Task TK-940TX-DF Blowing fuse

[momaka]

Re: Task TK-940TX-DF Blowing fuse

Then i connected 230V across the AC legs of the removed rectifier (with the bulb in series) and the bulb didn't glow - the voltage across the DC legs is 211V DC and 96V AC - is that good?

Looks like it.
Without a capacitor on the output of the bridge rectifier (i.e. between + and -), you get sinwave pulsed DC, which will read somewhat close to the RMS value, but not quite (will vary between multimeters)... so 211V DC doesn't seem bad. If you take the green X2-cap from the PSU's AC side and connect that across the bridge rectifier, then measure the voltage on the bridge rectifier, you should get around 340V DC stable. AC measurement would be irrelevant. Just make sure the X2 cap is rated for 250-275V AC / 400-450V DC (usually for X2 class, that is the minimum rating anyways, but always worth checking first before plugging things in. )

Tested across the DC holes in the board and get 35M ohm that slowly decreases.

OK, but while it is decreasing, does it still stay in the MOhm or high KOhm range after, say, 10-20 seconds? You can also try hooking that 16V laptop DC power adapter across the two caps (mind the proper polarity, of course) and see if it sparks / short-circuits. If yes, the short is still on the DC side somewhere.

Surely we must be getting close to finding the culprit by now!?

Close? We should have had the PSU fixed by now.
I still can't fanthom what is causing an apparently almost dead short-circuit with power applied, yet nothing when checked with a multimeter. There are not only no active (transistor) devices, but no high-current paths available on the primary DC side (aside from the caps being damaged somehow... but I doubt it.)

Bad comes to worse, you can take the two high voltage electrolytic caps, wire them in series with proper discharge/balancing resistors, then hook the bridge rectifier to them and test with the bulb. The caps should charge to around 340 DC total (about 160-170V across each cap) depending on line voltage and bulb shouldn't glow. If it does, either the caps are bad or you wired the circuit incorrectly.

If you do build that circuit, please stay safe!
Linking an Electroboom video here to show you what could go wrong.
https://www.youtube.com/watch?v=sI5Ftm1-jik

yea, i'd luv to have a 42.25 megawatt power plant! dont need to pay no stinkin' power bills no more!

Just pray it's not a nuclear power plant close by.
Then again, coal wouldn't be that much better either. *cough cough*

[ChaosLegionnaire]

Re: Task TK-940TX-DF Blowing fuse

And just for fun, the power dissipated from that resistor with that voltage and current would be 42250000 Watts (42.25 Mega-Watts!!!)

yea, i'd luv to have a 42.25 megawatt power plant! dont need to pay no stinkin' power bills no more!

[socketa]

Re: Task TK-940TX-DF Blowing fuse

i removed the rectifier and the bulb stopped glowing - so the AC side seems to be good.
Then i connected 230V across the AC legs of the removed rectifier (with the bulb in series) and the bulb didn't glow - the voltage across the DC legs is 211V DC and 96V AC - is that good?
Tested across the DC holes in the board and get 35M ohm that slowly decreases.
The bridge rectifier still tests good, according to that test that was done earlier on.
Surely we must be getting close to finding the culprit by now!?

[momaka]

Re: Task TK-940TX-DF Blowing fuse

Well, that was uneventful
The element was drawing 1.5kW during the period of six minutes, and the DC voltage across the caps (the rectifier DC terminals) only rose from 1V to 3V.
No fireworks either

Wow, something must be shorted "hard" for that much current to pass and nothing burn out.
Heck, I'm surprised the bridge rectifier survived. 1.5 KW @ 230V line is ~6.5 Amps of current. Probably was running it right at the limits.

3V across the caps? That's a little higher than 2 diode voltage drops added together (probably originating from the bridge rectifier when loaded with such current.)

That brings another item to check: is your EMI/RFI coil/choke oriented properly - i.e. AC coming in on one side and exiting the other through the other to the bridge rectifier. If the leads happen to be flipped by 90 degrees on the PCB, you'll be shorting out the AC input... though that would have showed up as s short-circuit when you checked the input of the PSU. So that may not be the issue here... but still check.

The rectifier tests good after that - so doesn't that suggest that the fault is either in the green X Capacitor or the 1Mohm resistor?

Perhaps it could be the green X-cap indeed. Though if the bulb doesn't glow when the bridge rectifier is removed, then the short has to be on the DC-side.

Try it and see so you can verify. The green X-cap is only there for EMI/RFI suppression, so you don't really need it for anything.

The 1 MOhn resistor can't be the issue, because, again, resistors always go open-circuit or high-resistance when they go bad. By Ohm's law, the input voltage at the wall would need to be 6500 KV (or 6,500,000V if you like) to drive 6.5 Amps through that 1 MOhm resistor.
And just for fun, the power dissipated from that resistor with that voltage and current would be 42250000 Watts (42.25 Mega-Watts!!!)

[socketa]

Re: Task TK-940TX-DF Blowing fuse

The rectifier tests good after that - so doesn't that suggest that the fault is either in the green X Capacitor or the 1Mohm resistor?

[socketa]

Re: Task TK-940TX-DF Blowing fuse

Well, that was uneventful
The element was drawing 1.5kW during the period of six minutes, and the DC voltage across the caps (the rectifier DC terminals) only rose from 1V to 3V.
No fireworks either

[momaka]

Re: Task TK-940TX-DF Blowing fuse

Could it be possible that that 1M ohm resistor, that's between the AC input rails, is faulty, and only shorts out when high voltage is applied?

No, resistor failure mode is always open-circuit or higher resistance.
Yes, you can remove it and do the bulb test again, but it shouldn't change anything.

Yeah i have a spare stove element that i can use.
How long should i "let it rip" for? Until the caps/rectifier explode or one of those voltages shows up on the meter?

Maybe a minute or two or 5 tops.
Something should pop/melt at that point if it's faulty / weak conductive path. If you do have a hard short-circuit on the primary of the PSU, though, the heating element will just run at full power (so beware how hot it can get.)

I usually do this with a 1200 Watt toaster oven, since it's all closed up and don't have to worry about heat from the heating element melting anything.

It really doesn't look like this PSU has those two balancing resistors that you speak of

That's poor design.
With two caps in series like this, balancing resistors of equal value should always be installed. Otherwise, one cap can take on a larger voltage than the other (due to differences in internal impedance and also capacitance), which can result in a cap failure over time.

Maybe I gave this PSU more credit initially than it deserved. If the manufacturer didn't get this right with the balancing resistors, who knows what else they got wrong / goofed up.

[socketa]

Re: Task TK-940TX-DF Blowing fuse

Could it be possible that that 1M ohm resistor, that's between the AC input rails, is faulty, and only shorts out when high voltage is applied?
Would it be OK to remove it and test again with the bulb?

Yeah i have a spare stove element that i can use.
How long should i "let it rip" for? Until the caps/rectifier explode or one of those voltages shows up on the meter?

It really doesn't look like this PSU has those two balancing resistors that you speak of - the only components that are connected to the fat trace where the primary caps connect together, are two MOVs. Apart from that there is no tee off.
You can clearly-enough see this in the last photo that i posted - (moving from right to left) there is the two holes where the two cap legs have gone through, there is a blob of solder, an empty hole that was originally empty, and then two holes for the MOVs. That's it. Not even any jumper! Can't even see two same value resistors on the photos of the board, that have a yellow multiplier.

[momaka]

Re: Task TK-940TX-DF Blowing fuse

I reinstalled R9, the 1.5Mohm resistor, that was across the -ve of one primary cap and the +ve of the other (AKA, the outside terminals).
Looking at other PSU schematics, it seems that the usual balancing resistors are something along the lines of two 330K resistors, one across each capacitor - but not looking like that in this PSU - so could it be that there is only one balancing resistor?

No, there should always be two balancing resistors of equal value. 300-500 KOhms will do (but again, both resistors have to have equal rated values.)

That alone can't cause the bulb to glow, though.

The two blue Y2 caps are not shorted (not even high resistance on the Mohms range) so is there any point in removing them?

I suppose not, because without the bridge rectifier, you already confirmed the bulb is not glowing.

The bulb still glows after bridging the fuse holder and connecting it between the red and brown live wires.
The rectifier tests as you say it should with a double diode reading one way and nothing the other way.
...
I checked the fuse holder that i installed, to make sure that i didn't do something silly - but that's on correctly, and not in contact with any other component
Also replaced the primary caps, with correct polarity orientation, with the other ones that i previously had in there; and still the bulb glows. (and managed to lift another trace while doing so)

Welp, I don't know what else there is to do at this point. I just can't explain what's eating up the power on that board. Maybe Nikola Tesla was right, and there is a way to transfer power wireless-ly.

On a more serious note, do you have something with a powerful heating element in your house? Perhaps a kettle or hair dryer or a loose heating element from a stove/oven? If yes, get the one with the highest power you can find (1 KilloWatts or more would do nicely), but also not something that is too high and would overload the breaker/plug/wires you're testing the PSU with. Then connect that heating element / device in place of the incandescent bulb, power it up, and let it rip. HINT HINT HINT: you may want to do this outside. Most likely you will end up blowing up the bridge rectifier and big caps if there is still a short-circuit on the PSU board. So if you have a set of crappy no-name primary caps from a junk PSU, you may want to put those in the PSU instead. Also put a bridge across the NTC thermistors if you'd like to save them (and to not limit current in any way to the heating element.) Extra credit: put multimeter to measure voltage across the primary caps while doing this experiment. If whatever is causing the short on the primary burns out before the bridge rectifier and the rest of the wires, then you will see a healthy ~150-170V DC across each cap (or 330-340V between [+] and [-] on the bridge rectifier.) The reason to use a big heating element is so that much more power can be passed through the primary side. So in the worst case, the heating element will fully power on, and the breaker in your house won't trip and you won't end up with melted plugs anywhere (except perhaps for melted traces on the PSU board.)

[socketa]

Re: Task TK-940TX-DF Blowing fuse

As you can see, in post #43, the bulb is wired across the fuse holder
Yes, the bulb is an old type incandescent filament type, that heats up
I reinstalled R9, the 1.5Mohm resistor, that was across the -ve of one primary cap and the +ve of the other (AKA, the outside terminals).
Looking at other PSU schematics, it seems that the usual balancing resistors are something along the lines of two 330K resistors, one across each capacitor - but not looking like that in this PSU - so could it be that there is only one balancing resistor?
The bulb still glows after bridging the fuse holder and connecting it between the red and brown live wires.
The rectifier tests as you say it should with a double diode reading one way and nothing the other way.
The two blue Y2 caps are not shorted (not even high resistance on the Mohms range) so is there any point in removing them?

I checked the fuse holder that i installed, to make sure that i didn't do something silly - but that's on correctly, and not in contact with any other component
Also replaced the primary caps, with correct polarity orientation, with the other ones that i previously had in there; and still the bulb glows. (and managed to lift another trace while doing so)

[momaka]

Re: Task TK-940TX-DF Blowing fuse

That is not possible.

How are you wiring the incandescent light bulb? Across where the fuse goes? Also, are you sure it's an incandescent light bulb and NOT LED or flourescent? Bridge rectifier still OK (showing two diode voltage drops when red multimeter probe is on [-] leg and black probe on [+] leg, with power disconnected from PSU of course.)

Looking at this picture...
https://www.badcaps.net/forum/attach...9&d=1595749921
... how about bridge the fuse and wire the bulb between where the brown wire and red wires attach instead (that is, separate connection between brown and red wire, then put incandescent bulb across there.)

All I can think of is that there is a mistake in the wiring of the bulb if bridge rectifier is OK. May also remove the two blue Y2 caps as well, but I doubt that will do the trick.

This is a single layer PCB and I don't see any solder bridges across traces, so it really isn't possible for the bulb to be glowing still.

Also, you said you removed everything from the primary side, but keep the high-resistance balancing resistors across the two electrolytic caps still in the circuit.

PCB doesn't look that bad at all, by the way. Still serviceable.

[socketa]

Re: Task TK-940TX-DF Blowing fuse

Have removed all of the primary DC components and thoroughly cleaned the backside of the board - but the bulb still glows (no surprise there)

[socketa]

Re: Task TK-940TX-DF Blowing fuse

Ok, my initial intuition was correct, in respect to you not mentioning those NTC thermistors again.
i didn't read the datasheet carefully; as i should have being looking under the zero power resistance column - thanks for the explanation - i get the gist of these now.
Will continue on, over the next day or so, as i'd like to find the culprit.

Can't be as bad as this, can it?

Thanks for the encouragement

[momaka]

Re: Task TK-940TX-DF Blowing fuse

Well i removed all of the components, two at a time, that were connected to the DC legs of the rectifier, and the bulb still glowed, so then i removed the green NTC thermistors that were noted earlier, and the bulb didn't glow.

The NTC thermistors are on the AC side, so when you remove them, AC no longer gets to the bridge rectifier... hence why the bulb stopped glowing.

Need to remove components on the DC side of the bridge rectifier to find out what causes this. If nothing, at least in the end you'll end up with a bag full of components for re-use.

The one that is cracked, (and a chunk of it fell off of it when removing it) measured 42 ohms and the other one 9 ohms - according to the datasheet it seems that they should be in the Mohms range, so they were shorting out the AC input - that's why i couldn't find the short, because i was testing further into the circuit.

The cracked one is reading high-resistance and is consistent with how one would fail. After all, these are just temperature-dependent resistors whose resistance goes down with an increase in temperature. So like other resistors, their failure mode is high-resistance or open-circuit.

The one that is measuring 9 Ohms might be OK, depending on the room temperature. SCK 054 should be about 5 Ohms at 25C ambient and higher resistance with colder temperature (or lower resistance with higher temperature.) Your multimeter may also add an Ohm or two...

So yeah, these aren't the issue of your PSU and they are not shorting out the primary - not unless you solder them accidentally in a spot meant for a Y2-class cap.

Maybe you meant "test" when you said "check"?

Sorry, yeah I use those two interchangeably sometimes.

The board is a mess now, with traces lifted in several spots, courtesy of my 'practising' on it, and have lost that big resistor, so there is no way that i would risk using it now.

Can't be as bad as this, can it?
https://www.badcaps.net/forum/showpo...&postcount=206

What sort of design is this PSU, and would you rate it on par with one of those ISO or Hyena power supplies, as opposed to a Liteon or Hipro?

I'd say this PSU is closer to the LiteOn and HiPro PSUs, since it uses a single-transistor forward design for the main PS, which is a little newer than the cheapo PSUs that use half-bridge (like the ISO and Hyena, IIRC.)

Is a 5VSB IC controller preferable to what this PSU has?

Yes, since it offers more protections and generally won't burn out when overloaded. 2-transistors 5VSB designs don't take it kind to bad caps or over-loading.

[socketa]

Re: Task TK-940TX-DF Blowing fuse

Well i removed all of the components, two at a time, that were connected to the DC legs of the rectifier, and the bulb still glowed, so then i removed the green NTC thermistors that were noted earlier, and the bulb didn't glow.
The one that is cracked, (and a chunk of it fell off of it when removing it) measured 42 ohms and the other one 9 ohms - according to the datasheet it seems that they should be in the Mohms range, so they were shorting out the AC input - that's why i couldn't find the short, because i was testing further into the circuit.
Maybe you meant "test" when you said "check"?
The board is a mess now, with traces lifted in several spots, courtesy of my 'practising' on it, and have lost that big resistor, so there is no way that i would risk using it now.
The initial reason why i wanted this PSU is no longer valid because i thought that i had to use a particular board with more than 2GB Ram for the internet; but the issue was with firefox (the current computer that i'm using has 3 GB of RAM and if FF pulls up to around 2GB it starts locking up the page with a circular throbber on top, and sometimes it just crashes - i'm going to get around to trying that version of Opera that you suggested)

What sort of design is this PSU, and would you rate it on par with one of those ISO or Hyena power supplies, as opposed to a Liteon or Hipro?
Is a 5VSB IC controller preferable to what this PSU has?

[socketa]

Re: Task TK-940TX-DF Blowing fuse

Ok, thanks for the explanation.
Have removed the transformer, and the bulb still glows.
Will continue on as you described, and will post back when it bulb ceases to glow.

[momaka]

Re: Task TK-940TX-DF Blowing fuse

There is two small BJT transistors (but ignore that, because i figured out that i have no idea what they are for)

Well, if they are on the primary, then I guess it doesn't hurt to remove them too, just in case (though I don't see how they can make the bulb glow when they are not connected in any way that will allow so much current to pass through them.)

Yes.
There is obviously a short, or low resistance happening somewhere, so why can't i find the it using the multimeter? - that would make this a lot easier.

Multimeter only measures DC resistance and based on a low-voltage test.
Parts can fail such that they only short-circuit or pass current with normal circuit voltage.
Active circuits will also only pass current with normal circuit voltage (i.e. diodes, transistors, etc.)
So that's why the issue is not showing itself.

and across the DC legs i initially get 30Mohms that falls to 9Mohms over the course of a few minutes and keeps falling. Does that sound normal? (shouldn't there initially be low resistance and then it be rising as the primarys slowly charge up?)

Capacitance on the primary caps is rather small. Likely they charge so quickly that the "low resistance" is possibly not captured by the multimeter. I find this to be the norm with slower multimeters and capacitors of 1000 uF or less.

Of course, to verify that something else is not interfering with this measurement, then it's best to repeat the same experiment with the caps out of the circuit.

What's the odds that the replacement primary caps are bad (or were destroyed after fitting them and turning it on), as well as the original ones?

Unlikely.
You'd need reverse voltage to kill them. You can take either the old or the new caps and put them back into another test PSU, then see if they work there. If they work and your cap meter is also reading the capacitance and ESR correctly, then the caps are fine.

I just removed one of the primary caps, and the bulb didn't glow.
Then put it back in, and then the bulb glows.
Swapped it for one of the original ones and the bulb glows
Swapped the other one with the original (so that both of the original ones were back in), and the bulb glows.
I measured the original caps with the ESR meter and got readings of 0.037 and 0.041 ohms

That further suggests there is some part of the circuit after the caps that is the issue.

If you have the tools, remove the main PS (big) transformer.
After this, test with bulb, and continue removing parts from the primary until it stops glowing and you're left with nothing but the primary caps, bridge rectifier, and balancing resistors between the two caps.

Do it one part at a time so you know which part made the bulb stop glowing. First start with any components that are directly connected to the + or - of the primary caps (again, with the exception of the bridge rectifier and balancing/bleeder resistors.)

[socketa]

Re: Task TK-940TX-DF Blowing fuse

Which two transistors? I thought all the ones on the primary side are gone.

There is two small BJT transistors (but ignore that, because i figured out that i have no idea what they are for)

Is your bridge rectifier installed correctly? - i.e. + and - polarity going properly to the big caps?

Yes.
There is obviously a short, or low resistance happening somewhere, so why can't i find the it using the multimeter? - that would make this a lot easier.
I get 1Mohm ohm across the AC legs of the installed rectifier (that 1Mohm resistor) which makes sense, and across the DC legs i initially get 30Mohms that falls to 9Mohms over the course of a few minutes and keeps falling. Does that sound normal? (shouldn't there initially be low resistance and then it be rising as the primarys slowly charge up?)
What's the odds that the replacement primary caps are bad (or were destroyed after fitting them and turning it on), as well as the original ones?
I just removed one of the primary caps, and the bulb didn't glow.
Then put it back in, and then the bulb glows.
Swapped it for one of the original ones and the bulb glows
Swapped the other one with the original (so that both of the original ones were back in), and the bulb glows.
I measured the original caps with the ESR meter and got readings of 0.037 and 0.041 ohms

[momaka]

Re: Task TK-940TX-DF Blowing fuse

Just in time - i was getting close to removing the two transistors

Which two transistors? I thought all the ones on the primary side are gone.

No - but i have just measured it now.
Across the +ve of one cap and the -ve of the other cap, as soon as the switch is flicked there 55-60V DC and then it quickly goes to 40V, and over the course of about a minute it drops in (roughly) exponential decay towards 21V (by then it's falling at a slow rate of 0.01V/s)
Just across the outer cap there is an initial voltage of 19V that instantly drops to 13V and then, in the same fashion as above, falls to 12.6V after about a minute
And ditto across the inner cap 30V > 22V > 10.4V

That's a very strange result.

It looks almost as if there is some active device somewhere pulling the voltage down to that level. Still, I assume this is with all primary transistors removed, so this shouldn't be happening.

Is your bridge rectifier installed correctly? - i.e. + and - polarity going properly to the big caps? With no primary transistors and no 5VSB transformer, there should be no current path to conduct across the two big caps. Keep removing components until you get somewhere. Clearly there must be something that we're missing that I can't seem to see in the pictures.

Once you get a result after removing/re-installing something, repeat once or twice more to confirm which component or part is causing the change in symptoms. In other words, we need repeatability. If you remove a component and the circuit stops (or starts) doing something new, then re-inserting same component should restore back old behavior. If not, check everything that you're doing to make sure you're not introducing an error somehow.

That doesn't happen to me with FF48.0.2 Windows7 Ultimate 32bit Intel Core2 Duo CPU E7500 @ 2.93GHz - i keep an eye on it with SysTrayMeter in the taskbar.

Well, that E7500 is quite powerful CPU (relatively speaking, compared to P4's and single core Athlon XP/64 chips), so indeed won't see it hog CPU resources (it would probably be down in the 10-20% CPU usage when it happens.)

[socketa]

Re: Task TK-940TX-DF Blowing fuse

Just in time - i was getting close to removing the two transistors

So did you try measuring the voltage across the two big caps when the PSU is plugged in and the bulb glowing?

No - but i have just measured it now.
Across the +ve of one cap and the -ve of the other cap, as soon as the switch is flicked there 55-60V DC and then it quickly goes to 40V, and over the course of about a minute it drops in (roughly) exponential decay towards 21V (by then it's falling at a slow rate of 0.01V/s)
Just across the outer cap there is an initial voltage of 19V that instantly drops to 13V and then, in the same fashion as above, falls to 12.6V after about a minute
And ditto across the inner cap 30V > 22V > 10.4V

Anything after FF24 is questionable to me.

Sounds good to me - i was reluctant to go to FF30, and then only used nightly

Thanks for the info on the 754.

all of the ESR and non-ESR versions of FF after 40 have been trashing my CPU usage in random intervals

That doesn't happen to me with FF48.0.2 Windows7 Ultimate 32bit Intel Core2 Duo CPU E7500 @ 2.93GHz - i keep an eye on it with SysTrayMeter in the taskbar.

Chrome? - wouldn't surprise me if that's what will be used when TPTB lure gullible folks and technophiles into being AI'ed