Help, switching p.supply doesn't boot but works

Thread Starter

Giovanni Gozzi

Joined Mar 11, 2017
5
Hi everybody, and thank you for your precious help.

I'm repairing 3 light scanners (club lighting) and they have the same problem, the power doesn't start up or in some cases after 2-3 seconds. The third one was working but after I tested it on bench without load (it has 36ohm internal resistor on +5V and about 100 on 12V) it starting have the same problem. If I give just a +12V pulse from an external battery after D14 (see figure) they all start and keep working perfectly connected to load (18A on 5V, 8A on 12V). They are up from 4years already 24/365 but in one I changed both power-transistors, both condensers C5-C6, the C9 (in my circuit is 2,2mF), even the controller TL494, tested all diodes and transistors junctions, and I made all test ever bypassing the components near TL494 dedicated for protection (short, overvoltage, overcurrent and so on).
So brief as I could be, I suppose the responsability is on the power stage, cause it produces only 0,87 to 1,4V (on D14) when connected to 220V on all the three exemplars I tested: that also means the switching is working.. at low condition. At this low power conditions in my opinion the controller doesn't work, not activate the couple T3-T4 and leave the power switching side to operate as minimum.
I need to overcome this problem but it's not easy, I already tried to short the NTC to improve the current at the start without success, I would increase the capacity of C9 to let more current flow.. and try to rise potential on D14 to let the controller work, but I'm afraid it is calculated to have resonance with the transformer or to rise up too much the current. Maybe I can have a try for some seconds adding 1mF to C9 and seeing if something changes. There is also C10-R11 which looks like a "cos-fi" reduction, instead of a trick to produce at boot a sort of peak to startup the controller, I don't think they may help the process.

Does someone here has faced the dame issue before? I'm quite disoriented. Thank you a lot!

PS: the circuit is really similar but not the same, differences are on the circuit fault checks areas
 
Last edited:

Thread Starter

Giovanni Gozzi

Joined Mar 11, 2017
5
I would give also a couple of important info. There is no transformer Tr3, the controller checks the current (probably) (with pin 16) by testing Tr1 where connects with D5. Also on Tr1 there is not the 3rd coil. The coil 1 of Tr2 is directly connected to the transistors T1-T2 for one side, and directly to C9 on the other.
 

Thread Starter

Giovanni Gozzi

Joined Mar 11, 2017
5
thank you from answer, I didn't try yet but I'm focusing more on rectifier.. the behaviour suggest to search fro components under stress, and I read about degradation of Schottky at load/high temperatures for long time. It may cause higher forward drop.. which cause to me the problem. Also it justify why it always boot after the first run on load.. the junction is hot and forward drop lows for this reason. I'm just looking for a sobstitute.. but it's not that easy
 
In some switchmode supplies,you need a way to kick it into oscillating, In other words, it's chicken and egg problem. You can't get Vcc on the PS chip (TL494) until the chip is up and running. There is no independent 5V supply. C11, C16 and C18 do provide some asymmetry.

So, I was looking for something that would act as a "short" briefly on start-up that would provide the start-up kick.
 

Thread Starter

Giovanni Gozzi

Joined Mar 11, 2017
5
mates, I decided that was time to cut the head's problem and I placed some old phone power charges to "kick up" the controller with a diode after D14. It works good and I can spend time on other real work issues (which make me money). Pity I couldn't figure out in the end what the problem was, I replaced the rectifiers but I just improved a bit the kick in, it was just enough to heat up a bit with iron the cooler to see it turning on.. but we can't consider it "repaired".
I send a big hug to all helped me here :)
 
Top