Can't make this circuit work

[rodd]

Hi,
I decided to build the zener diode tester that appeared on the Nuts & Volts magazize, Nov 2014.
The circuit is basically a voltage source built around a multi widing transformer.
The schematics and the full article are attached.
I can't make the primary side to oscillate for some reason.
I have changed the transistors and double checked the circuit, but I can't find anything wrong.
Can anyone list some probable causes for this circuit to fail?
Thanks to all,
Roger Zener Diode Tester _ Nuts & Volts Magazine.pdf (1889.83 kB - downloaded 104 times.)

[xavier60]

As well as other possible construction mistakes, double check the phasing of the windings.
Measure the DC voltages on Q1's pins. Is Q1 getting hot?
Do you have an oscilloscope?

[rodd]

I Thank you for your reply.
Votages around Q1 are: Collector= 8V, Base= 1.1V Emmiter approx 0V.
Yes, I have an oscilloscope, and the circuit is not oscillating.
I cross checked the windings  identifications with the transformer datasheet and they are as they should.
Roger

[xavier60]

I Thank you for your reply.
Votages around Q1 are: Collector= 8V, Base= 1.1V Emmiter approx 0V.
Yes, I have an oscilloscope, and the circuit is not oscillating.
I cross checked the windings  identifications with the transformer datasheet and they are as they should.
Roger

Those voltages indicate Q1 being faulty or damaged. This is where it gets messy. If Q1 started off being good and there is another fault that prevents oscillation, Q1 will sink about 100mA continuously, causing it to overheat and fail. Now you have 2 faults.
I suggest replacing Q1. Check or replace Q2, also check R5 in circuit.
Power it up gradually increasing the voltage while monitoring the current draw.
When the current gets up to about 100mA the voltage drop(0.6V) across R5 starts to turn Q2 on which shunts current away from the Base of Q1, preventing Q1's Collector current from rising further.
Also monitor the Collector of Q1 for oscillation in case of it deciding to work.

[rodd]

Thnak you very much for the instructions.
I will do that and post the results here.
Roger

[rodd]

Hi,
The Vbe of Q2 is 0.6V, so it should be on.
So it seems both transistors are On at the same time...
Does it makes any sense?
Roger

[rodd]

I forgot to mention that I am using two 2N2222A instead of the suggested 2N3904.
Is this a problem?

[xavier60]

Hi,
The Vbe of Q2 is 0.6V, so it should be on.
So it seems both transistors are On at the same time...
Does it makes any sense?
Roger

You need to be keeping in mind all I said in post 3, so giving me all relevant information.
 What are the voltages on Q1 with respect to ground? What is the supply voltage? What is the current draw? Is Q1 getting hot?
2N2222A should be ok.

[rodd]

Thnak you for the replies, Xavier60.
Ther circuit is powered by a 9V battery.
yes, Q1 is getting hot and the Voltage at the test terminals is a little lower than Vbat.
Total current is 3.3mA!!!
Why is Q1 getting hot? Power dissipation should be 3.3* 10^-3 x (8.2-0.6)= 6.2 mW !!! This is strange....
I checked with the scope and the circuit is NOT oscillating...
Voltages are:
Q1: Ve=0.6V, Vb=1.3V Vc=8.2V
Q2: Vb= 0.6V

[xavier60]

Thnak you for the replies, Xavier60.
Ther circuit is powered by a 9V battery.
yes, Q1 is getting hot and the Voltage at the test terminals is a little lower than Vbat.
Total current is 3.3mA!!!
Why is Q1 getting hot? Power dissipation should be 3.3* 10^-3 x (8.2-0.6)= 6.2 mW !!! This is strange....
I checked with the scope and the circuit is NOT oscillating...
Voltages are:
Q1: Ve=0.6V, Vb=1.3V Vc=8.2V
Q2: Vb= 0.6V

Those voltages seem normal for that circuit in its non-oscillating state.
But the current isn't right. With 0.6V across R5, it should be about 100mA which would explain Q1 being hot.
 

[gnuarm]

Hi,
The Vbe of Q2 is 0.6V, so it should be on.
So it seems both transistors are On at the same time...
Does it makes any sense?
Roger

Yes, the voltage on R5 would rise as the base current of Q1 rises.   But as the voltage on R5 rises to the point off turning on Q2, that will shunt current from Q1 base and limit the current through the Q1 collector and emitter.  A rising current through T1-L1 will provide a voltage across T1-L4, so the action of Q2 prevents this feedback from driving up the current excessively. 

It's not clear what the circuit is intended to do.  Do you simply put a meter across the diode being tested?  Is the Yellow LED a failure indicator? 

[rodd]

The circuit is a low current "moderate hight voltage" 60V-100VCC.
You just plug a zener there and read the value of the zener voltage.
It is possible do the same with a variable power supply and a current limiting resistor, but when Vz gets high it is more difficult to find a PS in that range.
The yellow led is an indication that the overvoltage protection has been activated.

[rodd]

Any idea of what is preventing the circuit to oscillate?
Thanks

[wasedadoc]

The yellow led is an indication that the overvoltage protection has been activated.

The yellow LED in combination with its series zener has two functions:

1.  It shows the circuit is oscillating.

2.  It limits the voltage spike at Q1 collector when Q1 turns off and thus prevents exceeding Q1's voltage rating.

If the circuit is not oscillating there should be zero voltage at the test terminals.  Most likely reason that you are seeing some voltage and no oscillation is that despite your checking, the transformer is not wired as it should be.  Either an error by you or by the manufacturer.

[xavier60]

Mouser stocks the transformer. Does anyone know of any cheaper sources?
Id like to have a go at building the circuit.

 https://au.mouser.com/ProductDetail/Coiltronics-Eaton/VPH2-1600-R?qs=l%2Fkuw%2F8%252BD9jBGNNlh%252BShRQ%3D%3D

[Kim Christensen]

Any idea of what is preventing the circuit to oscillate?
Thanks

Try swapping pins 11 & 8 of the transformer.

And see if it starts to oscillate. If not, swap them back because we don't want to introduce a 2nd fault if the phase shift was correct to begin with.

[gnuarm]

The circuit is a low current "moderate hight voltage" 60V-100VCC.
You just plug a zener there and read the value of the zener voltage.
It is possible do the same with a variable power supply and a current limiting resistor, but when Vz gets high it is more difficult to find a PS in that range.
The yellow led is an indication that the overvoltage protection has been activated.

Is there any reason to use feedback from the transformer to the oscillator?  Why not just use a freestanding oscillator?  I know how to make those easily. 

[rodd]

for those who want to build this project, I found a "simpler" design
Please take a look at
I am attaching the schematics below

[rodd]

 wasedadoc made very interesting observation!!!
If the circuit is wired correctly and the circuit is not oscillating it is NOT POSSIBLE to read any voltage on the secondary!
Since I can measure ~ VCC on the test terminals there must be a short of some kind between the primary and secondary sections ,OR the windings does not match the datasheet.
I will look at this later on the day.
Thank you all for the replies.

[xavier60]

I mocked up the original circuit using 2 Bc548 transistors. It will oscillate with just about anything resembling a transformer connected to it.
Currently it's 15 bifilar turns on a green ring core from a CMC.
Disconnecting the feedback winding causes oscillation to stop and 100mA current draw. Shorting the Base of Q1 to ground causes 5uS pulses to appear at the Collector of Q1 clamped at the Zener voltage. The pulse width will be different for other values of transformer inductance and Zener voltage. It should be about  0.6uS for the original circuit values, if I have done my sums right.

[rodd]

I have found that the digital voltimeter module I used had the VCC connected to the voltage input terminal, so that is why it was showing and output of ~Vbat.
Now the aoutput is zero.
I might have shortened one of the windings of the transformer causing it to open.
I will travel tomorrow, coming back on Monday afternoon, when I will resume the troubleshooting
Thank you all for the replis

[rodd]

Hi,
I am back,
I foud the problem o the primary side, a cold solder joint...
The circuit now oscillate, but the maximum volatge I get on the test points is about 18V.
I tried to add a 3rd secondary in series Pin 6+ connected to pin 2 and pin 3 connected to D3.
Any ideas?

[xavier60]

Hi,
I am back,
I foud the problem o the primary side, a cold solder joint...
The circuit now oscillate, but the maximum volatge I get on the test points is about 18V.
I tried to add a 3rd secondary in series Pin 6+ connected to pin 2 and pin 3 connected to D3.
Any ideas?

Use your oscilloscope to check waveforms. Amplitude and wave shape at the primary winding should first be noted.
Then comparison to the other windings to confirm that phasing is correct.

[wasedadoc]

If D2 LED is lighting up there is a zero to about 35 Volt pulse on the transistor collector which means about 26 Volt pulse across the transformer primary.  The two secondary windings in series should transform that to about 52 Volts and give close to that on the rectifier output.

[rodd]

Thnak you for the replies.
D2 is not lit.
There is a square wave in T1-L1, and the voltage at Pin9 with respect to ground is 12.1Vpp.
The voltage at the anode of D3 (1N4937) is 35.8 Vpp. with no zener connected to the test terminals.
However the rectified voltage at the test terminals is 15Vcc (maximum).
In my setup, C3 is made of two Epcos 0.1uF x 630V poliester capacitors.
What is wrong?