The need for such a device arises every time when repairing a welding inverter– you need to check a powerful IGBT or MOSFET transistor for serviceability, or select a pair for a working transistor, or when purchasing new transistors, make sure that it is not a “remarker”. This topic has been repeatedly raised on many forums, but having not found a ready-made (tested) or someone designed device, I decided to make it myself.
The idea is that you need to have some kind of database various types transistors, with which to compare the characteristics of the transistor under test, and if the characteristics fit within a certain framework, then it can be considered serviceable. All this should be done using some simplified method and simple equipment. Of course, you will have to collect the necessary database yourself, but this can all be solved.

The device allows:
- determine the serviceability (failure) of the transistor
- determine the gate voltage required to fully open the transistor
- determine the relative voltage drop across K-E conclusions open transistor
- determine the relative gate capacitance of the transistor, even in one batch of transistors there is a scatter and it can be seen indirectly
- select several transistors with the same parameters

Scheme

The schematic diagram of the device is shown in the figure.

It consists of a 16V power supply direct current, digital millivoltmeter 0-1V, voltage stabilizer +5V on LM7805 to power this millivoltmeter and power the “light clock” - flashing LED LD1, current stabilizer on the lamp - to power the transistor under test, current stabilizer to - to create adjustable voltage(at a stable current) on the gate of the transistor under test using a variable resistor, and two buttons to open and close the transistor.

The device is very simple in design and is assembled from publicly available parts. I had some kind of transformer with an overall power of about 40 W and a voltage on the secondary winding of 12 V. If desired, and if necessary, the device can be powered from a 12V / 0.6 Ah battery (for example). It was also in stock.

I decided to use power from a 220V network, because you can’t go to the market for shopping with the device, and the network is still more stable than a “dead” battery. But... it's a matter of taste.
Further, while studying and adapting the voltmeter, I discovered an interesting feature: if a voltage exceeding its upper measurement threshold (1V) is applied to its terminals L0 and HI, then the display simply goes out and it does not show anything, but if you reduce the voltage and everything returns to normal indication (this is all with constant nutrition+5V between terminals 0V and 5V). I decided to use this feature. I think that many digital “display meters” have the same feature. Take, for example, any Chinese digital tester, if in 20V mode you apply 200V to it, then nothing bad will happen, it will just display “1” and that’s it. Scoreboards similar to mine are now on sale.
Possible.

About the operation of the circuit

Next, I’ll tell you about four interesting points about the scheme and its operation:
1. The use of an incandescent lamp in the collector circuit of the transistor under test is due to the desire (initially there was such a desire) to visually see that the transistor has OPENED. In addition, the lamp performs 2 more functions here: protecting the circuit when connecting a “broken” transistor and some stabilization of the current (54-58 mA) flowing through the transistor when the network changes from 200 to 240V. But the “feature” of my voltmeter allowed me to ignore the first function, while even gaining in measurement accuracy, but more on that later...
2. The use of a current stabilizer made it possible NOT to accidentally burn out a variable resistor (when it is in the top position according to the circuit) and accidentally press two buttons at the same time, or when testing a “broken” transistor. The amount of limited current in this circuit even with short circuit equal to 12 mA.
3. Using 4 pieces of IN4148 diodes in the gate circuit of the transistor under test to slowly discharge the gate capacitance of the transistor when the voltage at its gate has already been removed and the transistor is still in the open state. They have some insignificant leakage current, which discharges the capacitance.
4. Use of a “blinking” LED as a time meter (light clock) when the gate capacitance is discharged.
From all of the above, it becomes absolutely clear how everything works, but more on this a little later...

Housing and layout

Next, a case was purchased and all these components are located inside.

Outwardly, it turned out not even bad, except for the fact that I still don’t know how to draw scales and inscriptions on a computer, but... The remains of some connectors worked great as sockets for the transistors under test. At the same time, an external cable was made for transistors with “clumsy” legs that would not fit into the connector.

Well, this is what it looks like in action:

How to use the device

1. We turn on the device into the network, the LED starts blinking, the “display meter” does not light up
2. Connect the transistor under test (as in the photo above)
3. Set the voltage regulator knob on the gate to the extreme left position (counterclockwise)
4. Press the “Open” button and at the same time slowly increase the voltage regulator clockwise until the “display meter” lights up
5. Stop, release the “Open” button, take readings from the regulator and record. This is the opening tension.
6. Turn the regulator all the way clockwise
7. Press the “Open” button, the “display meter” will light up, take readings from it and record it. This is the K-E voltage on an open transistor
8. It is possible that during the time spent on recording, the transistor has already closed, then we open it again with the button, and after that we release the “Open” button and press the “Close” button - the transistor should close and the “display meter” should go out accordingly. This is a check of the integrity of the transistor - it opens and closes
9. Again, open the transistor with the “Open” button (voltage regulator at maximum) and, having waited for the previously recorded readings, release the “Open” button while simultaneously starting to count the number of flashes (blinks) of the LED
10. After waiting for the “display meter” to go out, we record the number of LED flashes. This is the relative time of discharge of the transistor's gate capacitance or closing time (until the voltage drop across the closing transistor increases by more than 1V). The greater this time (quantity), the correspondingly greater the gate capacity.

Next, we check all the available transistors, and put all the data in a table.
It is from this table that it comes comparative analysis transistors - whether they are branded or “remarkers”, whether they correspond to their characteristics or not.

Below is the table that I came up with. Transistors that were not available are highlighted in yellow, but I definitely used them once, so I left them for the future. Of course, it doesn’t represent all the transistors that passed through my hands; I simply didn’t write down some of them, although I seem to always write. Of course, when repeating this device, someone may end up with a table with slightly different numbers, this is possible, because the numbers depend on many things: on the existing light bulb or transformer or battery, for example.

The table shows the difference between transistors, for example G30N60A4 from GP4068D. They differ in closing time. Both transistors are used in the same device - Telvin, Technique 164, only the first ones were used a little earlier (3, 4 years ago), and the second ones are used now. And the rest of the characteristics according to DATASHIT are approximately the same. And in this situation, everything is clearly visible - everything is there.

In addition, if you have a table of only 3-4 or 5 types of transistors, and the rest are simply not available, then you can probably calculate the coefficient of “consistency” of your numbers with my table and, using it, continue your table using numbers from my table. I think that the dependence of “consistency” in this situation will be linear. For the first time, it will probably be enough, and then you will adjust your table over time.
I spent about 3 days on this device, one of which was buying some small things, a housing, and another one for setting up and debugging. The rest is work.

Of course, the device has possible design options: for example, using a cheaper pointer millivoltmeter (you need to think about limiting the pointer's travel to the right when the transistor is closed), using another stabilizer instead of a light bulb, using a battery, installing an additional switch to test transistors with a p-channel, etc. .d. But the principle in the device will not change.

I repeat once again, the device does not measure the values ​​(digits) indicated in the DATASHEETS, it does almost the same thing, but in relative units, comparing one sample to another. The device does not measure characteristics in dynamic mode, it is only static, like a regular tester. But not all transistors can be checked with a tester, and not all parameters can be seen. On these I usually put a question mark “?”

You can also test it in dynamics, put a small PWM on the K176 series, or something similar.
But the device is generally simple and inexpensive, and most importantly, it ties all subjects to the same framework.

Sergey (s237)

Ukraine, Kyiv

My name is Sergey, I live in Kyiv, age 46 years. I have my own car, my own soldering iron, and even my own workplace in the kitchen, where I sculpt something interesting.

I love high-quality music on high-quality equipment. I have an ancient Technix, everything sounds on it. Married, has adult children.

Former military. I work as a master repairing and adjusting welding equipment, including inverter equipment, voltage stabilizers and much more, where electronics are present.

I don’t have any special achievements, except that I try to be methodical, consistent and, if possible, finish what I start. I came to you not only to take, but also, if possible, to give, to discuss, to talk. That's all briefly.

Reader vote

The article was approved by 75 readers.

To participate in the voting, register and log in to the site with your username and password.

There is probably no radio amateur who does not profess the cult of radio engineering laboratory equipment. First of all, these are attachments for them and probes, which for the most part are made independently. And since there are never too many measuring instruments and this is an axiom, I somehow assembled a transistor and diode tester that was small in size and had a very simple circuit. It’s been a long time since I’ve had a multimeter that’s not bad, but homemade tester, in many cases, I continue to use it as before.

Device diagram

The probe designer consists of only 7 electronic components + printed circuit board. It assembles quickly and starts working absolutely without any setup.

The circuit is assembled on a chip K155LN1 containing six inverters. When correct connection one of the LEDs (HL1 when N-P-N structure and HL2 at P-N-P). If faulty:

1. broken, both LEDs flash
2. has an internal break, both do not ignite

The diodes being tested are connected to terminals “K” and “E”. Depending on the polarity of the connection, HL1 or HL2 will light up.

There are not many circuit components, but it’s better to make them printed circuit board, it is troublesome to solder wires to the legs of the microcircuit directly.

And try not to forget to put a socket under the chip.

You can use the probe without installing it in the case, but if you spend a little more time on its manufacture, you will have a full-fledged, mobile probe that you can already take with you (for example, to the radio market). The case in the photo is made from the plastic case of a square battery, which has already served its purpose. All that was needed was to remove the previous contents and saw off the excess, drill holes for the LEDs and glue a strip with connectors for connecting the transistors being tested. It would be a good idea to “dress” the connectors with identification colors. A power button is required. The power supply is a AAA battery compartment screwed to the case with several screws.

The fastening screws are small in size, it is convenient to pass them through the positive contacts and tighten them with the obligatory use of nuts.

The tester is in full readiness. It would be optimal to use AAA batteries; four 1.2 volt batteries will give the best option supply voltage is 4.8 volts.

Good day everyone, I would like to present a probe for transistors that will definitely show whether it is working or not, because it is more reliable than simply testing its terminals with an ohmmeter like diodes. The diagram itself is shown below.

Probe circuit

As we can see, this is an ordinary blocking generator. It starts up easily - there are very few parts and it is difficult to mix anything up during assembly. What we need to build the circuit:

1. Bread board
2. LED of any color
3. Momentary button
4. 1K resistor
5. Ferrite ring
6. Varnished wire
7. Socket for microcircuits

Parts for assembly

Let's think about what we can pick up from where. You can make such a breadboard yourself or buy it; the easiest way is to assemble it with a canopy or on cardboard. The LED can be picked out from a lighter or from a Chinese toy. The non-latching button can be picked from the same Chinese toy, or from any burnt household device with similar controls.

The resistor does not have to have a nominal value of 1K - it can deviate from the specified nominal value within 100R to 10K. The ferrite ring can be obtained from energy saving lamp, and not necessarily a ring - you can also use ferrite transformers and ferrite rods, the number of turns is from 10 to 50 turns.

The wire is varnished, it is permissible to take almost any diameter from 0.5 to 0.9 mm, the number of turns is the same. You will learn how to connect the windings for proper operation during testing - if it doesn’t work, then simply swap the ends of the terminals. That's all, now a short video of the work.

Video of the tester working

This simple device schematic diagram which you see in the figure, is designed to identify hidden defects and control the reverse uncontrolled current in bipolar and BSIT transistors of any structure, at an operating voltage of 30...600 V. They can also check the reverse current of SCRs, triacs, diodes and determine the operating voltage gas-discharge lamps, varistors, zener diodes.

It is known that checking with a conventional multimeter semiconductor devices with a maximum operating voltage of more than 50 V does not give a complete picture of the serviceability of the part, since the test takes place at too low a voltage, which does not allow us to clearly judge how this part will behave when operating at its rated, much higher, voltage.

Those who have ever had to repair TVs or monitors can probably remember cases when a completely new powerful high-voltage transistor installed in a horizontal scanning module or switching power supply failed in the very first seconds of operation.

It is not uncommon to see “strange” behavior of triacs and thyristors in phase power regulators, which manifests itself as flickering of incandescent lamps connected as a load. At the same time, the thyristor usually begins to noticeably heat up even when operating with a 40 W load.

Numerous probes for testing "low voltage" bipolar transistors are not suitable for testing high-power high-voltage transistors. For example, KT840A, according to the reference book, has a maximum voltage of 400 V, with a 100 Ohm resistor connected between its base and emitter terminals, the reverse collector current at a temperature of 25°C should not exceed 0.1..3mA.

It is clear that 3 mA is the worst value at which the transistor can be considered conditionally serviceable. Several of the tested transistors of this type behaved “decently” only up to E-K voltage= 200...250 V. With a further increase in voltage, the reverse current increased sharply, exceeding the permissible value according to the reference data. When trying to install in pulse block power supply MP3-3, two such transistors failed in the first seconds of operation, taking each of the KU112A SCRs with them “to the grave”.

A lot of defective parts are also found among diodes, which can also be read well by a multimeter, but in reality can only work at low voltage.

It should be borne in mind that if the transistor being tested has an initial uncontrolled current that is worse than that given in the reference book, or is obviously worse than that of other transistors of the same type, then you may have in front of you not just a slightly low-quality specimen, but a so-called “fracture” - when under the guise of one transistor, you purchase another, but “unpopular” one in the same package, from which the old markings have been washed off and a new one has been applied.

Transistors and electrolytic capacitors.

Probe for checking transistors, diodes - first option

This circuit is based on a symmetrical multivibrator, but the negative connections through capacitors C1 and C2 are removed from the emitters of transistors VT1 and VT4. At the moment when VT2 is closed, the positive potential through the open VT1 creates a weak resistance at the input and thus increases the load quality sampler.

From the emitter VT1, a positive signal goes through C1 to the output. Through the open transistor VT2 and diode VD1, capacitor C1 is discharged, and therefore this circuit has low resistance.

The polarity of the output signal from the multivibrator outputs changes with a frequency of approximately 1 kHz and its amplitude is about 4 volts.

Pulses from one output of the multivibrator go to connector X3 of the probe (emitter of the transistor being tested), from the other output to connector X2 of the probe (base) through resistance R5, as well as to connector X1 of the probe (collector) through resistance R6, LEDs HL1, HL2 and speaker . If the transistor being tested is working properly, one of the LEDs will light up (for n-p-n - HL1, for p-n-p - HL2)

If at checks both LEDs are on - transistor broken, if none of them lights up, then most likely the transistor being tested has an internal break. When checking the diodes for serviceability, it is connected to connectors X1 and X3. If the diode is working properly, one of the LEDs will light up, depending on the polarity of the diode connection.

The probe also has a sound indication, which is very convenient when testing the wiring circuits of the device being repaired.

The second version of the probe for checking transistors

This circuit is functionally similar to the previous one, but the generator is built not on transistors, but on 3 NAND elements of the K555LA3 microcircuit.
Element DD1.4 is used as an output stage - an inverter. The frequency of the output pulses depends on the resistance R1 and capacitance C1. The sample can also be used for . Its contacts are connected to connectors X1 and X3. Alternate blinking of the LEDs indicates a working electrolytic capacitor. The time it takes for the LEDs to burn is related to the capacitance value of the capacitor.