Power Circuit Board Mystery, Harman Kardon HKTS220 SUB/230

[Yoake10]

Here is a power circuit board mystery to solve!

For 3 months now I have been trying to repair and error search Harman Kardon 220 sub power circuit board without any luck at all. Therefore I need some more electric minds and brains to solve this problem. I will attach the schematic below with the replaced components and voltage measurements.

Anyone have any idea whats going on?
The problem started with R23 burning out and recitifer bridge BR1 broken. It should be 35 VDC at DN4 and 3,3V DC at CN2.

 

[Yoake10]

It seems you need at least a USB scope to observe charge / discharge effects to isolate if it is insufficient /or/ excessive on 3.3VSB/0.1A. The 24V on +15 indicates it is certainly trying more than it should to reach 3.3V and that DC readings thus far point in this direction. Is there excessive voltage drop on the diode or excessive load on the 3.3 somewhere by observing the shape/ slope of the voltage on the storage cap. .. failures might be a fractured solder joints but the AC waveform will give more clues.

Thanks for more information!

I will take a look at this and be back with more information to you all.

 

[Yoake10]

My friend has offered to lend me his oscilloscope and help me with the readings in June/July.

I will be back then, hope ur still here.

 

[csabex]

Hi. Any update for this thread? I have similar issue and my toughts are about the T1, because at the secondary side i measure very low value at one of the coils...

 

[Nigel Goodwin]

Hi. Any update for this thread? I have similar issue and my toughts are about the T1, because at the secondary side i measure very low value at one of the coils...

As the outputs of T1 are low voltage, they will read very low resistance, as there aren't many turns of fairly thick wire.

While transformers do VERY occasionally fail, it is VERY, VERY rare - and would be down to design flaws in those specific transformers. I used to keep two types of transformers in stock (both for satellite receivers), as those two types did occasionally fail - I certainly never replaced more than single figures of those though, and I was repairing a few thousand SMPSU's every year.

 

[csabex]

As the outputs of T1 are low voltage, they will read very low resistance, as there aren't many turns of fairly thick wire.

While transformers do VERY occasionally fail, it is VERY, VERY rare - and would be down to design flaws in those specific transformers. I used to keep two types of transformers in stock (both for satellite receivers), as those two types did occasionally fail - I certainly never replaced more than single figures of those though, and I was repairing a few thousand SMPSU's every year.

Thank you for your reply Nigel, and I wish you Happy New Year I haven't notice it, sorry, i must have been to busy with the new year's eve party
I read this post by you, and i kept in consideration, but basically the circle is so small already that it points to this T1 failure. I got the same readings, i have the 320 V DC on primary side and nothing else at the secondary side, and the system depends on these voltages, because the stdby voltage is also dependent of T1. I have no other clue for now...

 

[rjenkinsgb]

It's far more likely to be the high voltage IC that drives T1, or a capacitor or diode - the transformer itself is the very last thing to suspect.

The transformer does not pass DC, the system is totally dependant on the ICE3B0365J single chip controller to drive the transformer primary.

Those are cheap enough, I'd try that first and see what happens?

https://www.mouser.co.uk/ProductDetail/Infineon-Technologies/ICE3B0365J?qs=OWKQh20churyEQ%2FC8E%2FCHQ%3D%3D

 

[csabex]

It's far more likely to be the high voltage IC that drives T1, or a capacitor or diode - the transformer itself is the very last thing to suspect.

The transformer does not pass DC, the system is totally dependant on the ICE3B0365J single chip controller to drive the transformer primary.

Those are cheap enough, I'd try that first and see what happens?

https://www.mouser.co.uk/ProductDetail/Infineon-Technologies/ICE3B0365J?qs=OWKQh20churyEQ%2FC8E%2FCHQ%3D%3D

Thank you for the reply. Tomorrow i will replace especialy the C8 and also the rest of the capacitors, and then i will buy this chip controller and i will get back with a reply to share the results. Thank you again

 

[csabex]

It's far more likely to be the high voltage IC that drives T1, or a capacitor or diode - the transformer itself is the very last thing to suspect.

The transformer does not pass DC, the system is totally dependant on the ICE3B0365J single chip controller to drive the transformer primary.

Those are cheap enough, I'd try that first and see what happens?

https://www.mouser.co.uk/ProductDetail/Infineon-Technologies/ICE3B0365J?qs=OWKQh20churyEQ%2FC8E%2FCHQ%3D%3D

I replaced the IC and the C8 capacitor and also one diode of the SB160 because i forgot and have buyed only one piece Still nothing... No voltages at all at CN2. Now i will replace all of the capacitors, the C5, C6, C26, C27 and C31 let's see what happens.

 

[Nigel Goodwin]

Have you checked the rectifiers aren't short?, D1, D3, D4 and D9 - any of those short will prevent the standby PSU starting up.

I presume you are experienced in switch-mode PSU repair, and fully aware that the primary side is live to the mains, the secondary side isn't, and that they share a common ground. So any primary side measurements need to be taken from PGND - also extreme caution needs to be taken, because you can't connect an earthed scope to the primary side, unless you take special precautions.

 

[csabex]

Have you checked the rectifiers aren't short?, D1, D3, D4 and D9 - any of those short will prevent the standby PSU starting up.

I presume you are experienced in switch-mode PSU repair, and fully aware that the primary side is live to the mains, the secondary side isn't, and that they share a common ground. So any primary side measurements need to be taken from PGND - also extreme caution needs to be taken, because you can't connect an earthed scope to the primary side, unless you take special precautions.

Yes, i have chaked and even replaced one of them, all are ok. Just middle experienced, but i do aware of the live side, I take precautions when measuring there. Btw, i measured the primary side and it was 315 V, so it is ok the live side, the capacitors also work, because after every power cycle i discharge them. I'm a little bit confused, really don't know in which direction should I continue

I don't remember if I mentioned before, but I started the repair with the popular R23 and 2pcs of FET (MDF 11N60) issue, but now if I look closer I see a capacitor, the C30 which is right before R23, so maybe I will replace that too, but as I now it is very rare to be damaged a ceramic capacitor.

 

[csabex]

Yes, i have chaked and even replaced one of them, all are ok. Just middle experienced, but i do aware of the live side, I take precautions when measuring there. Btw, i measured the primary side and it was 315 V, so it is ok the live side, the capacitors also work, because after every power cycle i discharge them. I'm a little bit confused, really don't know in which direction should I continue

I don't remember if I mentioned before, but I started the repair with the popular R23 and 2pcs of FET (MDF 11N60) issue, but now if I look closer I see a capacitor, the C30 which is right before R23, so maybe I will replace that too, but as I now it is very rare to be damaged a ceramic capacitor.

I have found out that the R20 and D5 are bad and replaced them, and also the opto coupler, but still nothing...
This board is really a mystery, never ever had so much headache with anything...
Any ideas somebody ?

 

[Nigel Goodwin]

I have found out that the R20 and D5 are bad and replaced them, and also the opto coupler, but still nothing...

Those two are likely to die if the chip fails - opto-couplers 'can' fail, but it's a very rare occurrence.

I always used to make up 'PSU repair kits' for anything I repaired on a regular basis, and some kits were available commercially.

Here's an example off my website, from LONG ago (last edited 2006), of the kind of items that 'fail' in those specific units.

PRD PSU repair details

And a list of various examples.

Satellite Index Page

 

[Tony Stewart]

U1 is a critical driver surrounded by dead parts ... means it is at fault

https://www.digikey.com/htmldatasheets/production/70535/0/0/1/ice3b0365jg-ice3b0565jg.html

 

[Olli]

Hi! I just got my hands on the same device, faulty.
I have similar results, CN4: no voltage, CN2 Pin1: Fluttering 2.5-3v, Pin3: ~6V and dropping constantly. No voltage at Pin4 at any times.
I have replaced the following: R23, D4, D5, C5, C6, C8, C19, C26, C27, C30, C31.
Measured resistors straight from the board and all seems fine.

After i replaced these, i plugged the board in, and got 7v in CN2 1 & 3, i think CN4 was still 0V
Waited a while, plugged it in again with CN2 & CN4 connected -> Board died again.

Lost motivation, found this thread.

So is the next step replace U1? And how can i find what component did it burn again?
I dont have a ESR Meter on hand, but at school there is one if i want to measure something.
(Can i measure capacitor straight from the board with ESR Meter? The pads on pcb are really brittle.)

 

[csabex]

As the outputs of T1 are low voltage, they will read very low resistance, as there aren't many turns of fairly thick wire.

While transformers do VERY occasionally fail, it is VERY, VERY rare - and would be down to design flaws in those specific transformers. I used to keep two types of transformers in stock (both for satellite receivers), as those two types did occasionally fail - I certainly never replaced more than single figures of those though, and I was repairing a few thousand SMPSU's every year.

Hi! I just got my hands on the same device, faulty.
I have similar results, CN4: no voltage, CN2 Pin1: Fluttering 2.5-3v, Pin3: ~6V and dropping constantly. No voltage at Pin4 at any times.
I have replaced the following: R23, D4, D5, C5, C6, C8, C19, C26, C27, C30, C31.
Measured resistors straight from the board and all seems fine.

After i replaced these, i plugged the board in, and got 7v in CN2 1 & 3, i think CN4 was still 0V
Waited a while, plugged it in again with CN2 & CN4 connected -> Board died again.

Lost motivation, found this thread.

So is the next step replace U1? And how can i find what component did it burn again?
I dont have a ESR Meter on hand, but at school there is one if i want to measure something.
(Can i measure capacitor straight from the board with ESR Meter? The pads on pcb are really brittle.)

First, I need to inform you, Nigel, that I am the exception that proves the rule because, finally, the faulty part was the T1... I don't know why and how he died, but after I ordered a new one and replaced it, the unit worked like a charm again.

Anyway, thank you again for your support and for the help that you offered to me all the way from the beginning of my posts on this topic. I do have a concern that it will fail again because there are extremly high vibrations, and I think that's the reason for the faulty components, because at this level of vibration, also the solderings can fail and do a mail function.

Olli:

First, check the voltages across the board based on the Yoake10 drawing with the voltage values he measured, and then also check the parts and mark them on the drawing the same way as he did, and then you will get a picture of the whole drawing, and then you can rule out good areas of the board and see which area is related to a faulty part that you measured. One thing is sure: without U1 there is no life, so everything starts there if in your case T1 is ok.

I replaced the same parts as you and many more (U1 also), but the difference was that I couldn't measure any voltages on CN2 in my case. Also measure the secondary side outputs of the T1 if you have 15V or 3,3V.

Be aware of the primary side, because there is lethal voltage there, so measure only if you know what you are doing, because it is dangerous if you are unsure about what you are doing, so be very careful or ask for qualified help.

 

[Nigel Goodwin]

Well done, and particularly on been able to source a transformer

 

[csabex]

Well done, and particularly on been able to source a transformer

Thank you I share the link if somebody needs a transformer in the future. They sent it from the Netherlands, and it took 5 days in my case to arrive here in Hungary.

I think in the future somehow I will take the PS board out from the sub body in a separate box to prevent soldering, or any other damage caused by extremly high vibrations. I hope that there will be no problem if the power connectors and speaker wire get longer by 20 cm max...

 

[Olli]

Hi, i made similar picture that Yoake10, measured places. But im baffeled about the voltages, how can those be so high.

Anyone got explanation, or suggestions what to do next?

Replaced components: R23, D4, D5, C5, C6, C8, C19, C26, C27, C30, C31

 

[Tony Stewart]

My areas of concern the intelligent switcher U1 and the Programmable Vref IC1.

The IC1 must have 2.5V on pin 1 to signal gnd but for 3.3VSB (standby) R22 should be 56k and not 10k. (??) Otherwise it should try to go much higher output to get 2.50V and cause problems.

The soft-start depends on U1-7 being stable for soft start to work properly, so check C8 then U1-1.
When drawing current U1-3 should rise to 200 mV for CSense then U1-2 FB regulates around 1.35V if IC1-3 ever gets to 2.1V with 3.3VSB out .

If the transformer voltages are not ratiometric, for AC across each low voltage winding pair then there is a fault in the transformer. The meter must be AC coupled for this. This can happen from a fault of lightning exceeding the protection design.

 

[Nigel Goodwin]

Hi, i made similar picture that Yoake10, measured places. But im baffeled about the voltages, how can those be so high.

Anyone got explanation, or suggestions what to do next?

Which voltages? - assuming the one you marked as AC?, you shouldn't be trying to measure AC, so that's your problem.

As the previous poster had a faulty transformer, then there's a good chance that's your problem as well - as I've mentioned previously, transformers rarely fail - BUT, if they do, it's usually down to a poorly designed or poorly constructed transformer, so it's likely to be a common issue.

Tony Stewarts suggestion that it could be down to lightning is extremely unlikely, any lightning event big enough to damage a transformer would leave massive signs that it had been struck by lightning. Tracks vaporised, components blown off the board or disintegrated - and even then, I've never seen a transformer damaged (and I've seen a LOT of lightning damaged domestic electronics).

And don't believe that lightning never strikes twice in the same place, we had the exact same three houses (in the middle of an estate) struck almost exactly 12 months apart.