Hi!

I have two driver boards for xenon flash light. I received them faulty, with some signs of previous repair attempts. For learning purposes I am also trying to fix...

Some notes:

• I rev. engineered the circuit board and put it into ltpice (hope I did it right). Since I did not found any xenon flash model, I replaced it with a voltage dependent switch and a resistor on the drawing on the right.
• I don't have any clue if the board was working at all anytime
• I don't know if the part types or values are the original ones or due to the previous repair attempts
• Transformer windings are measured with LCR meter, but don't know how to incorporate into the model the core properties (I took apart one of the power supply transformers, and it had around 40-50 windings on primaries, and around 900 on high voltage secondary side. I have the more precise notes somewhere...)
• The PCB had a marking PULSAR24V from which I suspect, that it was designed to be supplied from 24V (probably DC, though it has a diode at the power input terminal, but probably for polarity protection)

The power supply part seems to be a self oscillating flyback type to me. Something similar described in this topic:
https://forum.allaboutcircuits.com/threads/simple-rcc-flyback-switching-power-supplies.170463/

The BC639 transistor were surely faulty, which I replaced several times, but after applying the supply voltage starting from low voltage and turning up towards 24V at some higher voltage, the current suddenly shoots up and the transistor smokes again and again. Till this happen I don't see any oscillation activity starting so probably the transistor does not turn off.

I tried to build the power converter part of the circuit on a breadboard without the right part of the circuit (only till C8, omitting anything else to the right) and it starts to oscillate but at around 160kHz and no significant output voltage is produced on the secondary. Only a few volts.
The 160kHz seems to be a bit much, since testing the transformer standalone from a sig.gen. it seem to produce the higher output on the secondary somewhere about 20-25kHz.
But I cannot make the oscillating frequency lower than 110kHz by changing some part values. Oh, and it only works if I omit also the C5 capacitor. When I add it it does not oscillate, like in the PCB.
Maybe this oscillation starts with some parasitic effects caused by the breadboard or I don't know. But probably this is not how it was designed to operate

Any idea how I should progress with the investigation? What could be the error or what should I check/test?
Really would appreciate any help/guidance. Thanks!

(Btw the ltspice simulation more or less works...)

 

Since it does not seem to oscillate, I suggest checking the C7, R7, L2 string. L2 is what is required for feedback,which is needed for oscillation. also check D3 for being either open or shorted.That would cause problems.

 

Hi Itar,
Please post your LTS asc file.
E

 

Hi I,
I was clearing out unwanted stuff, and one is a camera flash.
I stipped it down and applied voltage, which started the familiar hissing sound from the switching section, then the LED came on and I shorted the 'button' and flash.
I can now try connecting meters and oscilloscope and 'see' what's going on.
Being carefull of the high voltages.
You may want to try finding an old flash unit to 'play' with.
Cheers, C.

 

The circuit should also function at a lower voltage that should not damage the transistor. If you disconnect one side of D3 then you will be able to investigate the oscillator circuit at a reasonable voltage to see if it will oscillate as intended. Do you know if the BC639 transistor is even the correct part? What parts appear to have been changed??

 

Thanks All for the suggestions so far!

MisterBill2: I have checked all components (for capacity, ESR, resitance, diode conduction...), and tried to replace some of them like C7,R7, D3, D4... I also tried to check the transformer with ring test for potential shorts and it seems to be ok, but it is a bit hard to decide for me (ringing of course dampens, but maybe not too quickly on all windings)...
I tried to play a bit with lower voltages, but not exhaustively, focused more on the breadboard setup (it was easier to change parts out of circuit than soldering/desoldering), but probably will go back to the circuit board and do some more tests on lower voltages then. I did not had a clue if oscillation would start only from some specific votage or not.
I don't know if BC639 is the original part or not, but at least these were in both of the faulty boards I got. It also seemed that they were already replaced because of weary solder pads of the transistor. So maybe I am chasing something with a wrong transistor...
ericgibbs: attached the .asc (with the included libs)
camerart: good idea, may try to dig up some old camera also...

 

Is this the one?
https://www.karillon2.com/Articles/pulsar/strlf.jpg

 

probably it would be a good choice to look up the specifications of the BC639 transistor and see if it's voltage and current ratings are adequate. If the system will oscillate at 6 or 12 volts then I would not go higher unless the light output was low.
Unless they were for some special application, 24 volts does not seem like a common photoflash package voltage.

 

Ian0: I'm not sure about the type, cause I only have the PCB-s. Made some photos...

Soldered everything back to PCB and the issue is the same. Tried with 5V and moved to 15 with 150mA limit on the power supply. No oscillation. When I turn the supply to 15V, the current consumption goes up from around 80-90mA to 150mA in 20-30 seconds and the supply limits more current to flow. And the transistor warms up considerably.

Made some scope measurements on L1 and to my surprise the voltage on top side is negative to bottom side. I assumed it should be the other way around after switching on.
On the scope picture:
Ch1 is top of L1 (compared to ground)
Ch2 is bottom of L1 (compared to ground)
Ch3 is Ch1-Ch2


When I remove C5 capacitor from the board it start to oscillate at 145kHz (like it did on breadboard) and has only produces 1.4V on output caps.

 

Given the ratio that I would guess the transformer has, that low output voltage makes no sense. Have you verified continuity of the secondary that is supposed to be producing the high voltage?? What I see in the picture of the scope trace does not show any oscillation. And if it oscillated at 145 KHZ, we should see that on the scope traces. So please show us the base to common voltage waveform when it is oscillating, and also the collector to common waveform as it oscillates.

The very low output voltage points to either an open circuit or a faulty big capacitor, or a failed D3. Or maybe an open connection in the charging system.

 

Previous scope captures were done when the circuit was not oscillating (it only start to oscillate when I remove C5). It shows the start of circuit...

Picture of removed transformer with signal generator driving L1 (CH1) and L3 (CH2). It shows around 20x increase in voltage, but without any load.




Scope captures with transformer in circuit (without C5), 5V on input:
1. base (CH1), collector (CH2):





2. base (CH1), L2 top (CH2):





3. base (CH1), L3 top (CH2):





I more and more suspect that one of the transformer winding is shorted, since there is almost no voltage is generated on the L3.

 

I think that I am seeing about 5 volts peak to peak on the collector. I am guessing an open connection on the high voltage one. If it will oscillate then both the base signal feedback winding and the collector winding must be OK, or it would not oscillate at all. An open high voltage winding will explain no high voltage. But so would an open D3. Much cheaper to replace, as well. Have you done a resistance check on the secondary winding yet??And on D3. Maybe try disconnecting one end of D3.

 

Resistance of high voltage winding is about 54ohm. I have checked D3 also and it is ok, not open.
Btw, shouldn't I see much higher voltage changes on high voltage winding even in case of open or disconnected D3?

 

Resistance of high voltage winding is about 54ohm. I have checked D3 also and it is ok, not open.
Btw, shouldn't I see much higher voltage changes on high voltage winding even in case of open or disconnected D3?

YES, you should. Is there any chance of getting to see a copy of the original circuit and parts?? There may be something we have missed.

 

any chance of getting to see a copy of the original circuit and parts??

What do you mean on "copy"? I attached photo of circuit board in #9 post. Do you need more?

In the meantime I have tried to drive the flash part of the circuit from a separate DC-DC converter. With 300V fed into right side of D3 (D3 disconnected) the flash works (on both of the two boards).

But I still have the ambition to fix the converter part or build one with same configuration which is working.
I tried to attach a different transformer, of course with different number of turns and inductance (lower) to see if it starts to oscillate, but unfortunately not. Played a bit with different R6,R7 resistor values without luck.

Is there any documentation with details of this converter type? How the different part values can be calculated? What is the purpose of C5 or C6? How parts should be modified if I have a different transformer like this one which I tried:
https://www.aliexpress.com/item/100...o.order_list.order_list_main.5.30c81802mbVrTh

 

Finally I made the converter work on a breadboard with the transformer I linked into the previous post. Since it has different turns ratio it produces only around 73V at the output from 10V input. It is not enough to drive the flash tube, but at least it works.
I had to reduce R6 significantly to make it work. This is what I put together:




I will try to look for a suitable transformer, because probably this was faulty on the original circuit board. Or I have to rewind it, but it looks challenging with so many secondary turns...

 

If it is a break in the high voltage secondary, it could be at a terminal of the transformer. A very close inspection with a good light and a good magnifier might find the break and save a whole lot of work. Certainly worth the effort. OR maybe it really is 6000 ohms..

 

If it is a break in the high voltage secondary, it could be at a terminal of the transformer. A very close inspection with a good light and a good magnifier might find the break and save a whole lot of work. Certainly worth the effort. OR maybe it really is 6000 ohms..

Thanks. I will inspect the transformer again. But where does this 6000 ohms come from? I measured 54 ohms for high voltage winding.
Btw, isn't it rather a shorted turn that could cause the problem? Sometime back I had an issue with a very thin high voltage winding in an old scope power supply. At that case it was a short.

Anyway, at least I could make that this type of self oscillating converter work, with a different transformer.

 

Certainly a shorted turn will be a source of a problem, as it will tend to increase the loss so that the oscillator will not work. So the transformer may have multiple failures.

 

Just to report the final result...

I rewound the transformer (~1000, 50, 44 turns) and the circuit works. Running it from 10V DC it flashes around every 1-2 seconds.
The resonance is not on a constant frequency, it starts low, then goes up to around 16-17kHz till Xenon tube flashes and output capacitors start to load again (up to around 280V). It is also audible...
I am not sure if this changing frequency is normal, but this is how it works.

All in all, is working! Thanks for the help!