This article discusses repair and calibration of the EICO 950A Resistance-Capacitance-Comparator Bridge, commonly used as a capacitor tester. Also discussed are the PACO C-20, EICO 950, and EICO 950B tester. The exact EICO 950A unit pictured in the photos below is for sale at $149 plus shipping, rebuilt and calibrated as described below. The PACO C-20 unit is for sale at $139 plus shipping. Extreme caution must be followed with any vintage capacitor tester because very high voltages are present during repair and while operating. They should only be used by knowledgeable technicians.

When the bridge is unbalanced, the magic eye will remain closed.

PACO C-20 tester

Eico offered three models, the 950, 950A, and 950B. They are very similar, and most of this article will be relevant to all three models. Likewise, I have updated this article to include coverage of the PACO C-20 unit, which is almost identical to the EICO 950B.

These models use a balanced bridge that measures capacitance from 10mmf to 5000mf (some models 2000mf) and resistance from 0.5 ohms to 500 Megohms (some models up to 200 Megohms). During component value testing, the magic eye tube serves as the null-indicator. When the bridge is far from balanced, the target area of the magic eye tube glows completely closed green, and in fact overlaps. As the pointer dial approaches balance point, first the overlapping disappears and eventually the entire target area of the eye tube is completely open (dark). The bridge is balanced when the maximum dark area is indicated, and you can then read the value of your component on the faceplate scale. This is a very high quality test instrument when properly rebuilt.

Any technician who repairs vintage tube equipment (such as tube amps, antique radios, vintage jukeboxes) will find that a quality capacitor tester is a highly valued test instrument, and once you realize all the benefit it offers, will probably receive as much usage as your tube tester. This model can properly load test a capacitor to more than 500V, which is the ONLY way to correctly test high voltage capacitors for leakage. ( Leakage is the most common problem with capacitors.) Similarly, many NOS capacitors need to be “reformed” (a process of slowing increasing the voltage to a capacitor over a period of time to allow the chemical formula inside the capacitor to reformulate). A premium-quality capacitor tester, such as any member of the 950-series, will allow you to reformulate capacitors.

Simply connect the cap that you want to reform to the right-hand terminals, set the range to either (’Paper/Mica’ or ‘Test Voltage’, name varies by model) for regular caps, or to ‘Electrolytic Test’ for electrolytics, and dial up the Voltage selector. Start low and gradually increase the voltage over a period of hours, with the goal of achieving the capacitor’s maximum rated voltage with no leakage detected. Stubborn caps may require 3 extended attempts (1 attempt per day). If the cap still shows leakage after 3 attempts, it is defective. Extreme caution must be exercised when removing electrolytics from testing. Be sure to slowly turn the voltage down to zero before removing the capacitor undergoing testing. Even then, you should carefully discharge the capacitor before touching any of the leads, just in case any charge was still stored.

Eico 950A leakage testing of regular capacitor. Open eye indicates no leakage.

Eico 950A electrolytic leakage. Lighting of Neon bulb for extended period indicates excess Leakage.



For the 950A, leakage for Paper/Mica caps is indicated by the #1629 magic eye tube. Leakage for electrolytics is indicated by the Neon bulb.

For the more common 950B model, leakage for either type of capacitor is indicated by the magic eye tube. Personally, I prefer the rare 950A for this one small reason — since the 950A does not use the magic eye tube to indicate leakage, you can install a small toggle switch in series with the Plate of the 1629 tube (pin 4 at 1629 socket) and put the 1629 tube into ‘Standby’ mode by switching off the Plate supply voltage. This is very practical because capacitor reforming will take many hours, and there is no reason to leave the expensive magic eye tube running during the reforming process. With the 950A model, you will still be able to monitor your progress because the Neon bulb is your indicator. If you make this modification with the 950B, since the magic eye tube itself is your indicator of leakage, you will not be able to monitor whether the cap is still showing leakage at a certain voltage level. This is a small benefit with the 950A model in my opinion. Of course, you have to add this small modification yourself, and does require drilling a hole in the faceplate and mounting the toggle switch. For that reason, I do not make this modification when I resell a rebuilt unit, but I do use the modification myself.

Testing a 50-Megohm resistor

Another practical use for this unit is to test for extremely high resistance. Most regular multimeters have a limited upper range. Cheap digital units may not go above 5 Meg ohms. Standard Fluke’s test up to 20-Megohm, and expensive Fluke DMM’ such as model 179 only test up to 50-Megohms. Well, this antique tube-based bridge will analyze up to 500-Megohm! Accuracy above 50-Megohms is a loose tolerance, but up to 50-Meg the unit is quite accurate. See example photo testing a precision 50-Megohm resistor. The eye tube indicator is balanced very close to the 50-Meg mark on the dial. In a similar test, I connected three of these precision 50-Megohm resistors in series (150-Megohms resistance), and the unit tests roughly between the 100 and 500 megohm indicators.

Testing 150-megohm resistance

Considering that my Fluke 179 would have only tested 1/3 of this resistance, and could not do anything with this test example, I would say that this degree of accuracy is very acceptable for these extremely high resistances.
Repair and Calibration

Check/Replace line cord. The unit uses two tubes internally, a 6X5GT rectifier and a 1629 magic eye tube. Test both tubes and replace as needed. For model 950A, remove the Neon bulb, clean bulb connections and treat bulb socket with deoxit. Clean all switches and potentiometers with Deoxit. As always, every resistor must be checked for accuracy and replaced where necessary. Replace all paper capacitors, making sure that the 2mfd capacitor is accurate. (Use a modern digital capacitor tester). It is critically important that the electrolytics, an 8mfd at 525V and a 4mfd at 250V, are replaced with precision accuracy. For example, using a 10mfd in place of the 8mfd is unacceptable in this circuit. For this reason, you may want to remove these two capacitors, test them with another cap tester, and reinstall them if they are good. To replace the 8mfd cap, your best option will probably be a series-connection of (1) 10mfd @450V + (1) 40mfd at 450V, and again, you MUST test these electrolytics prior to installation to verify their value. Most electrolytic capacitors have very loose tolerances, often as much as 50% variation in value, so do NOT assume their value is accurate.

Once resistors and capacitors have been tested and replaced, test for the voltages as indicated on the schematic. If your voltages differ, stop and troubleshoot. Next, connect your Fluke meter to the right-side terminals, set the function switch for Electrolytic Test, and turn the Voltage control. Your Fluke should read whatever voltage you are dialing. If the voltage is off slightly, reposition the knob to achieve best accuracy over the entire range of the dial.

Finally, take a resistor of known value of 200K or 300K (.2Meg or .3Meg), set the Range to ‘5Kohm — 5Mohm’, and turn the pointer until the bridge is balanced. Once balanced, loosen the pointer setscrew, and reposition pointer onto the shaft at correct faceplate position. In the example below, I am using an NOS carbon resistor which (due to age) has increased in value to almost exactly 300K. Hence, in this example we want the pointer to indicate 0.3Meg on the faceplate scale.

Accurate testing of 300K resistance

Fluke verification of resistor value.

Examples of Operation

Defective capacitor with leakage

Example #1: I am testing a clean NOS American-made Mallory 40mfd @450V capacitor. Even after an extended period of time, the capacitor still shows leakage at voltages as low as 250V. (All electrolytics will show leakage at first, but will subside if they are not defective). After attempting reformulation, this capacitor is defective — and would not have been found with modern digital capacitor testers, nor with lower voltage cap testers.

Eico 950A testing capacitor value

Testing the actual capacitance value of this capacitor, you see a value of approx 51mfd. As previously stated, capacitor tolerances are well-known to vary widely, often as much as 50% difference. In this example, approx 25% high. Again, it is important to note that modern capacitor testers would have told you this capacitance value, and told you no shorts, thereby giving you a misleading test result of ‘good’. In fact, the capacitor has excess leakage starting at 250V, and is defective for its rated working voltage. You can now see the value of the vintage high-voltage capacitor tester.

No leakage in this NOS Mallory 450V capacitor!

Example #2: you see no leakage in this NOS American-made Mallory 100mfd at 450V electrolytic capacitor, testing at full 450V voltage.

It took less than 10 minutes for this capacitor to show no leakage at full voltage. You can feel confident that this capacitor will work excellent in circuit.


The PACO C-20 tester is functionally almost identical to the EICO 950B, and any technician should be easily able to use either schematic to repair either unit. The very minor difference is that the PACO has a maximum rated capacitance measurement of 2000mf and maximum rated resistance measurement of 200 Megohms. This is a meaningless difference. All other functions, including DC Polarizing Voltage of 500V+ for capacitor leakage testing, is same.

PACO C-20 tester

This unit is for sale, also repaired and calibrated. Like the EICO 950B, this model uses the 1629 Magic Eye tube for all functionality — both measurements AND leakage testing. Otherwise, there is no significant difference between the C-20 or EICO 950, 950A, 950B.

The C-20 works great, the faceplate is in very good condition, and I have also repainted the cabinet very close to original (Hammertone Gray).


Here are photos of the Paco C-20 in action, testing the leaky Mallory 40mfd cap, and then the good Mallory 100mfd cap.

Paco C-20 testing defective capacitor with leakage. Notice Magic Eye tube is almost closed, even at less than 250V

Paco C-20 testing good capacitor without leakage. Notice open Magic Eye.



regards,

Bob Putnak.

eBay ID = rjputnak

Additional Photos (click each to enlarge)