12 V AC Desk Lamp Power Supply

[aknewhope]

Hello,

I'm having an issue with a friends desk lamp. When he initially brought it to me, he said he had put a 120 V AC rated bulb into the 12 V socket. After that it hasn't turned on. I don't know if that actually caused the problem, but please let me know.

Inside the lamp base is a small 120 V AC ---> 12 V AC adjustable output (via potentiometer) power supply. It's made by a Chinese company (http://www.chiachien.com/English/en_Profile.asp).
After taking apart the power supply, I measured the main input resistor as open, even though there were no signs of damage. According the resistor color bands, it is 0.5 ohm. After replacing the resistor, the power supply started to show life again, but at a reduced voltage. The output never goes above 1-1.5 volts with the pot turned all the way up. I measured the resistance of the pot and it is very low when turned up. The bulb lights up, but only very dim. I know the bulbs work correctly, because I have tested them on a lab power supply. I checked for shorts between wires and the light frame, but it looks all good.

I decided it must be a cap or transistor, so I purchased replacement components. I replaced all of the capacitors, transistors, most diodes (could try all if you think necessary, but I checked them with a meter and they come back as good), and even the little blue diac. It still shows the same behavior though. I can hear switching noise when I turn up the pot, and one of the large NPN transistors gets hot after a few minutes. Does anyone have any idea why a PS like this would function like this? What do you suggest I try?


If you need a circuit diagram, parts list, or anything else let me know. Below are the transistors, and diacs (ordered two different manufacturers because I couldn't identify the original) I ordered.

https://www.digikey.com/product-detail/en/micro-commercial-co/DB3-TP/DB3-TPMSCT-ND/2513488
https://www.digikey.com/products/en?keywords=497-3099-1-nd
https://www.digikey.com/products/en?keywords=497-12230-nd
https://www.digikey.com/products/en?keywords=2n5551bu-nd

The two white wires going to the middle of the board are from the potentiometer/switch.

[Armadillo]

Since you have already tested almost all the components, then why not probe the primary of the transformer?

Edit: Anyway, advise your friend to replace the power supply as it is potentially hazardous because shocking voltage lead all the way to the potentiometer/switch.

Edit: Measure the potentiometer resistance also. [likely high ohms or open, since this control the intensity/power]

[Rasz]

it looks like it was made in early eighties, or even seventies
im amazed you would actually waste time and money (parts) trying to fix it instead of outright replacing, almost as amazed as when I realized your friend wants to use filament bulbs

[james_s]

it looks like it was made in early eighties, or even seventies
im amazed you would actually waste time and money (parts) trying to fix it instead of outright replacing, almost as amazed as when I realized your friend wants to use filament bulbs

I'm guessing you weren't around in the 80s, and certainly not the 70s, low voltage lamps like this all had iron transformers back then, these dirt cheap electronic things didn't appear until the late 90s-early 2000's.

Yes, I know you don't understand why anyone would try to fix something rather than throw it out and replace it, so why even comment on repair threads? It doesn't matter why he wants to fix it and it's not your choice to decide if it's worth it or not, so your smug condescending reply contributes absolutely nothing to the thread. I mean I don't use filament lamps for much anymore either but so what? You clearly don't get enjoyment out of solving a puzzle and fixing something, those of us who do often fix things just because we can.


Anyhow, for the OP, these transformers are Royer oscillators very similar to the ballast in a CFL. The behavior you describe sounds like an output short, have you tried powering the unit removed from the lamp with no load on the secondary? If the problem occurred when the wrong bulb was installed it's possible that the socket was damaged by forcing in a bulb with the wrong base. If it still misbehaves disconnected from the rest of the lamp, look closely for damaged windings on the transformer.

[capt bullshot]

This is one of these simple self-oscillating designs, having kind of a line-frequency amplitude modulated some 10kHz waveform at their output. Very common for this purpose. And the best feature: completely auto-destructing in case of failure.
First guess - if the transistors are still OK, check the red capacitors, at least one of them forms the resonant tank circuit together with the output transformer and is heavily loaded by that. Check the small electrolytic cap, this might have lost its properties due to age and heat
Second guess - throw it away and get a new one, these units are disposable by design

Edit: after reading the thread again: Third guess - operating the supply with too light load (120V light bulb) may have damaged the circuit or the transformer. Sometime there are warnings on these things: do not operate below minimum load!

[Rasz]

I'm guessing you weren't around in the 80s, and certainly not the 70s, low voltage lamps like this all had iron transformers back then, these dirt cheap electronic things didn't appear until the late 90s-early 2000's.

you are joking, right? it looks to be made by little chinese hands on manufacturing line of same vintage as this "beauty":
even C64 pcb is more modern.

Yes, I know you don't understand why anyone would try to fix something rather than throw it out and replace it

something worth fixing, not chinese death trap piece of shit made with seventies level of technical prowess, which apparently self destructed with bad load?

It doesn't matter why he wants to fix it and it's not your choice to decide if it's worth it or not, so your smug condescending reply contributes absolutely nothing to the thread.

Informing someone when its not worth fixing is valuable in itself. OP is not in Cuba nor north Korea, he isnt restoring rare piece of vintage history,  there is no reason to reanimate the dead when replacement is $10 on Amazon/Wallmart.

[aknewhope]

Thanks everyone. I will try the things you mentioned. In regards to just replacing the unit: Please provide a link. I would gladly buy a replacement if I knew it was that easy.

[james_s]

While you can get halogen lighting transformers that are essentially the same as that power supply, IMHO it's not really worth it unless that lamp has some kind of sentimental value. You can buy a whole new lamp for not very much money and get one that has a modern LED source.

I would still take a crack at repairing the original though just because it would bug me not knowing what was wrong. Have you verified that the fault is indeed in the power supply rather than a short in the socket or wiring?

[Ian.M]

If the potentiometer/switch has a plastic body and shaft or is an edge wheel type, with a plastic wheel, or is fitted to a grounded metal case, its probably safe - the manufacturer went to the trouble to sleeve the secondary to provide reinforced insulation, and the creepage distance appears to be over 4mm.  I've seen a lot worse in cheap Chinese PSUs.
OTOH if its a metal body pot with a two pin mains plug, its a deathtrap and best *NOT* rebuilt unless you can reliably ground it by retrofitting a three pin mains lead, or can replace the pot with a plastic one.

One cheap option would be to retrofit a DC-DC buck converter module from Ebay in the lamp base, hacking it to use the pot to control the output voltage so it dims, and feed it from an old laptop PSU.

[martinr33]

When I have this kind of problem, it is usually because I have mixed up the power transistors in some way. If one is hot and the other not, it is not oscillating properly.

[Rasz]

IMHO it's not really worth it unless that lamp has some kind of sentimental value. You can buy a whole new lamp for not very much money and get one that has a modern LED source.

damn it james, you trolled me

Ian even super cheap CFL ballasts come with a fuse and thermistor, this looks like an accident waiting to happen. Not to mention lack of filtering and primitive construction, I bet it generates lovely squiggles all over house installation.
btw how did the diodes come back good when first picture lower left corner one has a hole in it?

[Ian.M]

Its got a fusible resistor for inrush limiting + overcurrent protection, and is obviously designed to have a fairly good power factor, what more do you want?   

Yeah there's the little matter of interference across all the AM broadcast bands, but most users don't care about AM, they've all gone FM or digital.