12V, 1.5A PSU Adaptor Inductor Overheating

[mindmend]

This 12V, 1.5A 240V mains adaptor is being used to power an adsl modem router. Usually it runs quite cool when it changed and I noticed the case was getting excessively hot.

Opening the case revealed the mains input filter inductor is getting far too hot. I replaced it. The replacement also gets hot but not as hot, it reaches 70C after being on for 10 mins supplying 0.7A (cools down off load) - this is with the case open. See attached images.

I don't know how hot is normal for this inductor, please anyone with experience I would value your opinion.

The mains input circuit is a bit unusual in that bridge rectifier diodes are connected directly to the mains input. Output from the rectifier feeds the inductor coils. Most PSU's I've seen have the rectifier after this inductor. There are two reservoir caps one connected to the output of the bridge rectifier the second on the output of the inductor coils.

As I'm writing this a thought comes to mind if the bridge rectifier reservoir capacitor is open circuit then all the ripple current will be flowing through the inductor coils to the second reservoir capacitor connected to the output of the inductor coils, it might explain the heat. Anyway I'll post this to see what your thoughts are.

Photo's:

[mindmend]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

Update:
Replacing the reservoir cap connected to the bridge rectifier has not reduced the temperature of the inductor.

[redwire]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

I would look for an open rectifer diode, or low ohms value NTC inrush limiter, or some other reason the ripple current is so high through the inductor.
There's two electrolytic caps on the primary side? I'd check what value the other has.

[mindmend]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

I would look for an open rectifer diode, or low ohms value NTC inrush limiter, or some other reason the ripple current is so high through the inductor.
There's two electrolytic caps on the primary side? I'd check what value the other has.

Thank you for your reply.

The four rectifier diodes measure correctly. The cap on the input (output of the bridge rectifier) to the inductor coils is 10uF 400V the output of the filter has a cap of 22uF 400V.
The mains input connects through a fuse directly to the bridge rectifier, I don't see anything to limit the current into the filter inductor.

I have tried putting a 33uF cap in place of the 10uF to get less ripple before the filter but this experiment didn't seem to change the temperature noticeably.

I have no idea what temperature to expect the inductor to be with a load of 0.7A - any comment on that please?

[redwire]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

0.7A is pretty much the limit for most of those CM chokes around 5mH or less, it depends on the technology (nanocrystaline vs ferrite). I think the DCR around 1/2 ohm makes the heat. They are rated for a 60C rise at their full current, which is hot.

For an 18W adapter, 0.7A seems way high... unless the multimeter is not true RMS or the power factor is really terrible. The caps have over 340V on them?
I wonder if something else is making the adaptor's efficiency really low or the 12V load is high past 1.5A

I thought I see a black NTC next to the blue mov in heatshrink. If it's high value they get very hot and low value the riple current peaks get high. But likely not a problem.

[mindmend]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

For an 18W adapter, 0.7A seems way high... unless the multimeter is not true RMS or the power factor is really terrible. The caps have over 340V on them? I wonder if something else is making the adaptor's efficiency really low or the 12V load is high past 1.5A

The mains input current is 0.037A (meter is not true rms) with the 12V load of 0.7A, sorry for ambiguity.

The voltage on the caps is 329V, with mains input of 239V AC.

0.7A is pretty much the limit for most of those CM chokes around 5mH or less, it depends on the technology (nanocrystaline vs ferrite). I think the DCR around 1/2 ohm makes the heat. They are rated for a 60C rise at their full current, which is hot.

This is useful info, in this case the rise above ambient is about 50C but that's under half load.

I thought I see a black NTC next to the blue mov in heatshrink. If it's high value they get very hot and low value the riple current peaks get high. But likely not a problem.

There is an NTC, it is running just warm. All other components are running warm ~38C. Except the switching (FET?) on the heatsink it's running about 45-50C with 12V loaded to 0.7A.

Before the PSU was overheating the case felt warm maybe to the touch ~42C, if the choke is the main source of heat the choke must have been getting pretty hot for the case to feel warm. Did it eventually giveup the ghost.
Do you think this is just a 'bad' design or something is causing the choke to overheat?

[Hitto]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

Is the first capacitor placed before the inductor?

The cap on the input (output of the bridge rectifier) to the inductor coils is 10uF 400V the output of the filter has a cap of 22uF 400V.

That inductor is made for AC (230V). If it is placed after the main capacitor or after the bridge rectifier then it will work with rectified DC (400V DC).

[mindmend]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

Is the first capacitor placed before the inductor?

That inductor is made for AC (230V). If it is placed after the main capacitor or after the bridge rectifier then it will work with rectified DC (400V DC).

Before, it's only 10uF, the second cap is 22uF after the CM choke. Ripple current will be passing through the choke to the second cap.

I don't have an oscilloscope so cannot see the ripple.

[Hitto]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

Usually the choke is placed before the bridge rectifier:


It looks like, in your case, it is placed after the bridge rectifier. You can remove it and bypass with a pair of fuses/wires.

Do not worry too much about removing it: this power supply is a flyback switching power supply and the ripple on the secondary will be quite high regardless. And this design lacks some components (the blue capacitor, ceramic, at CY1).

[mindmend]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

Usually the choke is placed before the bridge rectifier.
It looks like, in your case, it is placed after the bridge rectifier. You can remove it and bypass with a pair of fuses/wires.

Yes that's right Hitto, the choke is connected after the bridge rectifier.

The thought of bypassing the choke did come to mind. I haven't done that (yet!) for these reasons:

• Something changed - the PSU did not previously run hot.
• Switching noise injecting back into the mains supply.
• Reduced protection from incoming voltage spikes on the mains supply.

My first choice is to find out what caused the psu to suddenly get much hotter than before. And to know what temperature would one expect the choke to reach when supplying the 12V output load of 0.7A.

[petehall347]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

does it stay cold without a load ?
i should think you could get an idea on ripple using a cap on your meter probe .

[mindmend]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

does it stay cold without a load ?
i should think you could get an idea on ripple using a cap on your meter probe .

Yes the choke remains cold off load.

[Hitto]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

Ripple on the mains is usually not so bad to cause an inductor to overheat that much. You may place a X2 100nF capacitor just before the inductor, it will help filtering high frequencies.

Remove that white glue: maybe it has decayed and is leaking current and/or has became capacitive.

Is a mylar between the 22uF electrolytic and the inductor? Desolder it and measure it.

Are the pins, on the inductor, soldered well? Reflow solder between copper wires and the four pins.

Check for any ceramic SMD capacitor: they like to fail short or develop cracks. Since they tollerate a lot of current they may be defective while appearing good.

[mindmend]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

Ripple on the mains is usually not so bad to cause an inductor to overheat that much. You may place a X2 100nF capacitor just before the inductor, it will help filtering high frequencies.

Remove that white glue: maybe it has decayed and is leaking current and/or has became capacitive.

Removed the white rubbery glue. I don't have an X2 100nF capacitor to hand.

Is a mylar between the 22uF electrolytic and the inductor? Desolder it and measure it.

Not mylar, it's a 1nF ceramic disk connected in parallel with the 22uF cap. I will replace and report.

Are the pins, on the inductor, soldered well? Reflow solder between copper wires and the four pins.

I had already replaced the choke (the original one looked and smelt like it was seriously cooked). The solder joints are good.

Check for any ceramic SMD capacitor: they like to fail short or develop cracks. Since they tollerate a lot of current they may be defective while appearing good.

There are no SMD capacitors anywhere on the PCB.

The mains input current is 0.037A (not true rms meter) = ~9W.
The output load is 0.67A @ 11.78V = ~8W.
Looking at the efficeincy it appears quite good. Which makes me wonder if the circuit is funcitoning correctly now the choke is replaced.
How hot would you expect the choke to get given this load?

[budm]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

What is the Voltage drops on the windings of that common mode choke with load and no load, check in DCV and ACV.

[mindmend]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

What is the Voltage drops on the windings of that common mode choke with load and no load, check in DCV and ACV.

On load:
400mV DC
30mV AC

Off load it's close to zero.

As a last resort I will remove some windings from the choke.

[mindmend]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

Check for any ceramic SMD capacitor: they like to fail short or develop cracks. Since they tollerate a lot of current they may be defective while appearing good.

Update:
Sorry Hitto, I should have said no SMD caps in the input rectifier and filter areas.

[redwire]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

If the CM choke has say 0.4V across each winding that is say 1 ohm, then it's around 0.32W it is dissipating and I would not expect it to get way hot. Manufactures will sometimes stuff in the highest value mH (but high DCR) to ensure the device passes EMC tests.

But you say it's changed so the only thing left I can think of is the SMPS transformer could be saturating? It might be from mains going up from say 230VAC to 250VAC. You could try a variac with a little lower mains.
edit: It's a current-mode controller after spotting R18,R19,R20 2R7.

[mindmend]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

If the CM choke has say 0.4V across each winding that is say 1 ohm, then it's around 0.32W it is dissipating and I would not expect it to get way hot. Manufactures will sometimes stuff in the highest value mH (but high DCR) to ensure the device passes EMC tests.

But you say it's changed so the only thing left I can think of is the SMPS transformer could be saturating? It might be from mains going up from say 230VAC to 250VAC. You could try a variac with a little lower mains.
edit: It's a current-mode controller after spotting R18,R19,R20 2R7.

It did significantly change. On opening the case the original choke had obviously overheated and was getting hotter than 70C but I didn't measure that - I just replaced it.
I'm wondering if in this design the choke always ran hot could it have broken down and seriously overheated.
I don't know normally how hot the choke gets with this design, which makes me wonder if 70C is normal. That's what I was hoping someone could give an opinion on.
I will reassemble the case and power the modem like before that's the only way I can evaluate the case temperature. I will do that and give an update.

I don't have access to a variac.

[mindmend]

Re: 12V, 1.5A PSU Adaptor Inductor Overheating

I will reassemble the case and power the modem like before that's the only way I can evaluate the case temperature. I will do that and give an update.

With psu reassembled and original modem load the case temperature is warm but not hot. The CM choke must have always been running hot and with time destroyed itself.

To minimise this recurring I shall unwind half the turns from each coil of the choke.

A big thank you to every-one who contributed to this Sherlock Holmes story. I have learnt much about the switch mode psu and the willingness of others to help solve problems.