but the large 7 pin socket, and the long lead subminiature socket have not.

The test buttons are all in a line down the right hand side. There is no separate button for the 117N7 valve. The meter is calibrated with a 0 to 120 scale which is useful to compare the valve to a 100% rating. There are charts in the manual and on the net, which give a cross reference to mutual conductance values. The lid has the valve data book, a pin straightener and a couple of tube adapters. There is also a pilot lamp.

The SHORTS and the RANGE control have been combined. The L and R controls have been renamed BIAS and SHUNT. The FILAMENT switches have been expanded, with two others added to the existing voltage switch, to select the actual filament pins, and labeled with letters of the alphabet, set using the valve data book. The two valve pin selector controls have been expanded to provide 5 independent selectors for GRID, PLATE, SCREEN, CATHODE, and SUPPRESSOR electrodes, and each have 9 positions. This tester is more versatile than the I-177 valve tester.

Picture 11: TV-7B/U Tube Tester

The I-177 is a nice portable tester in a rugged box. It does not test modern valves, and only covers American valves use in World War II. It can test more valves with the MX-949 adapter box. The physical layout could have been improved slightly, by arranging the valve sockets in a straight line, with the duplicated sockets underneath each other. Also, the test buttons could be grouped together, as was done in the later TV-7.

There is the easy possibility of damaging a valve, if the wrong button is pressed. But of course, you can also damage the valve by setting the wrong filament voltage! The 4.7 volts RMS applied to the valve grid, may be too large for some modern valves.

The quality could have been better by the use of ceramic valve sockets, and ceramic switches. This may also reduce the possibility of carbon tracking in switches and sockets.


TM-2627 Manual (from BAMA)

Rick Putnak, Signal Corp I-177 Tube Tester Calibration

Nolan E. Lee, Operation, Calibration, and Maintenance of US Military Tube Testers

Ray Robinson