SMPS repair

[IO390]

I have a CT-258C PSU out of an Optoma HD600X projector. It's roughly 6 years old.

I've never repaired an SMPS before and usually I'm happy to dive in and learn for myself, but given that there's mains voltage on this and the possibility of stuff blowing up (well more than usual) I thought I'd ask for some advice.

I've read various troubleshooting guides but I just want some second opinions if that's OK.

First of all, there is nothing on the board that's visibly damaged. No signs of anything overheating and nothing has blown up. I have tested all the FETs and diodes on the board and they all appear to be OK. I've read that it's best not to use an oscilloscope on PSUs like this, I have an isolation transformer but I still don't want to damage the scope (or probes) so I haven't tried that route yet. I've been doing what I can with my multimeter (HP34401A).

The PSU has an ICE2PCS02 PFC controller right after the rectifier, and there is the 335VDC present afterwards so that appears to be working. There is an SG5841 PWM controller and an LNK364 (335VDC was measured at the D pin). I do suspect that the problem is with one of the PWM controllers, but how could I check for this? The parts aren't expensive so I could just order all 3 of the controllers and try them, but I'd like to be more precise if possible. Does anyone have any tips on how to find out what's broken, and if it is possible to use the scope? I've not played with SMPSs at all before, I understand how they work but this one is rather more complex than the others I've seen so I'm a bit stuck.

Thanks in advance

[retiredcaps]

Well, you didn't say what is wrong with the projector, but if SG5841 is responsible for the standby voltage and you are missing standby voltage, then you can check to see if 16V is present on pin 7 (VDD).  Also check the voltage across the capacitor that feeds pin 7.  It should be 16V steady and stable.  Often these caps can dry up (high ESR) and have no visible bulges.  Pin 1 on the SG5841 is your GND.

I use a normal hand held meter to take readings, but am careful when working on the hot side. I use clips if I can for no hands measurement.

I can't answer any scope questions because I don't have one.

In addition, a clear focused picture of the board would help us help you.

[IO390]

Thanks for the reply. There is nothing coming out the board whatsoever so hence why the projector won't work. I'll get some pictures tomorrow. Cheers

[voltz]

Hi, just some general thoughts, sorry nothing specific to your PSU or problem.

Switch mode supplies are notoriously difficult to repair. Some obstacles include:

Dangerous high side voltages connected directly to the mains without isolation. Earth on scope probe will go bang if connected!
Feedback by design meaning its hard to get a handle to isolate the problem, one fault can show as another.
They can shut down in a short circuit load situation misleading us into thinking the PSU is at fault when its not.
'Turn on' signals not reaching the supply (like a PC power supply for example with its power up button).

Its way safer as already stated to use a multimeter to measure voltages and if you are going to use a scope beware! You would need to use an isolation transformer connected to the PSU (not to the scope mains, thats dangerous as the chassis can go live once connected to the smps). Or,  using either a differential scope probe (expensive) or two x 100 probes and putting the scope in 'Maths' differential input mode if you have it. That means only the inner points of the probes are used and at no point are the grounds of the probes connected to the supply, keeping the scope isolated.

All in all, its a pain! and personally, i tend to back out of repairing switchers. Other than for checking obvious faults like the DC rectified voltage is present (330v) across the mains caps. Then, test diodes and switching devices for shorts.. After that it gets tricky. Its not uncommon for the transformer to break down too. And probably near impossible to replace. Many are a custom winding.

I hope some of this helps, and good luck with it. But take a look at replacing the whole supply, would that be cost effective?.

Best.

[tautech]

Other than checking the IC supply resistors are not OC and the IC bulk cap values are OK the next course of action is to externally power the IC's and check they are providing the correct outputs. Non mains powered at the same time of course.
If there is a FB winding and supply diode check that too.

[station240]

First thing I'd do is replace any low voltage (not 400V) capacitors on the hot side. These are so called start up caps, but just filter voltages for switching devices.

Also possible a fault on the output rails (bad capacitor, shorted diode) is shutting down the PSU.
Again, suspect capacitors, especially as this is from a projector, and those run hot.

[SeanB]

Also check any high value resistors, as they are often used, along with the small capacitor, to provide a start up supply for the SMPS controller. No voltage because a resistor is high in value is common, as often they use a few SMD resistors in series, or a single 0.3W unit which fails as it is being run at too high a voltage and breaks down.

[IO390]

Thanks for the help everyone, it's greatly appreciated.

I've made a little bit of progress, not been playing with electronics much over the past few months, it's amazing how quickly I forget things.

There were some small caps on the hot side which I replaced. There's some pictures of the board here: http://www.aliexpress.com/store/product/Projector-Accessories-CT-258C-mains-power-supply-for-LG-BS275-BX275/1040500_2011114602.html

Third picture down shows 2 transformers, I'm assuming the smaller one is for standby power. In the same picture there is a small 2 pin connector on the top left which appears to be connected to the supply of the SG5841. The connector goes to a microswitch which is only closed when the lamp is installed - with the projector dismantled and the switch disconnected this is obviously open. For testing purposes I have substituted the microswitch with an M3 nut.

I've measured every resistor on the board and they appear to be in tolerance. My multimeter has a problem with the ohms range though so I'll need to double check it all when I have access to another meter.

When powered from mains, the supply pin of the ICE2PCS02 PFC controller is receiving only 1.2V so there is definitely a problem there. The supply circuit consists of some SOT-23 transistors and 0603 resistors and I have lost my magnifying glass, when I find it I can draw the circuit.

[Towger]

I spent ages messing with an infocus power supply, only to discover that one of the leads from the supply which disappeared into the guts of the projector was actually connected to a blown thermal fuse.

[xavier60]

The LNK364 should be for the standby supply, so has to run first. Other control ICs are sometimes powered from rectified waveform from a winding on the standby transformer.

[rob77]

just few generic hints from my experience:

always check also the diodes on secondary side... i repaired many SMPS which had a shorted diode on the output and the smps was not starting up because of the overload protection. you can easily detect this kind of fault by a scoping the current sensing feedback (battery powered scope or use isolation transformer).

also a dried input filter cap can cause many issue - i had a case when the PFC was pumping into a dried out cap and creating high voltage spikes which killed the standby switcher - rendering the whole SMPS unusable.

if you missing the standby voltage (e.g. Vsb on ATX supply) and there is no visible damage/burn marks and the rectified mains is present, then in most cases there is a open fusible resistor which is powering the standby supply. ( so replace both the standby switcher and the fusible resistor )

[rob77]

335V after PFC doesn't seem right to me. It should be considerably higher than rectified mains voltage. A normal voltage is 385V.
This voltage is usually set to higher than maximum possible rectified voltage, so for a universal input device, it should be higher than 265V*1.414=375V.

i agree 335V is on the low end... 360-400V is the usual range with a 450V rated capacitor.
but i also met several of them with PFC and approx. the mentioned 335V , using 350V rated capacitor.

[_Wim_]

always check also the diodes on secondary side... i repaired many SMPS which had a shorted diode on the output and the smps was not starting up because of the overload protection. you can easily detect this kind of fault by a scoping the current sensing feedback (battery powered scope or use isolation transformer).

Same here, I repaired (only) 2 SMPSes so far, but both failures were caused by a failing diode (not shorted however when measured with a multimeter)...

Be carefull with an isolation transformer, when touching 2 different parts of the SMPS, you (or your equipment) still can get into big trouble. I used an Tektronix A6902B isolator, these pop-up on Ebay every now and then for aprox 100€ (in the US there a plenty available). By using this, you can isolate 2 channels from your scope and probe around safely...

[Jane]

Speaking of an isolation transformer and measuring with an oscilloscope on SMPS.
Does it mean that I must  use SMPS AND oscilloscope as well with  an isolation transformer or ONLY SMPS with isolation transformer but  not oscilloscope when measuring on SMPS?

[rob77]

Speaking of an isolation transformer and measuring with an oscilloscope on SMPS.
Does it mean that I must  use SMPS AND oscilloscope as well with  an isolation transformer or ONLY SMPS with isolation transformer but  not oscilloscope when measuring on SMPS?

the point is to have one of the two floating either the scope or the smps but i would prefer to float the smps and have the scope grounded properly. but the safest and recommended way is to use a high voltage differential probe i'm  personally using a battery powered handheld scope (hantek dso1102b) to work on SMPS repairs so i'm not using an isolation transformer.

[Jane]

@rob77: What high voltage differential probe  do you use?

[_Wim_]

Speaking of an isolation transformer and measuring with an oscilloscope on SMPS.
Does it mean that I must  use SMPS AND oscilloscope as well with  an isolation transformer or ONLY SMPS with isolation transformer but  not oscilloscope when measuring on SMPS?

To be honest, when you need to ask this question, it better (safer) to use high voltage differential probes in combination with an isolation transformer.
An isolation transformer by itself can help for certain measurements, but it does not make measurements safe. Connecting both scope AND smps to the transformer does not enable you to perform measurements on the primary side (floating side) of the smps. When connecting the smps only to the transformer, you CAN measure on the primary side of the SMPS, BUT when you connect the ground lead of your scope, this will reference that part of the SMPS to ground, which means (all) other parts of the SMPS (parts that you would not always expect) will be floating on 230Vac voltage (for example the DC output of a non-isolated SMPS). This make this approach very dangerous to get chocked, but it does allow you to make the measurements (so you really need to know what you are doing)

In my opinion, an isolation transformer is only useful in COMBINATION with other safety measures, for example connecting the SMPS to the isolation transformer, and then measuring with isolated probes on the scope. The isolated probes on the scope ensure the SMPS does not get referenced to ground somewhere, and the isolation transformer ensures that you not get chocked by touching a single point (you can get chocked by toughing 2 points, so always probe around "with one hand in your pocket"...) Same approach can be used when probing around with a multimeter. Floating the scope I would not recommend, because when you connect the scope to the SMPS, parts of the scope can become at life potential, chocking you when you touch the scope.
So please be very careful, and read up on the subject before approaching a project like this without all the necessary tools.

[Jane]

@ _Wim_:Can you please let me know what kind of the isolated probes to use?Which would you recommend?

[tautech]

@ _Wim_:Can you please let me know what kind of the isolated probes to use?Which would you recommend?

Pintek is a trusted brand, their DP25 differential probe is satisfactory for hobbyist use.
http://www.pintek.com.tw/

[_Wim_]

@ _Wim_:Can you please let me know what kind of the isolated probes to use?Which would you recommend?

I use an Tektronix A6902B which can be found quite cheaply on Ebay. This is not a differential probe, only a 2 channel isolator... The DUT is powered by a medical grade isolation transformer.

[Bondguy]

Does it mean I can not make a good measurement on primary side without a differential probe( when using an isolation transformer)?

[rob77]

Does it mean I can not make a good measurement on primary side without a differential probe( when using an isolation transformer)?

why not ? i'm not using a diff probe , but you just must be well aware what are you doing, and of course you MUST NOT exceed the scope's input specs.  if you  don't pay enough attention or you don't know what you're doing => you can easily get shocked or damage the DUT and/or scope. so the best way is to use a high voltage differential probe.

and one more note, usually you don't poke around with a scope in a SMPS... you just take a look at few point.. mainly the current sensing feedback and gate drive and for those you don't need a high voltage diff probe, you just need a battery powered (floating) scope or floating DUT (isolation transformer). but for example scoping the drain of a mosfet on the primary side of the SMPS you have to have a high voltage diff probe - those voltages exceed the scopes's input rating.

but in general ,a high voltage differential probe is the best option for ANY work on a SMPS. but don't forget that a isolation transformer + hv diff probe doesn't make you immortal ! you still can get shocked/killed by touching the DUT's guts