Badcaps.net Forum
Go Back   Badcaps Forums > Troubleshooting Hardware & Devices and Electronics Theory > Troubleshooting Power Supplies and Power Supply Design
Register FAQ Calendar Search Today's Posts Mark Forums Read

 
Thread Tools Display Modes
Old 09-21-2020, 01:19 PM   #81
momaka
master hoarder
 
momaka's Avatar
 
Join Date: May 2008
City & State: VA (NoVA)
My Country: U.S.A.
Line Voltage: 120 VAC, 60 Hz
I'm a: Hobbyist Tech
Posts: 9,639
Default Re: Task TK-940TX-DF Blowing fuse

Buzzing for single-transistor forward PSUs typically means over-load condition.

However, since you have a series light bulb, it could also be a false over-load condition, as some PSUs simply don't like high-impedance on their input.

So I think there are three possibilities for what you are seeing here:

1) There is a short-circuit or over-current condition on the main PS secondary side, causing the primary to over-load. (Could be a faulty rectifier on one of the rails, for example, or an open feedback path.)

2) There is a component fault on the primary side of the main PS, causing the PWM IC to over-drive main PS MOSFET until its protection kicks in (open opto-coupler, open opto-coupler feedback path, faulty snubber component, or sense resistor going high resistance.)

3) PSU's main PS design might be sensitive to high-impedance on the input, which means the series light bulb could be the reason why you're seeing this behavior. (On that note, Active PFC PSUs... or PSUs with APFC... are even more sensitive, and generally cannot be tested with a series bulb less than 200-300 Watts.) If that's the case, then using several 60-100 Watt incandescent bulbs in parallel... or one 200-300 Watt heating element... should provide sufficiently lower impedance on the output to make the PSU function.

I suggest you start by assuming case #1 above and check all secondary-side components. Start with the large power-handling components first, like large rectifiers and diodes for the 3.3V, 5V, 12V, and -12V rails. To do that, check resistance between each of these rails and ground. Note that there _may_ be minimum-load resistors on the output, which could possibly make you think that a rail is shorted to ground (common mistake when someone uses continuity test on their multimeter to check if a rail is shorted to ground, since most multimeters will beep at resistance under 100 Ohms, and load resistors on the 3.3V and 5V rails typically are 100 Ohms or less.) An alternative (and somewhat easier) method to checking the rails for short-circuit is to use a 20-40 Watt, 12V, incandescent/halogen bulb and backfeed an external voltage to each rail. If you have a spare ATX PSU, simply wire the output of that PSU to the rail you want to test. For example, if you want to test if the 12V rail is shorted, connected ground on both PSUs together, then connect the 12V rail of one PSU to the 12V rail of the other PSU _through_ the indancesence/halogen bulb (i.e. similar to how you use the series light bulb on the input.) Then, plug in only the working PSU and turn it On (the PSU under test should remain unplugged from wall.) If the bulb glows steadily, there is likely a short-circuit on that rail. If not, then rail is probably OK. Just make sure to always connect the right rails together - i.e. only use 12V rail to test 12V rail on the PSU in question, or 5V rail to test the 5V rail, and 3.3V rail to test the 3.3V rail. If you accidentally mix then up (particularly inserting a higher voltage on a lower-voltage rated rail), you could blow up some caps.

If you find that none of the rails are shorted with the above tests, then we could assume case #3 for a bit, and simply add a few more incandescent bulbs in parallel to your single bulb on the input, or simply replace that with a 200-300 Watt heating element. Then try powering up the PSU and see if it buzzes again. If it does, go back to assuming case #1, and check all secondary-side components again. After this, assume case #2 and check all primary side components related to the main PS.

Quote:
Originally Posted by socketa View Post
As far as i can hear, the light buzzing is coming from the small 472 Mohm blue cap (red line), and after i turn it off and on until the loud buzzing happens, it is coming from the adjacent side of the big white resistor (green line), possibly the other small 221 ohm small blue cap or the larger one.
The numbers on those caps indicates their capacitance and not their resistance.

caps with numbers xx5 = x.x uF
caps with numbers xx4 = 0.xx uF
caps with numbers xx3 = 0.0xx uF = xx nF
caps with numbers xx2 = 0.00xx uF = x.x nF
caps with numbers xx1 = 0.000xx uF = 0.xx nF

* where xx = 2-digit number
(Also, nF = nanoFarads, uF = microFarads)

So 472 = 4.7 nF and 221 = 0.22 nF

Last edited by momaka; 09-21-2020 at 01:27 PM..
momaka is offline   Reply With Quote
Old 09-22-2020, 01:36 PM   #82
socketa
Badcaps Veteran
 
socketa's Avatar
 
Join Date: Jun 2014
City & State: hell
My Country: samsara
Line Voltage: 230v
I'm a: Knowledge Seeker
Posts: 391
Default Re: Task TK-940TX-DF Blowing fuse

----------
Last night
----------

No shorts on the output rails. (3.3V 15ohm, 5V 43ohm, 12V 233 ohm, and no shorts on the blue, white, or grey wires)
So added a 100W bulb in parallel with the current 60W bulb.
Turned it on, the fan spins and keeps spinning, there is buzzing for a half second and then the bulbs light up and the buzzing increases in volume, but the voltages are now good.
Swapped the 60W with a 75W, same result as before except that the bulb takes a second to light up.
The the voltages are good from the moment the PSU is turned on, to the moment that i turn it off.

-----------
This morning
-----------

Added another bulb holder (so now have all three bulbs in parallel) and the buzzing was still there.
Wiggled a primary cap and noticed a spark, and noticed that i hadn't soldered it in so great.

But, foolishly, instead of just soldering it up good , i removed them and put the originals back in.
The buzzing didn't happen, the bulb briefly lights up, the fan spins for a second then stops, and i hear a brief tiny squeak sound from the PSU after turning it off for about 10 seconds.
No squeak sound when turning it on.

If i just use the 100W bulb alone, the bulb stays lit, i hear a much louder squeak sound when turning it on, the fan twitches, and i hear the same soft squeak (an un-squeak?) sound after about 10 seconds, as before, after turning it off.

I get the same result now, regardless of which of the three pairs of caps i use.

So i don't know if replacing the caps caused another fault.
I suspect that it fixed the buzzing, but the buzzing (which i'm assuming was a loose primary cap leg that was arcing with the surrounding solder) may have created a condition whereby the PSU could power up with good voltages. So maybe there was two problems and i fixed one, which subsequently brought an underlying one to the surface.
Or maybe the impedance of the parallel bulbs is still too high for the PSU to start?

Quote:
The numbers on those caps indicates their capacitance and not their resistance.
Doh! I got tricked by the "472M" that was printed on the cap, and bypassed my logic gate.
Thanks for the additional info.

P.S. When i put the components back on, i replaced two shorted zeners with ones that looked exactly the same, from the primary of another PSU

Last edited by socketa; 09-22-2020 at 02:07 PM..
socketa is offline   Reply With Quote
Old 09-24-2020, 11:44 AM   #83
momaka
master hoarder
 
momaka's Avatar
 
Join Date: May 2008
City & State: VA (NoVA)
My Country: U.S.A.
Line Voltage: 120 VAC, 60 Hz
I'm a: Hobbyist Tech
Posts: 9,639
Default Re: Task TK-940TX-DF Blowing fuse

Quote:
Originally Posted by socketa View Post
Last night
----------

No shorts on the output rails. (3.3V 15ohm, 5V 43ohm, 12V 233 ohm, and no shorts on the blue, white, or grey wires)
So added a 100W bulb in parallel with the current 60W bulb.
Turned it on, the fan spins and keeps spinning, there is buzzing for a half second and then the bulbs light up and the buzzing increases in volume, but the voltages are now good.
Swapped the 60W with a 75W, same result as before except that the bulb takes a second to light up.
The the voltages are good from the moment the PSU is turned on, to the moment that i turn it off.
That's an interesting result.

It suggests the PSU might just be OK, but just not liking the still relatively high impedance of the bulbs on the input. However, I haven't seen a PSU that doesn't work with 2-3 bulbs in parallel, so there may still be something iffy going on - in particular something pulling too much current on the secondary side that is making the primary overload. The fact that the output voltages are stable and have correct values means the secondary side feedback is OK. Given the results, though, I suggest to check the secondary side snubber circuits for each rail. These snubber circuits would be a low-value resistor (typically 2.2 to 10 Ohms and about 1/2 Watt rating) placed in series with either a ceramic or small metal film capacitor (generally in the range of 0.01 to 0.1 uF... or 103 to 104 printed on the cap.) You'll find these snubber circuits for each of the major rails (5V, 12V, and maybe 3.3V) close to either the heatsink-mounted rectifiers or close to the secondary side output of the main transformer. Typical failure mode for these snubber circuits is either the resistor will overheat and go open-circuit or the ceramic/film cap will go short-circuit. Since these snubber circuits are connected directly across the transformer output pins, you won't be able to check the components in circuit - you will need to remove either the cap or the resistor, or lift one leg on these two components from the PCB to take proper measurement.

Quote:
Originally Posted by socketa View Post
This morning
-----------

Added another bulb holder (so now have all three bulbs in parallel) and the buzzing was still there.
Wiggled a primary cap and noticed a spark, and noticed that i hadn't soldered it in so great.

But, foolishly, instead of just soldering it up good , i removed them and put the originals back in.
The buzzing didn't happen, the bulb briefly lights up, the fan spins for a second then stops, and i hear a brief tiny squeak sound from the PSU after turning it off for about 10 seconds.
No squeak sound when turning it on.
Is the squeak sound heard when simply disconnecting the PS_ON wire from ground (to turn Off the main PS) or when you unplug the PSU from the wall?

The latter (unplugging PSU from wall outlet) is actually normal for 2-transistor 5VSB circuits. I have a lot of PSUs with a 2-transistor 5VSB circuit, and at least half of them do a fairly loud chirp/squak (typically ending with a raise in pitch as time passes) as the PSU is unplugged from the wall.

Quote:
Originally Posted by socketa View Post
If i just use the 100W bulb alone, the bulb stays lit, i hear a much louder squeak sound when turning it on, the fan twitches, and i hear the same soft squeak (an un-squeak?) sound after about 10 seconds, as before, after turning it off.

I get the same result now, regardless of which of the three pairs of caps i use.
Well at least we have consistency now... and the bulb is no longer glowing right away when the PSU is plugged in.

Quote:
Originally Posted by socketa View Post
P.S. When i put the components back on, i replaced two shorted zeners with ones that looked exactly the same, from the primary of another PSU
That is certainly something worth investigating. Zener diodes can have different Zener voltage ratings, despite looking all the same way. Need to know which Zener diodes this is in the circuit so hopefully we can try to guess what voltage it should be rated for - that is, if you don't have the original shorted Zener diodes, which will tell you that by reading the parts numbers written on them.
momaka is offline   Reply With Quote
Old 09-25-2020, 12:15 AM   #84
socketa
Badcaps Veteran
 
socketa's Avatar
 
Join Date: Jun 2014
City & State: hell
My Country: samsara
Line Voltage: 230v
I'm a: Knowledge Seeker
Posts: 391
Default Re: Task TK-940TX-DF Blowing fuse

Quote:
Is the squeak sound heard when simply disconnecting the PS_ON wire from ground (to turn Off the main PS) or when you unplug the PSU from the wall?

The latter (unplugging PSU from wall outlet) is actually normal for 2-transistor 5VSB circuits. I have a lot of PSUs with a 2-transistor 5VSB circuit, and at least half of them do a fairly loud chirp/squak (typically ending with a raise in pitch as time passes) as the PSU is unplugged from the wall.
Yes, that is indeed the the case here - switching off at the wall

I added another bulb (60W), so that's four bulbs in total - no noticeable change from when there was three.

The PSU will not power up, when turning on at the wall, until after the un-squeek squeek is heard (which happens when the primary caps have discharged to approx 30V)
Monitoring the rail voltages, i notice that 12V rail gets to 12.2V, 5V rail gets to 5.08V and the 3.3V rail gets to only 2.7V, before the PSU shut's itself down - so maybe there is a clue, or pointer, there that suggests looking at the 3.3V snubber circuit.

I'll investigate those snubber circuits to see if the resistors and ceramic/film caps are free of the faults that you mentioned, and then post results here when done.

--------

Zener Diodes

Looking back through the posts, the two zeners that i replaced (i eventually threw the originals away) are:

ZD8, next to the heatshrinked resistor,
Quote:
the shorted Zener is probably part of the problem. But since it appears connected to the Gate drive of the main PS MOSFETs
and ZD1
Quote:
I removed zener diode, ZD1, and it tests as a shorted; and the short across R23A dropped right off.

Last edited by socketa; 09-25-2020 at 12:38 AM..
socketa is offline   Reply With Quote
Old 09-26-2020, 03:34 AM   #85
socketa
Badcaps Veteran
 
socketa's Avatar
 
Join Date: Jun 2014
City & State: hell
My Country: samsara
Line Voltage: 230v
I'm a: Knowledge Seeker
Posts: 391
Default Re: Task TK-940TX-DF Blowing fuse

I found the 3 snubber circuits, all 4.7 Ohm 1/2 watt sized resistors. And there are no shorted caps or open resistors

Quote:
Since these snubber circuits are connected directly across the transformer output pins, you won't be able to check the components in circuit
I saw no need to lift any component, because, according to the following logic:
Each circuit is essentially a resistor and capacitor connected together in a closed loop.
If you place the meter leads across the resistor (which is the same as "placing them across the capacitor") and get a measurement that matches the resistor, then that means that the resistor is good, and that the capacitor is not shorted. If the capacitor was shorted then the meter would read as a short because one could rightly say that it's parallel with the resistor.

For quite some time i've being inclined question what i'm told, because from a young age i refused to eat animal meat, but, at that time, someone told me that sausages were mostly bread crumbs, so i became somewhat addicted to them, and sausage rolls, for many years; not really understanding, or considering, what i was doing - until one day not long after i started eating mince pies, i noticed a piece of artery inside one. That was the end of that 'venture', and also the beginning of being vegetarian. I still didn't know about why cows make milk (incredibly many folk don't!) ... could go on with a long story, but suffice it to say that i'm constantly amazed at how easy it can be to miss stuff, that in hindsight, is blatantly obvious after the application of minimal thought, quietude, or/and assistance

Attached is some photos of that resistors (which all tested good, at 4.7 ohms) and caps.
The first pic looks like "102" is printed on the cap, and the other two caps have "472" printed on them.
The colours on the three resistors are yellow, violet, gold; although due to lack of sunshine, the yellow looks green in the photos.
Also attached, is a photo of the back of the board, with the three snubber circuits drawn in yellow.
The snubber at the bottom has a jumper on the right hand side (vertical yellow line) which i initially missed, and so i was looking for the path back to the transformer (hence the irrelevant diode that i drew in)






Attached Images
File Type: jpg 2020-09-26_05-43-08-p.m..jpg (33.6 KB, 10 views)
File Type: jpg 2020-09-26_05-44-57-p.m..jpg (36.6 KB, 10 views)
File Type: jpg 2020-09-26_05-45-29-p.m..jpg (29.0 KB, 3 views)
File Type: jpg 2020-09-26_05-45-59-p.m..jpg (36.6 KB, 10 views)
File Type: jpg 2020-09-26_05-46-49-p.m..jpg (41.5 KB, 10 views)
File Type: jpg 20200926_165412.jpg (397.7 KB, 11 views)
File Type: jpg Snubbers.jpg (375.2 KB, 10 views)

Last edited by socketa; 09-26-2020 at 04:15 AM..
socketa is offline   Reply With Quote
Old Yesterday, 11:35 AM   #86
momaka
master hoarder
 
momaka's Avatar
 
Join Date: May 2008
City & State: VA (NoVA)
My Country: U.S.A.
Line Voltage: 120 VAC, 60 Hz
I'm a: Hobbyist Tech
Posts: 9,639
Default Re: Task TK-940TX-DF Blowing fuse

Quote:
Originally Posted by socketa View Post
Yes, that is indeed the the case here - switching off at the wall

I added another bulb (60W), so that's four bulbs in total - no noticeable change from when there was three.

The PSU will not power up, when turning on at the wall, until after the un-squeek squeek is heard (which happens when the primary caps have discharged to approx 30V)
Monitoring the rail voltages, i notice that 12V rail gets to 12.2V, 5V rail gets to 5.08V and the 3.3V rail gets to only 2.7V, before the PSU shut's itself down - so maybe there is a clue, or pointer, there that suggests looking at the 3.3V snubber circuit.
The PSU not able to start until you hear the "un-squeak squeak" is probably because a short-circuit or over-power protection is latching.

Yes, the 12V and 5V rails coming up but the 3.3V rail not regulating properly is definitely a clue here.

Since this PSU derives the 3.3V rail through a mag-amp circuit, it relies on the primary side's switching action to regulate the 3.3V rail through a transistor and a saturation coil on the secondary side. If something is not quite right with the PWM switching, the 3.3V rail can get affected by that and not regulate properly. Or it could also be that something on the 3.3V rail's mag-amp circuit is bad, causing it to not regulate properly. If you trace the entire power-side of the 3.3V rail (i.e. only big traces that carry power and not small feedback/signal traces), you should find a transistor, a diode or two, and also a 431 shunt regulator with some resistors all connected to the 3.3V rail - all of these parts are for the mag-amp circuit and should be checked out.

You should also try feeding back 3.3V from another PSU into this PSU's 3.3V rail and see if anything gives or overheats (particularly the transistor responsible for regulating the 3.3V rail.)

Quote:
Originally Posted by socketa View Post
I found the 3 snubber circuits, all 4.7 Ohm 1/2 watt sized resistors. And there are no shorted caps or open resistors
...
I saw no need to lift any component, because, according to the following logic:
Each circuit is essentially a resistor and capacitor connected together in a closed loop.
If you place the meter leads across the resistor (which is the same as "placing them across the capacitor") and get a measurement that matches the resistor, then that means that the resistor is good, and that the capacitor is not shorted. If the capacitor was shorted then the meter would read as a short because one could rightly say that it's parallel with the resistor.
That is true.

So looks like your snubber circuits are OK.

Quote:
Originally Posted by socketa View Post
Zener Diodes

Looking back through the posts, the two zeners that i replaced (i eventually threw the originals away) are:

ZD8, next to the heatshrinked resistor, and ZD1
Ok, looking back at the circuit once again...

ZD8 is related to the 5VSB circuit only: it appears to be used as a limiter for the voltage on the Gate of the 5VSB MOSFET. Since the 5VSB seems to be working normally, I suppose everything is good there.

However, that does raise the question: when you try to power on the main PS and you hear the PSU squeak, what happens to the 5VSB? In other words, try measuring the voltage on the 5VSB when you do this experiment. We want to see if the 5VSB drops out or not. If it is, that would be a problem. To get more accurate result, first put a load on the 5VSB - something in the range of 0.25 to 0.5 Amps should do (so connect something like a 10-20 Ohm resistor to the output of the 5VSB - that 1.5 KW heating element should have about 6.5 Ohms resistance, and probably even that will do the trick.) Then plug in the PSU, but keep main PS turned OFF, then see if the 5VSB still outputs 5V. If it does, try powering on the main PS and see if the 5VSB continued to supply 5V. The reason we want to do this is because the 5VSB circuit also has a primary side auxiliary winding that feeds the PWM controller on the primary side for the main PS. If 5VSB is unstable or dropping out, the main PS will not work either.

As for ZD1.....
That one is for Gate voltage protection for the main PS MOSFETs.

With that said, both ZD1 and ZD8 should have Zener voltage ratings that are less than the maximum rated Gate voltage of the 5VSB and main PS MOSFETs.
i.e.
ZD1 Zener voltage < V_GSS of main PS MOSFETs
and
ZD8 Zener voltage < V_GSS of 5VSB MOSFET

However, the Zener voltage of these Zener diodes should also not be too low, because then the MOSFET's may not get fully driven On.

This means the Zener voltage of these Zener diodes should be close to the rated V_GSS of the MOSFETs, but just slightly lower.

So if you used an improper Zener diode for ZD1, that could cause the MOSFET to not get fully driven On and thus not switch properly, leading to OPP/SC protection kicking in.

Last edited by momaka; Yesterday at 11:37 AM..
momaka is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump



Badcaps.net Technical Forums 2003 - 2020
Powered by vBulletin ®
Copyright ©2000 - 2020, Jelsoft Enterprises Ltd.
All times are GMT -6. The time now is 01:50 PM.
Did you find this forum helpful?