Troubleshooting, Restoring and Repairing The EICO 666/667 Tube Testers
by Mike at MDBVentures.com 31 March 2019 - updated 31 Dec 2023 -med

  http://www.MDBVentures.com - Great prices on great tubes!

Be sure to see the companion file http://www.fourwater.com/files/restr666.txt
which decribes how to calibrate the testers as well as what to look for 
when buying one. It also describes several modifications you can make to 
help protect the 666/667 from potential damage.

The file http://www.fourwater.com/files/how-eico666-works.txt provides a description
of how the testers work to test tubes. 

Also the file http://www.fourwater.com/files/eico666parasitics.txt describes a potential
problem you may encounter with the 666 and 667 testers, for which there is no easy fix. 


I do have a few replacement parts for the 666 and 667 available, including the Grid 
and Plate control pots noted in the article below. To see the parts I have available, 
go to http://www.fourwater.com/equip/eico666-667-parts.txt

**************************************************
Note: This is incomplete. I'm still working on it. 
I will upload the updates as I add to the article. 
*************************************************

So, you got your Eico 666 or 667 tester calibrated (see the restr666.txt file) 
and it still doesn't seem to work properly, or maybe you just can't get it to work 
at all. Here is a detailed set of procedures that describes how to check your 
Eico tester and fix the most common problems. 

These tests are valid for the Eico 666 and the Eico 667 tube testers. 
The primary difference between the two testers is that the 667 replaces the preoctal 
sockets with Compactron and Nuvistor sockets so that you don't have to use the 
Eico 610 adapter to test Compactron and Nuvistor tubes. There are a few other minor 
changes between the testers depending on when the 667 was made, but for most of the 
circuits they are the same. 

Note: This article assumes you have already performed the checks noted in the 
restr666.txt file, so if you haven't down-loaded and read the 
http://www.fourwater.com/files/restr666.txt file yet, do so before you continue 
on with this information.
 
You should also read the eico666tester-meter-check.txt file to learn how to test the 
meter circuits. It is preferable that the meter calibration is working correctly, 
however even if you have a problem with the meter calibration circuits, this test 
can still be performed. The meter calibration / leak test circuits are mostly 
separate from the rest of the tube test circuits (they share some switches and derive 
power from the 50 Volt section of the filament transformer). 

Also note that the test description below assumes you are doing the tests in the 
order written. If you do them out of order, the resulting problem description 
may not be correct for the test results seen. 

***WARNING***
Caution! Be careful, there are lethal voltages present and exposed with the tester 
out of it's case. If you don't know what you are doing or feel uncomfortable doing 
the work, get a qualified technician to do it for you. 
***WARNING***

Before you start, consider this:
Sometimes a switch can go bad over a long time of sitting, especially in a damp location. 
This is caused by build up of grime on the contact surfaces, similar to how dirt builds 
up on your TV screen over time. This can often be "fixed" by just moving the switch to 
clear the grime off the contact surfaces. So first, try moving the associated switches 
a bunch of times to see if it will fix the problem. 

Some of the following tests require the tester to be removed from the case. 
You will need an AC/DC volt meter capable of reading 0 volts to 250 volts. 
The meter should be 20,000 ohms per volt or better. (A digital volt meter should be fine.)


How to remove the tester from its case:

Remove the 14 sheet metal screws around the outside edge of the tester. 
(Don't remove the four screws next to the paper scroll knobs, they hold
the paper scroll assembly to the faceplate.) 
Set all the levers to position "1". 

Holding the front panel firmly to the case, carefully turn it over and 
set the faceplate on a cushioned flat work surface. 
Now carefully pull up on the case. You will probably need to wiggle it a bit
to clear the wires that will likely stick out a bit. 

***WARNING***
Caution! Be careful, there are lethal voltages present and exposed with the tester 
out of it's case. If you don't know what you are doing or feel uncomfortable doing 
the calibration, get a qualified technician to do it for you. 
***WARNING***


Putting the tester back in it's case:

To install the case back on the tester, set all the levers to position "1" and 
place the tester on it's face on a cushioned surface. With the case handle towards 
the paper scroll assembly, carefully place the case on the tester faceplate. 
You will probably have to wiggle it around a bit to clear some of the wires.
Make sure that there are no wires pinched between the case edge and the tester
faceplate. 

Holding the front panel firmly to the case, carefully turn it over and 
set the tester down on a flat surface. Reinstall the 14 screws around the edge 
of the tester. 


How to remove the paper scroll assembly:

Remove the four screws next to the paper scroll knobs. 
The scroll assembly will now come out. Be careful not to rip the paper. 
Also be careful handling the assembly. The tension of the metal bracket is what 
holds the paper scroll in the bracket. 
Note: When you reinstall it, make sure that you have the scroll assembly in 
the right way up so that the chart is facing the correct direction before you
put the screws back in the assembly.


Sticky Push Buttons Switches:

A common problem with kit builds is sticky push button switches, although I've seen it 
on factory wired units too. The buttons don't stay pushed, or won't release when another 
button is pushed, or when the release button is pressed. This is normally caused by the 
button knob rubbing against the hole it goes through. There is not much slop allowed 
for misalignment issues. 

If the tester has sticky buttons that won't stay down or won't release when you 
push another button or the Reset button, caused by the button rubbing against the hole
it goes through, the first thought is to loosen the screws that hold the button switch 
assembly to it's mount plate in order to try to realign the assembly. That won't work. 
The screw mount holes for the switch assembly are precision and do not provide any slop 
to adjust the assembly that way. Nor is it a good idea to try to bend the mount plate 
to try to fix it. The fix is actually rather simple once you know the trick. 

First place the filament switch to "0", the line adjust to "AC Off", the grid control to 
"100" and the Plate control to "0". This is so you will know what position the controls
are in to make it easier to put the knobs back on. It also makes it easier to access the
knob set screws. Remove the knobs from the Filament, Line Adjust, Grid and plate controls.
(Use a small flat blade screw driver to loosen the set screw inside the flat end of the knob.)
Now loosen the mount nuts that hold all four controls to the front panel (don't remove them, 
just loosen them). 

You can now move the push button assembly mount plate around to center the knobs in the holes. 
Once the push button knobs move freely and no longer stick, tighted the mount nuts on the controls.
If the push buttons still stick, loosen the control mount nuts and try again, the assembly may 
have moved as you tightened the control nuts. If after you tighten the control mount nuts it 
still doesn't fix the problem, you may have a more serious issue of a bent panel possibly caused 
by the tester being dropped. Good luck on fixing that... 

When you replace the control knobs, be sure to put the knobs back correctly. The set screws in the 
Filament and Line Adjust knobs go against the flat area on the control shaft. 

Check that the knobs on the plate and grid controls are placed properly. 
To check the grid control knob, rotate the grid control fully clockwise (towards 
the "100" position). The knob pointer should be pointing to the "100" tick mark. 
If it is not, loosen the knob and move it so it is pointing at the "100" tick.
You loosen the knob by using a small flat blade screw driver to loosen the 
screw located at the opposite side of the knob from the pointer. After adjusting 
the knob, retighten the knob screw and rotate the knob fully towards the "100" 
position to verify that it is now pointing at the "100" as the end stop. 
Rotate the knob counter clockwise towards the "0" position. You should feel 
the switch start to make just before the "0" position. The knob will normally 
rotate slightly past the "0" position about 3 to 4 tick marks distance when the 
switch is activated. 

To check the plate control knob, rotate the knob counter clockwise to the 
"0" position. The knob pointer should be pointing to the "0" tick mark. 
If it is not, adjust the knob so that it is pointing correctly (see the note 
on adjusting the Grid control knob above). Rotate the knob clockwise towards 
the "100" tick mark. It should stop at approximately the "100" tick mark
(plus or minus a tick mark or two is ok). If it stops significantly before or 
after the last tick mark, then the plate control is not the proper control 
for the tester which will result in incorrect test results. 

Note: The plate control used on some of the 667 testers don't match the tick marks
on the front panel. It will rotate past the 100 position by 3 or 4 tick marks.
That can cause a slight error when testing tubes that are sensitive to the plate setting
when set at the higher plate control setting (above 90). Not much that can be done about it.
I suppose you could compromise and try to set the known to have the equal offset at both 
ends of the scale, but then you increase the error at the lower settings. So there is no
good solution, But the error is small, less than 5%. 

See the companion file http://www.fourwater.com/files/restr666.txt to check the calibration 
of the tester. 


Checking the switches:

After the two calibration controls, the second biggest failure issue on the testers is the 
push button switches. They are overly sensitive to contact problems and can cause all sorts 
of test problems including the filament not lighting up, weak test results or no test results 
at all. All of the tube pin wiring goes through the push button switches, so they can lead 
to test problems if the contacts are bad. 

Before we get started, a note about the switches. 
If the tester has been sitting in storage for a long time, especially in an area that is 
subject to damp conditions at times, such as a basement, the attic, or a storage shed, 
the switches will likely be oxidized and covered in grime and may give you trouble 
(resistive connections or no connection at all). The first step before you begin the tests 
is to clear the switches. If they are not too badly oxidized or covered with grime, 
you can clear them by moving the switch a whole bunch of times. I usually move the switch 
though all positions 50 times. For a tester that has been stored well, that will usually 
clear the switches. If the tester has been stored in damp conditions, you will likely need 
to get more drastic and spray them with a contact cleaner. 

I used to recommend Deoxit, but I have found that it tends to cause the switches to
go bad again rather quickly because it leaves a residue behind that collects grime. 
This is especially true of the push button switches which are overly sensitive to 
any contact problems. WD-40 contact cleaner (Not the lubricant!) or CRC contact cleaner 
should be a good alternative. I still use Deoxit to clean the plate and grid controls if 
they need it as regular contact cleaner is not as effective on the wirewound controls. 
For the plate and grid controls I just spray a bit of Deoxit on a Q-Tip and rub that on the 
control wires (don't spray the stuff on the control directly, and don't rub too hard or you 
might damage the control). See the section on checking the grid and plate controls below 
for details.

Before you start cleaning, I recommend removing the paper scroll assembly so that you 
don't get the stuff on the paper since it won't come off.
With the tester out of the case, Set all the switches to position "1". 
Then place the tester on it's face on a workspace with a large piece of cardboard 
under it to catch the dripping mess and to protect the meter from scuffs and scratches.  

Carefully put a little bit of contact cleaner spray on each contact surface. Be vary 
sparing with the spray. You don't want the stuff all over the place. It makes a mess that 
is difficult to cleanup. After you spray each switch, move the switch through all the 
positions 50 times to work the cleaner into the contacts. No matter how hard you try to 
be careful, you are going to end up with a dripping mess (thus the need for the cardboard 
and to remove the paper scroll). Clean it up as best you can, but don't try to dig into 
the tester to clean it out amoung the wires, you will likely cause more problems.
The stuff will dry out on its own in time.  

Now we wait... Leave the tester to sit for at least a couple hours (24 hours is better). 
That gives the contact cleaner time to dry out. You don't want to run the tester while 
there is still a mess of contact cleaner all over it as that will cause leakage currents 
and give you bad results. 

Note: Be vary careful not to get the cleaner on the Line Adj and leakage calibration pots
that will cause them to go bad and you will either have to clean them or replace them. 
The contact cleaner will even seep into the sealed pots. 

I leave the tester on its face while spraying the contacts and leave it there while 
waiting for it to dry out. 

After you have let it dry out, move all the switches again 50 times through all the 
positions. That should clear up any remaining oxidation and grime. 
Hopefully that will clear up any problems. If it doesn't, you may have to repeat 
the process. If the tester has not been stored properly, it can be difficult to get 
the switches properly cleaned. 

Checking the push button switches:

With the tester out of the case, make sure all the levers are set to position "6".
Press the "Reset" button to make sure all the push button switches are released. 
With the tester on it's face, locate the push button bank. On the top you will 
see a row of ears with no wires attach to them. On the side of the tester where the 
"Line" push button is located, connect one lead of an Ohm meter to the ear on that 
top row that is closest to the Line button. Now follow the Tester's meter "+" lead 
to where it attaches to the transistor test switch. Connect the other Ohm meter lead 
to the ear on the transistor test switch where the tester's "+" lead connects. 
(Usually this lead is red in color, but not all kit builder's follow the rules, 
so it could be a different color.)
  
Now look at the Ohm meter you should read zero Ohms. Actually, it will probably 
read anywhere form 1 ohm to 3 or 4 ohms as it has to pass through 21 switch contacts 
(28 contacts on the 667 tester).
If you see more than 3 ohms, you may want to consider using contact cleaner on
the switches to improve the connection.  
Now push the "1" button. You should see an open circuit. Press the release button and 
you should see the resistance return to what you saw before you pushed the "1" button.  
Repeat this for each number button and the "C" button. Push the Line button and you 
should see an open circuit. Release the button and you should see the original resistance
reading. Pull the merit switch and you should again see and open circuit and a return to
the orignal resistance when you release the Merit lever. 
If the resistance changes significantly (more than 1 Ohm), the contacts are dirty/oxidized 
and need to be cleaned. 

Now go back to where the Ohm meter lead was attached to the push button bank ear next to 
the "Line" push button. Below that ear on the end of the switch closest to the front panel
is a similar ear without a wire connected to it. Move the Ohm meter lead to that ear.
Connect the other lead of the Ohm meter to the center lead of the Plate control. 
As with the top of the push button bank, you should see zero Ohms (meaning in reality
no more than 3 or 4 ohms). As before, push each of the push button switches and release
them to check to see that you get an open circuit and a return to the original resistance. 
As before, a significant change in resistance indicates that the switch contacts are 
dirty/oxidized and need to be cleaned. 
Note: the "Line" and "Merit" switches are not involved in this particular check. 

Move the Ohm meter lead back to the Transistor test switch where the Tester's meter "+" 
lead is connected. Move the other Ohm meter lead to the Plate control ear closest to the 
"Merit" lever. You should se an open circuit. Pull the Merit lever and you should see 
less than 2 Ohms. This checks the plate supply voltage circuit through the Merit switch. 
If you see more than 2 Ohms, the Merit switch needs to be cleaned. 



Now we will check the continuity of the switches in "test mode". 
Place all the lever switches to position "6".
Press the "reset" button to release all the pushbuttons.
Move one of the Ohm meter leads to the center ear of the "Line" switch. 
There is a wire that connects between the push button bank and the "1" lever switch.
Connect the other Ohm meter lead to that wire. 

-- sorry, I'm still working on this part ---


--------------------------------------------------------------------------------
<--- to do - write a comprehensive test procedure on how to check all the switches --->
--------------------------------------------------------------------------------


Checking the wiring:

This test can be performed without removing the tester from the case. 
Before you get into checking the voltages, first you need to make sure the wiring 
is correct. This can sometimes be a problem with the testers that were built from a kit. 
For this test you will need an ohm meter. 

Note: The tester MUST be unpluged for this test. 

First set all the lever switches to position "6" (position "4" for the "V" lever). 
The position of the other switches and controls don't matter at this point. 
I recommend using a resistor lead attached to the meter probe with an aligator clip jumper 
(if you have them) to make this test easier to perform. 

Note: The transistor circuits will be checked during the separate section on testing that 
part of the tester below. (The transistor circuits are a separate test circuit.)

This test checks that all the socket pins are wired correctly. 
Connect one meter lead into pin 1 of the 9 pin miniature socket (use the 12 pin compactron 
socket on the Eico 667). Place the other meter lead into pin 1 of each of the other sockets.
You should see zero ohms for each socket check. 
Repeat this with pin 2 on the sockets. Repeat the test with each of the socket pins.
(Obviously there is no need to check socket pins for a particular socket beyond the number 
of pins that socket has.) 

If you do not see zero ohms on the socket pin, mark it down on a 
piece of paper because it indicates that socket pin may not be properly connected. 

Note: If the socket is oxidized or has a lot of grime in it, it may be a bit difficult to 
get a good reading. You may need to scrape the probe (or resistor lead that you poke into 
the socket pin) around a bit to get a good connection. 


The following test checks for shorts between socket pins. 

Set all the lever switches to position "6". Press the "Release" button to release
any latched push button. Place one of the ohm meter leads into pin 1 of the 9 pin miniature 
socket (use the 12 pin compactron socket on the Eico 667). Place the other ohm meter lead into 
pin 2 of the socket. You should see infinate ohms (open circuit). Move the meter probe from 
pin 2 to pin 3, pin 4 etc. You should see infinate ohms (open circuit) for each socket checked. 
Also check between pin 1 and the faceplace (ground). 
All should show infinate resistance. Now move the first meter lead to pin 2 on the 
socket and repeat the check against all the other socket pins on that socket and the 
faceplate (ground). 
Repeat this by moving the meter leads to each of the other socket pins. 
Note: You only need to do this to the 9 pin socket (the 12 pin compactron socket if you are 
checking an Eico 667), since all the tube sockets are connected together. 

If you see a short circuit between any of the pins or the faceplate, it indicates that there 
is a short between them. Unfortunately, there is no easy way to determine which socket has 
the short other than to remove the tester from the case and visually inspect each socket's 
wiring for a possible short. This can be caused by a solder blob, or wire tail, or the pin 
connections touching each other. Pay particular attention to the small sockets for the 
subminiature tubes as the pins are very close together which makes them difficult to 
solder the leads without causing a short circuit.  
It is also possible that incorrect wiring may be the cause of the short. Again a visual 
inspection is the only easy way to check this. 


--------------------------------------------------------------------------------
---> todo - write the procedure on checking the wiring to the lever switches <----
--------------------------------------------------------------------------------


Part checks:

Some of the parts in the Eico tube testers need to be checked to make sure they are still
within spec. This is especially true of the carbon composition resistors which have a tendancy
to increase in resistance with age. 

The carbon resistors in the tester are R2 (680 Ohm), R4 (1K Ohm), R8 and R19 (270K Ohm) and
R15 (27K Ohm 1 Watt). See the Transistor test section checks for checking the R2 and R4 
resistors. The R8 and R19 resistors can be checked in place without disconnecting them.
To test R15, move the "S" switch to position "1" then you can check it in place without
disconnecting it. 

Note: if you need to replace R15, I recommend that you use a 2 Watt resistor instead of
a 1 Watt as specified. The 1 Watt resistor can actually be damaged under a specific 
setup condition on the tester. 
To replicate this condition, set the "S" switch to position "6", set the "V" lever to 
position "3" and set one of the pin switches to position "4" and press it's associated 
push button, then pull the Merit lever. If you leave the Merit lever pulled
for a long time, it can potentially damage the resistor because that places 165VAC 
across the 24K resistor. (180V-15V=165V/24K=0.0075Amp*165V=1.2375W) 
Actually, without a tube in the tester to load down the voltages, they will be even higher,
as much as 200VAC plate and 18VAC grid (200V-18V=182V/24K=0.00758A*182V=1.38W).

That being said, I've never seen one of the 24K resistors get burned out. 
The switch setup is not one that is used in any of the test setups, so normally
it would never be encountered. In addition, the 1 Watt resistors can normally 
handle slight overloads like that for short duration (a few seconds) without problems,
and since this only happens when you pull the Merit lever, it is unlikely that 
the excess power would be applied to the resistor for enough time to cause damage. 
Still, you should always check the resistor to make sure it is ok. 

P.S. As a side note, the 667 schematics show pins 5 and 6 of the S switch as being
tied together. This is incorrect. Pin 6 of the "S" switch is only tied to the 24K resistor.
The dot on the schematic showing pin 6 tied to pin 5 is not correct, it is not connected.
However, when the "S" switch is placed in the "6" position, this results in the
pins 5 and 6 being connected together since pin 5 is connected to the wiper of the switch. 
This is important to note because pin 6 is _only_ connected to pin 5 when the switch 
is set to position 6, thereby avoiding the overload situation noted above under normal 
test setup conditions. 

The metal film resistors are usually not a problem, but they can be checked too. 
To check the meter shunt resistors (R10, R11, R12, R13, R14), place the "S" switch to 
posiiton "6". You can now check the resistors in place without disconnecting them. 
To check the Transistor test section metal film resistors (R3, R5, R6), set the Transistor 
test switch to the Tube position. You can now check the resistors in place without 
disconnecting them. 

There are two wirewound resistors in the tester. I've never seen them go bad, but it is 
always worth checking resistors to make sure they are ok. 
To check R17 (400 Ohm), set the Grid control to the "50" position. You can now measure the 
resistor in place without disconnecting it. To measure the ballast filament resistor 
R20 (3K Ohm), set the filament switch to "0". You can now check the resistor in place 
without disconnecting it. 

If any of the resistors are out of tolerance, or if they look like they are burnt, 
you should replace them. 

There are two electrolytic capacitors in the tester (C1 and C2). 
Usually it is best to simply replace electrolytic capacitors when doing a restoration 
as they have limited life spans. However, the Eico testers don't push the capacitors 
very hard so unless there was a problem that damaged the capacitor, they often are still
in good working condition. To test them properly, you will need to disconnect one end of 
the capacitor and use a capacitance tester to check it. Since many people don't have a 
capacitance meter to check the condition of the capacitor, it is usually easier to 
just replace it with a new one. If you would rather not replace the capacitors unless
you have to, follow the voltage checks for the calibration/leak test circuit and the 
transistor test circuit voltage checks. If they check ok, that is an indication that the 
capacitors are still functional.

There are two diodes in the tester (CR1 and CR2), these connect to the two electrolytic 
capacitors. See the calibration checks on how to test the CR1 rectifier diode and
the transistor section checks for the CR2 diode. As with the capacitor checks, if
the voltages are ok, then the diodes are still functional and don't need to be replaced.

See the companion file http://www.fourwater.com/files/restr666.txt to check the calibration 
of the tester. 

The tests for the other parts; switches, potentiometers and transformer are described 
in other sections of this document. 


Testing the filament voltages:

Set the Transistor test switch to "Tube". 
Press the "Reset" button so that all the latching push buttons are released.
Set all the levers to the "1" position. Set the Line Adj control to 120.
Set the Grid and Plate controls to the center postion and the Filament switch to 6.3V. 
Also, make sure that you do not have a tube plugged into one of the test
sockets (we don't want it loading down the voltage). 

Press the Line switch and adjust the Line Adj control so that the meter needle 
is on top of the Line Adj mark. 

If the meter isn't working you can still do this test, but the voltages measured 
may not be as accurate. In this case just inititally set the Line Adj control to 
120V (center position). 

Note: The following tests all assume that the tester has been calibrated and 
The line adjust control properly set. If the tester is not calibrated, 
you can still do these tests, but you should expect to see a larger varience in 
the specified voltages. 

Release the Line switch. 
Press the "Reset" button so that all the latching push buttons are released.
Set the filament switch to the 117 position. 
Set the 1. switch to the 1. position.
Set the 2. switch to the 2. position. 
Set all the other levers to the "1" position.
Set the meter to read 150 Volts AC or the next highest selection. 
Place the meter probes into one of the tube sockets at pins 1 and 2. 
You should read about 120 Volts AC +-5 Volts. 
 (Yes this is higher than 117V, The voltage will sag a bit when a tube is in place
 and drawing current from the transformer.)

Note: If the meter calibration is not working, and the 117 filament voltage is 
close to to desired 120 Volts, go ahead and adjust the Line Adj control to 
get the 120 Volts reading and leave it there for the all of the following tests. 

Note: The octal socket is usually the easiest socket to use for the meter probes
as its pins are close to the size of most meter probes. However in a well used tester, 
or one that has socket problems it can be difficult to get a good connection. 
An alternate test point selection is to connect the negative lead of the meter to 
the tester faceplace. Connect the positive meter lead to the tester's tube cap lead 
(such as with an aligator clip) and set the "C" switch to position "2" and press 
the "C" button instead of the "2" button. (Be very careful as the cap lead will have 
120 volts on it, so don't touch it or let it short to the chassis.)

Move the filament switch to each of the possible positions. You should read the 
indicated voltage +-5%. The voltages should normally be slightly high since there 
is no tube installed in the test socket. If the voltage is low, try moving the 
filament switch back and forth a few times to see if that clears up the problem. 
You should see the 120 Volts at the "Z" position. It is connected to the 120V 
position through a 3K ohm resistor. 

You will likely see a small voltage on the "0" position if you are using a meter 
with a high input impedance. This is caused by leakage current in the switch. 
The zero position is not connected to anything. To verify that it is zero volts 
(ie no connection), you can place a resistor (1K ohm to 10K ohm) across the meter 
leads to provide a load to drain off the leakage current in the switch. Depending 
on the meter, you may still see a small residual voltage being measured of around
0.1 to 0.3 Volts. This is just the meter being overly sensitive and not 
a normal concern. 

Note: In operation, the exact filament voltage will be variable depending 
on the current draw of tube being tested. Remember to always do the Line Adj 
with the tube being tested in place and it's heater on and warmed up for the 
best test result. 

Usually the tester passes the filament test without problems as it is very rare
for the filament transformer or switch to fail. The most common problem is 
the switch being mis-wired for instruments that were built from a kit. 
If the voltages are low, check that your line voltage is ok. If you don't see the 
filament voltage; 1. There is no power to the tester. 2. The fuse may be blown. 
3. It may be the tester was improperly wired or the filament switch is bad, or 
4. The push button switch or pin selection switch is not making good contact. 
(See the sections for testing the push buttons and pin selector switches.)
5. You really do have a bad transformer (I've never seen that happen). 

Note: If you don't see any filament voltages and the fuse is ok (if the pilot
lamp is on, then there is power to the tester), you can check to see if the problem 
is a bad push button switch by removing the tester from it's case and connecting
the volt meter between the front panel and the ear for position #2 on the pin 
selection switches (they are all tied together). If you see the selected filament 
voltage, then it is likely that either the pin selector switch is bad, or the 
push button switch is bad. This is usually caused by grime build up on the push 
button switch and can normally be cleared with contact cleaner. 
See the sections on testing the push button and selector switches for more information.

Note: the push buttons line up over the lever switches on the 666 tester. 
On the early 667 testers, they do not line up above the levers, so double check which 
button and lever you are selecting if you are using one of the early 667 testers. 

Testing the plate, screen and grid supply voltages:

Press the "Reset" button so that all the latching push buttons are released.
Set the filament switch to the "0" position. 
Set the "1" switch to the "1" position.
Set the "2" switch to the "4" position. 
Press the "2" button (in the buttons above the "2" switch). 
Set the "V" lever to the "1" position. 
Set all the other levers to the "1" position.
Set the grid control to the "100" position (maximum clockwise position).
Set the meter to read 250 Volts AC. 
Place the meter probes into one of the tube sockets at pins 1 and 2. 

Note: If you are using the alternate test point of the tube cap lead
(see filament checks above) set the "C" switch to the "4" position 
(instead of using the "2" switch) and press the "C" button instead of 
the "2" button. 

Note: the position of the plate control doesn't matter for this test. 
It is used while measuring plate current, not the plate supply voltage,
so just leave it at the center postion. 
Also, make sure that you do not have a tube plugged into one of the test
sockets (we don't want it loading down the voltage). 

The voltages indicated below assumes a calibrated tester. 
If it is not calibrated, the voltages may be slightly different. 

With the "V" switch in position "1"
Pull the Merit lever. 
*You should read 54 Volts +-3 Volts (plate supply voltage). 
Move the "2" (or "C") switch lever to the "3" position.
*With the merit level pulled, you should read 18 Volts +-2 Volts (screen supply voltage). 
Move the "2" (or "C") switch lever to the "5" position.
*With the merit level pulled, you should read 6 Volts +-2 Volts (grid bias voltage). 

Move the "V" lever to postion "2". 
Move the "2" (or "C") switch lever to the "4" position.
Pull the Merit lever. 
*You should read 106 Volts +-5 Volts (plate supply voltage). 
Move the "2" (or "C") switch lever to the "3" position.
*With the merit level pulled, you should read 54 Volts +-3 Volts (screen supply voltage). 
Move the "2" (or "C") switch lever to the "5" position.
*With the merit level pulled, you should read 18 Volts +-2 Volts (grid bias voltage). 

Move the "V" lever to postion "3". 
Move the "2" (or "C") switch lever to the "4" position.
Pull the Merit lever. 
*You should read 215 Volts +-10 Volts (plate supply voltage). 
Move the "2" (or "C") switch lever to the "3" position.
*With the merit level pulled, you should read 106 Volts +-5 Volts (screen supply voltage). 
Move the "2" (or "C") switch lever to the "5" position.
*With the merit level pulled, you should read 18 Volts +-2 Volts (grid bias voltage). 
  (Yes, the grid bias for "V" position "3" is the same as for "V" position "2".)

Move the "V" lever to postion "4". 
Move the "2" (or "C") switch lever to the "4" position.
Pull the Merit lever. 
*You should read 215 Volts +-10 Volts (plate supply voltage). 
Move the "2" (or "C") switch lever to the "3" position.
*With the merit level pulled, you should read 106 Volts +-5 Volts (screen supply voltage). 
Move the "2" (or "C") switch lever to the "5" position.
*With the merit level pulled, you should read 54 Volts +-3 Volts (grid bias voltage). 
 (Yes, the plate and screen supply voltages are the same for "V" positions "3" and "4".
  The difference for those switch positions is the grid bias voltage.)

Note: If the grid voltage is bad, it is normally caused by a bad grid control which
is another common failure of the Eico tube testers (the calibration controls
going bad is the most common failure). The test step below will test the grid bias control. 
However, it can also be a bad push button switch or pin selection switch. 
See the associated section on testing the switches for more information. 


Interpreting the above plate, screen and grid voltage test results:
Normally these voltage checks are good with the exception of the grid control voltage
which can sometimes be bad (see below for additional testing of the grid voltage).
The most common problem with a bad grid voltage is the grid control is bad, although 
it too is subject to wiring problems for kit built testers. 
However, the voltage checks are all going through the push button switches and lever 
switches, so if there is a switch problem, that can also be the source of the bad reading. 
(See the section on testing the switches.)


Testing the grid bias control:

Press the "Reset" button so that all the latching push buttons are released.
Set the filament switch to the "0" position. 
Set the "1" switch to the "1" position.
Set the "2" switch to the "5" position. 
Press the "2" button (the button above the "2" switch). 
Set the "V" lever to the "4" position. 
Set all the other levers to the "1" position.
Set the grid control to the "100" position (maximum clockwise position).
Set the meter to read 250 Volts AC. 
Place the meter probes into one of the tube sockets at pins 1 and 2. 

Note: If you are using the alternate test point of the tube cap lead
(see filament checks above) set the "C" switch to the "5" position and press 
the "C" button instead of the "2" button. 

Note: the position of the plate control doesn't matter for this test. 
It is used while measuring plate current, not the grid bias voltage,
so just leave it at the center postion. 

With the grid control set at "100".
Pull the Merit lever. 
*You should read 54 Volts +-3 Volts (grid bias voltage). 
With the merit lever pulled, slowly rotate the grid control towards the "50" position. 
The voltage should smoothly reduce to 26 Volts with no drop outs or sudden voltage changes. 
Continue to move the grid control towards the "0" position. 
The voltage should smoothly reduce to 0 Volts with no drop outs or sudden voltage changes. 
If the voltage doesn't change, you may have a bad grid control (burned out), although it 
could be a bad push button or level switch, or bad wiring.

If you see drop outs, that indicates that the control wires are corroded. If you can get 
the back off the control, you can usually clean it with contact cleaner, or dab it with 
a little bit of Deoxit then rotate the control back and forth to work it in. 
If the there is no voltage, then likely the control is open. Time for a new control. 
However it is possible that it just may be excessively corroded, so if you can get the 
back off the control, you might be able to try to clean it. Although before you do that,
visually check the control windings. They should be smooth and no deformaties. 
If they look warped or burnt, then the control is bad and will need to be replaced. 
(See below for replacement.)

Also note that this test assumes that the push button switch is in good order as is the
pin selector switch. If they are dirty or broken, you will get bad readings. 
To check if it is a switch problem, remove the tester from the case and connect one 
volt meter lead to the center ear of the grid control and the other volt meter lead
to the faceplate, then repeat the grid control test. If you get the same bad result, 
then the problem is with the grid control. If the test checks ok, then the problem 
is with the switches. (See the section on testing the switches.)

Note:
You can quickly test the grid control voltage by setting the grid control to 100, 
lever switch 1 to position 1, and lever switch C to position 5. Set all the others to 1.
Set the V and S levers to position 1. Make sure the Transitor switch is set to Tube. 
The plate and filament control positions don't matter for this test.  
Press the Line Adj button and make sure the meter is centered on the Line adjust mark. 
Press the "C" push button to latch it. Place one AC meter lead on the front panel metal
(if you have clip leads on your meter, you can clip the ground lead to the "C" lever).
connect the other meter lead to the cap lead. Pull the merit lever, you should see 
around 6 Volts AC (+-0.5V). If you see something lower (such as 2 volts), the control 
may be bad. Slowly move the Grid control towards zero. It should smoothly change the 
voltage to zero. If you don't see this, there is something wrong. Do the detailed check
noted above to get a better idea of where the problem might be. If you installed the 
grid control protection lamp modification, the lamp is probably burned out. 
(See http://www.fourwater.com/files/restr666.txt)


Grid control Rectifier switch test: 

Once you get at or near the "0" position on the grid control, the rectifier test switch 
will activate. We use a resistance check to see if the switch is working correctly. 
This test requires the tester to be removed from the case. 

Unplug the tester. Using an Ohm meter, connect the meter probes across the 400 Ohm 
resistor located on the grid control switch. Move the grid control past the zero position
to activate the switch (ie fully counter-clockwise). You should see 400 Ohms 
resistance +- 40 Ohms). If you see more than 400 ohms, the resistor may be bad. 
If you see less than 400 Ohms, the switch may be shorted internally or the circuit 
is incorrectly wired. 

Now move the grid control to the "10" position. You should see zero Ohms (+-0.25 ohm). 
If you still see 400 Ohms or anything higher than a fraction of an Ohm, the switch is bad.
Time for a new control. (If you see an Ohm or two, you can sometimes clear it by operating 
the switch a bunch of time to clear off the grime/oxidation on the switch.)

Note: The calibration of the Grid control knob is to set the control fully clockwise 
("100" position). The knob pointer should be pointing at the "100" tick. 
If not, loosen the knob and adjust it to point at "100".  Now move the knob back 
to "0". As you approach "0", you should feel the switch activate. A correct pot will 
engage the switch at or just before "0" and the knob will move slightly past the "0" 
position (typically four or five tick marks worth).  


Replacing the Grid bias control:

There is a CTS control that appears to be a possible replacement for the grid 
control (5K 5W wirewound pot with switch) CTS part number 026TB32R502B1B1. 
The part is available from Digikey or Mouser. However the pot is 20% rather than 
the specified 5% tolerance. It also has a shaft that is 1/4 inch too long. 

I've used the part in a restoration and it appears to work ok, although you may 
want to cut the shaft down a bit so that the knob is closer to the front panel. 
The shaft is 5/8 inch long, but the Eico is expecting a 3/8 inch long shaft. 
Since the shaft is made of aluminium, it is easy to cut with a hacksaw.  

Use a Sharpie to mark the shaft 3/8 inch from the bushing (1/4 inch from the tip 
of the shaft). Place the tip of the shaft in a vise with the marked portion exposed. 
Use the hacksaw to cut off the tip of the shaft. Be sure to hold onto the control 
as you finish the cut so it doesn't fall on the floor. You may want to use a file to 
file down the sharp edges on the cut end of the shaft when you are done. 

CTS does have a specification for a grid control pot that would be a proper replacement
(CTS part number 026TB20R502D1B1), however I have not found anyone that sells it so it 
would require a custom order from CTS which is rather expensive (you could buy a whole 
new working Eico tester for less). 

Note: How bad is it to use a 20% tolerance part verses a 5% tolerance part? 
While it is preferable to have the correct tolerance, the 20% tolerance is not as bad 
as it seems. First consider that the pot is used as a voltage divider to reduce the 
transformer voltage to the selected grid bias voltage. In an unloaded voltage divider, 
the resistance of the pot would actually not matter as long as the control is linear. 
However since the control is connected to the grid of the tube under test, the load 
will be the grid current of the tube. This is variable depending on the tube and the 
voltages used to test the tube. This causes the voltage on the grid to be reduced 
slightly as the tube will load down the voltage divider (equivalent to adding a 
resistor from the pot wiper arm to ground). The position of the grid control sets 
the grid current and thus the added load to the voltage divider for which the grid 
control is used. 

So what does this mean? It means that the actual tolerance of the grid voltage is much
less than the tolerance of the grid control pot resistance. How much less is a difficult 
thing to measure and highly dependant upon the grid current of the tube being tested.
However, the grid voltage tolerance is definately less than the tolerance of the 
control itself, and it is the grid voltage that is the important part of the circuit.
In addition, the 20% is the worse case tolerance. The pots that I ordered showed a 
tolerance of 8% which is not too bad for how they are being used.

Finally keep in mind that having a working tube tester is better than no tester even 
if the tolerance is off by a few percentage points. Given the typical accuracy of the 
tube test charts, it is likely that you will not notice much if any difference. 

See http://www.fourwater.com/files/restr666.txt on a modification to help protect the 
grid control from being burned out. 


Testing the plate control:

With the tester unplugged, press the reset button so that all the latching buttons are 
released. Set all the levers to position "6". Move the plate control fully 
counter-clockwise (position "0"). The plate control knob pointer should be pointing 
at the "0" tick mark. If not, loosen the knob and adjust it to point at "0". 

With the plate control set at the "0" position, attach an Ohm meter to the two wires 
that attach to the plate potentiometer. The meter should read 3000 Ohms. 
If it doesn't, the pot is bad. Often this is just corrosion or dirt on the control, 
so see below on how to try to fix it.

Set the plate control to position "0". 
You should see 3000 Ohms on the meter +-150 Ohms. 
Now slowly move the control towards the "100" position. 
The resistance should slowly decrease with no sudden changes or drop outs. 

If the there is no resistance (open), then it is probable that the control is bad. 
However it is possible that it just may be dirty or corroded (which is the most common 
problem with the plate control), so if you can get the back off of the control, you 
might be able to try to clean it (see below). Although before you do that, 
visually check the control windings. They should be smooth with no deformaties. 
If they look warped or burnt, then the control is bad and will need to be replaced. 

Cleaning the plate control:

If you see drop outs, that indicates that the control wires are dirty or corroded or the 
control is burned open. If you can get the back off the control, you can usually clean it 
up with contact cleaner, or dab a Q-tip with a little bit of Deoxit and rub it on the control 
resistance wires along the top (where the rotor contacts the wires), then rotate the 
control back and forth to work it in. 

Warning: Don't scrub the control very hard or you will damage the wires. Just lightly 
rub the Qtip over the wires. 

If you use Doxit, leave everything sit for a couple of hours after you spray it and work 
the control to left the cleaner work on the control. Wipe the control with the Qtip again 
to clean off the oxidation that the cleaner disolved. Then rotate the control back and forth 
again from end to end to make sure it is worked in. 

If the control has a lot of corrosion, Deoxit is better than regular contact cleaner, but 
it will leave behind a residue which can collect dirt and grime creating future problems. 
A follow up with contact cleaner can help remove the excess Deoxit from the control. 

If the control is heavily corroded or dirty, you may need to repeat the cleaning 
procedure two or three times. 

Note: I recommend removing the paper roll assembly before you do this so as to not 
accidently spray the paper with the cleaning solution which will permanently stain 
the paper. 

If you suspect the rotor itself being corroded, dab a bit of Deoxit on a piece of 
paper and insert it between the rotor contact and the resistance wires. Do this 
by laying the paper on the wires and slowly rotate the control until the rotor moves 
over the paper, then move the control back and forth to work in the cleaner. 
Do NOT try to pry up the rotor you _will_ damage it if you try that. 

Cleaning the rotor center contact:

Don't forget to spray the rotor contact in the middle of the control as well. If the 
resistor coil contact has problems, then the rotor will also likely have problems. 
Don't try to lift the contact, just spray it or use the paper trick and work the knob 
back and forth to get the cleaning solution into the contact area. You can slip a small 
piece of paper under the contact and move it around a bit to help remove any debris. 

Note: This trick is for pots that have a dimpled contact riding on the center of the 
rotor assembly. 

If the problem persists, try using a relay contact burnisher to slip under the rotor
contact and swip it a couple of times. Do not scrub it, and don't swip it more than a
few times. Don't do this if you are not having problems with the control as there is 
a risk of damaging the contact. This is a last resort trick that you need to be very 
careful with. The purpose is to get rid of the dirt stuck under the contact without 
damaging the contact metal. 

Unfortuntely I have yet to find a control that is an acceptable replacement (3K 4W 5%) 
for the plate control. You cannot substitute a different resistance control as that will throw 
off the calibration and you will get bad readings. The control *must* be 3000 Ohms. 
(The 667 manual lists this part as having 10% tolerance, whereas the 666 manual says 5%.)

Note: I have heard of one fellow who replaced the control with a 24 position switch 
with resistors wired to the switch to emulate the plate rheostat. A bit drastic but 
if you don't have any other option it is a solution that could work. 
 
I have seen some 3.3K 5W 5% pots offered on eBay from China. You could possibly use this pot
as a replacement. You can improve the accuracy by connecting a 27K resistor between the two 
wires that connect to the plate control. That makes it slightly less linear, but improves
the resistance error to less than 10%, which would meet (barely) the spec for the 667 part. 

As with the Grid control, having a working tube tester is better than no tester even 
if the tolerance is off by a few percentage points. Given the typical accurcy of the 
tube test charts, it is likely that you will not notice much if any difference. 

The plate control normally doesn't get burned out as the meter will likely be destroyed
before the control will receive an overload long enough to damage it. 
The exception being a bad wiring job on the control, the wire attached to the control 
being pinched by the case when putting the tester back in it's case (so be careful!), 
or a fumble fingered person accidently shorting out the control.

Unfortunately installing the diode protection on the meter increases the possibility of
failures causing damage to the plate control, but the diode protection of the irreplacable 
meter is a better safety value. The plate control can be a more easy and cheaper replacement
compared to trying to find a replacement meter given that the meters are custom and have
not been made for many years, or only a meter from a parted out tester is going to be 
available, and the vast majority of the testers are parted out because the meter failed. 

Note: The plate controls used on later 667 testers don't match the tick marks on the front 
panel. It will rotate past the 100 position by 3 or 4 tick marks. That can cause a slight 
error when testing tubes that are sensitive to the plate setting when set at the higher 
plate control setting (above 90). Not much that can be done about it.
I suppose you could compromise and try to set the knob to have the equal offset at both 
ends of the scale, but then you increase the error at the lower settings. So there is no
good solution, But the error is small, less than 5%. 
The control is different because they couldn't get the old control and had to use a 
different one. Unfortuantely it didn't match the markings on the front panel, and they 
didn't change the front panel markings to match the new control. 



Testing the Transitor tester portion of the Eico 666/667 Tester:

The transistor test portion of the 666/667 testers is like the tube test section, fairly 
simple. It supplies a bias voltage to the base/emitter of the transistor and measures the 
output at the collector/emitter junction. This is similar to the emission test of a tube. 
The difference in the current applied to the base/emitter Vs. the collector/emitter 
current is the gain of the transistor. For the 666/667 transitor test, there are two 
selections for PNP transistors and two selections for NPN transistors to test low and 
medium gain transistors. The test is more of a go/nogo test as it is not very accurate 
and they do not show actual HFE gain readings on the meter scale. If you want to do 
detailed testing of transistors, you should consider getting a proper transistor tester. 


Checking the transistor test power supply wiring:

The Eico 666/667 testers use a germanium diode as a rectifier, obtaining it's voltage from 
the 7.5V AC filament tap of the power transformer. The germanium diode is used as a half wave 
rectifier to provide power for the transistor tester section. I've seldom seen these go bad 
other than by fumble fingered technicians or kit builders destroying them accidently. 
If you do need to replace the germanium diode, it can be replaced with pretty much any other 
good quality germanium diode. However, there is no need to replace it (and I do not recommend it) 
unless there is a problem with the original part. 

Note: It is possible to destroy the germanium diode accidently while testing a transistor. 
If you set the test switch to the PNP1 or PNP2 position, and accidently short the "E" (emitter) 
pin of the transistor socket to the chassis, that will cause the diode's cathode to be 
shorted to ground, which will destroy the diode since that effectively puts it directly
across the 7.5V AC filament tranformer connection. 

You can check if the diode is bad by looking at the DC voltage at the output. With the 
Line Adjust set to the calibration point, the germanium diode should have +5 Volts DC +-1V 
at the cathode end (positive terminal of the 10uf capacitor) relative to the chasis ground. 
However keep in mind that a degraded or open 10uF filter capacitor will cause the voltage to 
show lower. That is a problem with the capacitor, not the diode. Solid state diodes seldom 
fail partially. They usually short out or open up when they fail. Electrolytic capacitors 
on the other hand usually open up when they fail. 

You can do this test without removing the tester from the case. Place the positive 
DC Voltmeter probe in the transistor "E" socket pin and the DC volt meter negative probe 
connected to the case. Set the Transistor test switch to PNP1 or PNP2. 
The meter should show +5 Volts DC +-1 Volt. Now switch the meter to measure AC Volts. 
You should see 2.5 Volts +-0.75 Volt AC on the meter. 
To do the test with the tester out of the case, connect the volt meter positive lead to the
positive terminal of the 10uF capacitor located on the transistor test switch and the 
meter's negative lead goes to the tester's face plate. You should see the DC and AC voltages 
noted above. 

Note: The germanium diode used in the transistor circuit can be replaced with a 
silicon diode (such as the 1N4148), however it may affect the transistor test results 
as the silicon diode will have a lower voltage drop causing the transitor test power to be 
slightly higher, however this is a minimal effect (power; germanium=4.8V vs. silicon=5.3V DC), 
so if you don't have a germanium diode, you can use the silicon diode instead. However, there 
is normally no need to replace it unless it has been damaged.  

A design note here, the 680 ohm resistor (R2 across the C2 capacitor causes the charge on 
the 10uF capacitor to bleed off faster (1tc=6.8mS) than the 60Hz charge rate (t=16.7mS)
and since the capacitor is only charged on 1/2 wave cycle, the charge time is restricted 
to the peak portion of 1/2 wave at 60Hz. What this means is that there is a 2.5V AC 
ripple riding on the 5 Volt DC power. I'm not sure why they designed it this way, 
unless it was to try to give the transistor an AC signal during the test. 

Note: If you replace the diode, don't use a power diode like the 1N4004, it has too much 
voltage drop. Use a 1N4148 or similar diode which is more appropriate as a substitute diode. 
However, a germanium diode like a 1N151 (an improved version of the 1N34) is a better replacement. 

What this means though is that it is important that the 10uF capacitor and the 
680 ohm resistor be in good condition to be able to get accurate test results for the 
transistor test since the capacitor is being used outside its normal operating condition.
And most importantly, do not replace the 10uF capacitor with a different capacitance.

Luckily from my experience the capacitor seems to survive this abuse. Unfortuantely, 
the choice of the carbon composition resistor has been a problem in that they seemed to
have selected a supplier that produced some resistors that fall out of spec over time, 
which can happen with some carbon composition resistors. So be sure to check the 
resistance of the resistor as a part of the restoration process.  

To check the resistor, place the ohm meter probes across the 680 ohm resistor. 
You should see 680 ohms +-34 ohms. If you see much less (such as 250 ohms), reverse the 
probes (when you see the low resistance, you are measuring the power supply diode not 
the resistor). If the resistor shows higher resistance than 714 ohms, you should replace 
the resistor. I would recommend replacing it with a modern metal film resistor. 

Note: if you see less than 10 ohms, that usually means the germanium diode is shorted
and needs to be replaced. Although it is possible the capacitor is the source of the 
short, but my experience is that the problem is with the diode. Still if you change
the diode you should replace the 10uF capacitor too as a shorted diode will cause damage 
to the capacitor since it will be exposed to direct AC power. 
It is also possible that the short is caused by a wiring problem. 
The connections on the transistor test switch are somewhat crowded, so wiring shorts 
are a possibility. 

Another aspect is that the capacitor should be a generic electrolytic capacitor. 
Do not use a tantalum or ceramic type capacitor, or even a low ESR electrolytic capacitor 
as that will put excess strain on the germanium diode which was never intended to be used 
the way they are using it. However, if you do decide to use a silicon diode to replace the
germanium diode, then you should consider using a capacitor that can handle high ripple 
(these are usually rated for use in power supply circuits). 

If you are never going to use the transistor test portion of the Eico 666/667 Tester, 
then you can ignore this section about the transistor tester. Other than the transitor 
switch setting up the proper meter connections, the transistor test circuits are entirely 
separate from the tube test portion of the tester and will have no effect on the tube 
tests even if the transitor test portion is not working. Conversely, other than the meter, 
proper line adjust and deriving it's power from the 7.5V filament tap, the transistor test 
portion is not affected by the tube test portion of the tester. 


Checking the Transistor Test Switch:

This test can be done without removing the tester from it's case. 

Set the Transistor switch to the Tube position. 
Press the Line push button switch and adjust the Line Adjust control to center the 
Eico's meter needle on the Line Adj mark. 

Note: If you cannot calibrate the meter (refer to the restr666.txt document), 
you can still do this test, but the voltages will not be as accurate. 

***WARNING***
BE VERY CAREFUL with this test. If you accidently short the "E" pin to the front panel 
or other parts of the tester, you can destroy the Transistor test power circuit. 

To do this test you may want to clip the DC Volt meter lead to a small resistor 
lead and poke that into the socket. I recommend using a 10K Ohm resistor with leads 
thin enough to fit in the transistor socket without damaging it. 
Connect the meter lead to one end of the resistor and poke the other resistor 
lead into the transistor socket. You can do the test by placing the meter lead directly 
in the transistor socket, but be very careful you don't accidently short it to anything 
else or you may damage the transistor test section power supply. 

Set the transistor switch to the PNP2 position. 
Connect a DC Volt meter positive lead to the "E" pin of the transistor socket.
Connect the Volt meter's negative lead to the tester's faceplate. 
You should see +5VDC +-1Volt. Move the Transistor test switch to the PNP1 
position. You should still see the +5VDC. Move the Transistor switch to the 
Tube position. The +5VDC should go away. It should remain at 0V when you move 
the test switch to the NPN1 and NPN2 positions. 

With the Transistor test switch in the NPN2 postion. Move the DC Volt meter 
positive lead from the "E" pin of the transistor socket to the "B" pin on the socket. 
You should see the +5VDC, although it may show as slightly less depending on the input 
resistance of the DC Volt meter (It is reading the +5VDC signal though a 
200K Ohm resistor). 

Now unplug the tester. The remaining Transistor switch tests require the power to be off.

The following tests use an Ohm meter. Preferably a digital VOM meter that can 
measure resistance.

Connect the negative lead of the Ohm meter to the tester's front panel. 
Connect the positive lead of the ohm meter to the "B" pin of the transistor socket.
For this test don't feed the Ohm meter through the 10K resistor. Instead, connect the
meter lead to the end of the resistor that plugs into the transistor socket, or use a 
small piece of thin wire. 

With the Ohm meter probe connected to the "B" pin, set the Transisor test switch to 
the PNP2 position. You should see 200K Ohms. 
Move the Transistor test switch to the PNP1 position. You should see infinite resistance. 
Move the switch to the Tube position and the NPN1 position. The Ohm meter should still
show infinite resistance (ie open circuit). Move the switch to the NPN2 position. 
You should see 200K Ohms again (actually slightly higher since the return is through 
the Transistor power supply circuit). 

Move the Ohm meter positive lead to the "E" pin of the Transistor test switch. 
With the switch set to the PNP2 position, you should see between 700 Ohms and 200 Ohms 
(the resistance of the Transistor power supply). Move the switch to the PNP1 position. 
You should continue to see the 700 Ohms to 200 Ohms resistance. 
Move the switch to the Tube position. You should see infinite resistance (open circuit).
Move the switch to the NPN1 position. You should see 114 Ohms +-2 Ohms. Move the switch 
to the NPN2 position. You should see 48 Ohms +-2 Ohms. 
Note: The resistance tolerance may be greater depending on the tolerance of your Ohm 
meter's resistance meaurement circuit. 
The resistance will be slightly less than the 129 and 50 ohm resistors in the transistor 
test circuit because the Eico meter will be in parallel with the resistors.

Move the Ohm meter positive lead to the "C" pin on the transistor socket. 
With the switch set to the NPN2 position, you should see between 1700 Ohms and 1200 Ohms 
(1000 Ohms plus the resistance of the Transistor power supply). Move the switch to the 
NPN1 position. You should continue to see the 1700 Ohms to 1200 Ohms resistance. 
Move the switch to the Tube position. You should see infinite resistance (open circuit).
Move the switch to the PNP1 position. You should see 114 Ohms +-2 Ohms. Move the switch 
to the PNP2 position. You should see 48 Ohms +-2 Ohms. 
Note: The resistance tolerance may be greater depending on the tolerance of your Ohm 
meter's resistance meaurement circuit. 
The resistance will be slightly less than the 129 and 50 ohm resistors in the transistor 
test circuit because the Eico meter will be in parallel with the resistors.

<eof>

--------------------------------------------------------------------------------
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!

<eof>