Single Transistor Continuity Tester

[simonyovink]

Hi! Long time lurker / EEVBlog viewer. Excited to get started on the forum.

I am making a continuity tester based on a single BJT powered by a CR2032 coin cell. What I have so far is a circuit that works, but to be honest, I am not sure why.

I'm using a single PNP 2N2907 (found them in an ancient parts bin at work) and I found the circuit is very dependent on that particular transistor. I've added diodes D2 & D3 to increase its sensitivity to around 22ohm. Anything above that and the LED won't light.

Also I should note R4 is serving as the device under test the simulation. I've attached an LTSpice sim too.

Thanks everyone! Looking forward to hearing from the great minds of this forum.

[PGPG]

Hi! Long time lurker / EEVBlog viewer. Excited to get started on the forum.

I am here since few days, but electronic is my hobby since 1968 (when in age of 10 I build my first detector radio receiver) and job since 1983.

I am making a continuity tester based on a single BJT powered by a CR2032 coin cell.

Explain what is the reason to use relatively complicated schematic. If cell voltage is enough to light the LED then connecting LED to cell through some current limiting resistor + tested resistance should light LED if there is connection (continuity) and not light LED if there are no connection. So tester would contain only cell, current limiting resistor and LED.
Do you want to get LED being light when the tested resistance is so high that LED connected through it to cell will light so faint that it won't be visible?
If yes than what is the expected limit value of the resistor (below this value LED lights, above not)?

[simonyovink]

Hi! Thanks for your input.

I am here since few days, but electronic is my hobby since 1968 (when in age of 10 I build my first detector radio receiver) and job since 1983.

What a long career! What kind of specialties have you developed? I do traffic signal engineering right now.

Explain what is the reason to use relatively complicated schematic. If cell voltage is enough to light the LED then connecting LED to cell through some current limiting resistor + tested resistance should light LED if there is connection (continuity) and not light LED if there are no connection. So tester would contain only cell, current limiting resistor and LED.

You make a good point - however, I find that simple circuit too insensitive. For example, if I test and resistance is 100ohm, LED will still light fairly bright, even though there is not true continuity in the circuit. If I make current limit resistor too high, then the LED will not light brightly even with perfect 0ohm test.

If yes than what is the expected limit value of the resistor (below this value LED lights, above not)?

I should have mentioned this in the original post - I would LED to light with test value <50ohm (in simulation, R4<50ohm, LED has current). Above that, no light or very dim light.

Eager to hear more of your thoughts.

[PGPG]

What a long career! What kind of specialties have you developed? I do traffic signal engineering right now.

At first I was working as academic teacher (profesor was giving lecture and 4 instructors had exercises (each with 3 student groups)). After 3 years older colleagues asked me how I am doing it that since 3 years each time my students achieved the best results in the exam. I widened my eyes because I didn't know that after each exam they sort the list by the instructor and count the average grade for each of us. At first they thought I was lucky to come across better groups, but 3 times in a row - not possible.
Those time in Poland (a communist country then) for my salary I was able to live but if I wanted to change it into $ my month working was worth $13 (thirteen and a month, not a mistake). Communist economics is incomprehensible to anyone from the normal world.
Working still there I setup (with my brother) our small business (imagine the capital we had for beginning). Our first products were Eprom programmer and GAL and serial EEPROM programmer (I found them here: https://ep.com.pl/files/7162.pdf). Soon we both quit our jobs. In 1993 we designed 8051 education system (http://pliki.micromade.pl/pdf/dsm_kk.pdf). Then we focused on access control systems and we continue to do so today.

I would LED to light with test value <50ohm

Replace D2 with resistor.
For D1 from Mouser i found datasheet = Error 404.
So assuming red LED is about 1.8V, and may be need 10mA I get R=(3-1.8-0.2)/0.01 = 100Ω (0.2 is assumed voltage at Q1 switched on).
But we currently use RED LEDs that need only 2mA to shine bright. So if you need less current than inverse proportionally change the resistor value.

In your current connection when Q1 switches on the determined voltage elements are connected directly to cell. As cell has no infinite current efficiency than its voltage drops in such situation.

If Q1 output is 10mA and its h21 is may be 200 then base current is 0.05mA. Multiplied by 47k gives you 2.3V. So the output current will rise from 0mA to 10mA when voltage (betwean D3R2 common point and +3 will vary from 0.7V to 2.3+0.7 = 3V. It not gives you a point for fast switch form 0 to 10mA.

Change R2 from 47k to 1k to make it only protect transistor base against too high current.

Now experiment with R1 value.
You can add second 1N4148 in serie with D3. That way you will get smaller current going through R4 so you will have bigger value of R1 and less load for 3V cell. It is not intended for high currents load.