Paulmann electronic lighting transformer

[stevenA]

Another day, another project!

This transformer (Paulmann 105H) failed ages ago, and I looked at repairing it today. It is a 230v AC 50Hz to 12v AC 40khz transformer for halogen lighting.

The actual fault was the microtherm thermal fuse glued to the top transistor in the attached photo.

I replaced this with a wire link to test and it then worked - for a few minutes! The thermal fuse was obviously doing its job. This was the only apparent fault.

The transistor is now out (short circuit) and it took a few resistors and another transistor with it.

I have a few questions relating to repair of this supply:

1. Is the transistor failing the likely root cause of the problem?
2. The transitor beside the one that failed was a fairchild e13009f - is the one that failed probably the same in this basic smps (the fuse was glued to it and no markings show)?
3. I have identified an PHE13009 from NXP as a replacement - very similar, but not identical - should this be OK?
4. I cannot find a similar thermal fuse - does anyone have any pointers, or should I just leave it out?

Thanks for your help

[Th3_uN1Qu3]

Re: Paulmann electronic lighting transformer

Any 13009 will be good. For the fuse you can use a regular one.

[budm]

Re: Paulmann electronic lighting transformer

Well it uses Thermal fuse for a reason, it will go open if the temp reaches over 100c or the circuits draw more than 1 A, I would find another Thermal fuse.

[japlytic]

Re: Paulmann electronic lighting transformer

In the second picture, I see that a current sensing resistor is blown (but still readable).

[budm]

Re: Paulmann electronic lighting transformer

Good eyes japlytic!

[stevenA]

Re: Paulmann electronic lighting transformer

Thanks for your replies.

In fact, a good part of the resistors were taken out when TR2 shorted and put 370v across most of the circuit (see the attached image with most of the bad ones removed)

I have matched with symmetry and through reading various application notes most of the resistor values, and am now checking them against the theory of a half bridge inverter - reading the values on some of these resistors is no game for the colour blind, and R4 is illisible!

Attached is my latest working circuit diagram for posterity in case someone is interested, but no guarantees as to its accuracy (plus the component coding as some items were not marked on the board)

Unfortunately, I will probably not even repair this - these supplies are so cheap today that postage costs from component suppliers are more than the cost of a new (and probably better quality) Osram supply.

Nevertheless, a great learning exercise, and in case anyone else tries to fathom this circuit out, the secret for the unitiated is that the two secondaries on TI are inversed - when one goes high, the other goes low which sets up the oscillator.

[stevenA]

Re: Paulmann electronic lighting transformer

After analysing this circuit, I have a number of questions!

1. It appears that the circuit failed because TR2 "shorted". TR2 and TR1 have similar loads except that TR2 is also the starter, initially switched on by the DIAC. In many similar circuits, a resistor of 10ohms is placed after the DIAC, pesumably to protect the transistor,although there is nothing in this one. How do you calculate the effect of the DIAC conducting on the base of TR2 (the max current through the DIAC and R9 would not appear to be enough at 2mA to turn it on)?

2. The safety circuit of TR3 against shorts appears to be redundant! The max power of the output is 105W, which means that the collector current of TR2 and TR1 is around 0.5A in its maximal operating mode. The max current of the thermal fuse (attached physically to TR2) is 2A. At 2A, the VBE of TR3 will be .3v, (through R7, and R3,R4) well below the saturation voltage of 2v and would appear insufficient to cause TR3 to protect TR2 and the thermal fuse by stopping the DIAC triggering. Is my analysis correct?

This appears to be how the circuit failed once I had shorted the thermal fuse - before TR3 could conduct, the current in R7 - a 1W resistor - rose to a level that caused it to OC (somewhere above 1.5A). This meant that VBe rose to 60v on TR3 and that in turn failed.

Would a solution be to increase the power of R7, and to change the value of R3 to 480k so that at 2A, VBE of TR3 would be at least 0.6v - or have I got it all wrong and nothing can protect the circuit if TR2 starts to fail

Any input or corrections, would be appreciated.

[budm]

Re: Paulmann electronic lighting transformer

The circuit is setup as an self oscillate bridge circuit. TR1 and TR2 wil turn on alternately, not at the same time otherwise they will blow up.
The over current protection uses D4 to rectifier the AC signal on R7, when that AC is high enough, it will trun on TR3, the collector of TR should be connected to the base of TR2 to turn off TR2, right now it is drawn as a crowbar of the output which will blow up TR3. You need to look at the AC level at R7 with the scope.
I do not see the 10 Ohms resistor in your drawing. The bias for TR2 is from R9 for kickk start, not sure about the DIAC in your drawing. What is trigger voltage of that DIAC?

[ben7]

Re: Paulmann electronic lighting transformer

I bet one or both of the two red film caps caused the transistors to oscillate too fast, hence the overheated thermal fuse. I would replace the electrolytic cap and the film caps too.

[stevenA]

Re: Paulmann electronic lighting transformer

Thanks to both of you for your input.

Budm, my analysis sugested that in principle, the voltage across R7 would be insufficient to turn off TR3. When TR3 is off, the Diac (32v DB3) will no longer fire TR2 at the start of the cycle and so it will shut down. Are you saying that as drawn (I believe it is correct, and corresponds to app notes that I have read), even if TR3 is switched on, the safety circuit will not work?

Ben 7, if I understand correctly you are suggesting that the thermal fuse blew because of increased frequency, and not because of imminent transistor failure? I will replace the caps and see what happens. How do these two red caps influence the frequency?

[budm]

Re: Paulmann electronic lighting transformer

"my analysis suggested that in principle, the voltage across R7 would be insufficient to turn off TR3" The voltage has to be enough to turn on TR3, not turning OFF, you will need the scope to see the AC signal which has to be higher than 0.6V to turn on that rectifier, it has diode and filter cap to convert the AC into DC to turn on TR3. If it is DC detection, then there will be no need for the diode and the filter cap. If the Collector of TR3 is connected to the Base of TR2, when TR3 is turn on if the R3 voltage gets too high, it will short the Base- Emitter of TR2 which will turn TR2 OFF which will stop the Oscillation of the circuit.

[stevenA]

Re: Paulmann electronic lighting transformer

Thanks Budm - and sorry for the typo; I meant on (not off) as you say!

I think that we agree - to turn on TR3 requires 0.6v which means greater than 3A peak through R7.

In the original circuit, this is never achievable, as the thermal fuse will have already failed, being rate at 2A max (although not the designed failure mode, this is 50% over rating).

I have verified the circuit, and it is correct - TR2 appears to be designed to stop the DIAC triggering, and not to sink the current once it has triggered.