SMPS Controller IC and Diodes

[Madcat]

Hey folks, sorry for my lapse of attendance. I've been off computers because of my dead PSU on my main machine. Too much drama at home, and too many projects to build. This PSU has been on the backburner since mid-December, so yeah.....I need my PC gaming fix.

Last few days I've been pouring over the datasheet for the SMPS controller IC my PSU uses. Its an Infineon ICE2A0565Z. My last thread was about suitable replacement caps for my Corsair HX620, which is really a SeaSonic M12II 620. When I was re-capping my PSU, I had the iron at 400C, which I learned was up to 50C higher than SMD components should take. I blew up two zener diodes and a diode on the underside of the PSU. I traced the lines for days making absolutely sure what those diodes did. Good thing I traced because one of the zeners is a shunt regulator for the IC's VCC. What I dont get is the VCC comes from one of the two AC input pins of the main rectifier. From AC, its through three series resistors with a total of 99K Ohms and then through a forward-biased diode. After that, is the zener and diode I destroyed. VCC also comes from the 5VSB transformer. On the primary, the chamfered corner pin is connected to a 10 ohm resistor and then through a forward-biased diode to VCC. So Im not quite sure why theres two connections from the mains to VCC. Also, I dont get why theres 10 ohm resistors connected to the 5VSB transformer then to VCC. Arent both transformer primaries connected to mains?

The other zener is across the phototransistor side of the one of the 3 optocouplers, and looks like its used as a shunt. Its emitter is grounded and the collector is the output going to the Feedback pin of the controller IC. The zener is reverse-biased across. Someone told me it might be a backup in case the phototransistor went bad. Im beginning to think its there to limit voltage to below the maximum rating of the IC. Im just not sure, but Im scavenging some used zeners from tv boards with low knee voltages to test.

If you need more photos or info let me know. My biggest issue is not being sure what diodes to use. Im not crazy about capping the PSU, then trying to run it without components. The VCC zener is ZD103, by its side is D102, and the zener across the opto is ZD101. If need be Ill try to take photos with light under for traces. Any advice is greatly appreciated. Im stuck at the moment.

Top
Bottom
Closeup
Datasheet

[Th3_uN1Qu3]

Re: SMPS Controller IC and Diodes

Also, I dont get why theres 10 ohm resistors connected to the 5VSB transformer then to VCC. Arent both transformer primaries connected to mains?

Look at the datasheet. The 10 ohm resistor is there to provide current sensing for the controller. The zener for the controller is likely to be 15 volts.

The other zener is across the phototransistor side of the one of the 3 optocouplers, and looks like its used as a shunt. Its emitter is grounded and the collector is the output going to the Feedback pin of the controller IC. The zener is reverse-biased across. Someone told me it might be a backup in case the phototransistor went bad.

That is likely correct, but it also might be limiting the voltage to that phototransistor. What does reverse biased mean? If the zener's cathode (denoted by black band) is on the transistor collector, then it does what i said (limiting the voltage). The FB pin seems to have an internal pull-up to 6.5v, so that zener would likely be 5v1.

Don't take my word on it tho. After replacing the zeners run the PSU with a 40W lightbulb in series, so if anything goes wrong the bulb lights up and prevents damage. If you get correct 5vsb and the bulb is dark, the PSU is okay.

[Madcat]

Re: SMPS Controller IC and Diodes

Look at the datasheet. The 10 ohm resistor is there to provide current sensing for the controller. The zener for the controller is likely to be 15 volts.

Alright cool, I was guessing all along a good zener voltage for the VCC zener. According to datasheet 13.5V is the VCC turn-on threshold, so 15 volts sure as hell wouldnt hurt. It didnt hit me until you replied, that 10ohm resistor might be the R(start-up) resistor. The current sensing resistor is actually on pin 3, a grey through-hole. Now for the zener 5V1, is that 5.1V or part of the part number?

[joshnz]

Re: SMPS Controller IC and Diodes

5v1 stands for 5.1volts
That apply's to almost all zeners.

[Madcat]

Re: SMPS Controller IC and Diodes

Alright cool. Thanks for the confirmation. Ordering some 15V and 5.1V zemers. For the regular diode, I'll try a 1N4148. Also, never thought of the light bulb in series, I'll try that.

[Longbow]

Re: SMPS Controller IC and Diodes

Good thing I traced because one of the zeners is a shunt regulator for the IC's VCC.

So, what is the resistance from the VCC terminal of the IC to ground? Is the IC shorted?

What I dont get is the VCC comes from one of the two AC input pins of the main rectifier. From AC, its through three series resistors with a total of 99K Ohms and then through a forward-biased diode. After that, is the zener and diode

A common source of startup voltage for the IC. Once the supply starts switching, power to the IC is supplied by the small winding on the transformer as shown on the block diagram. Typical run voltage for this type of IC would be 12-16 volts dc. Your startup will most likely only be a pulse for turn-on, which you would see with a scope. Check those 3 100K SMD resistors because they will open up once in a while.

The other zener is across the phototransistor side of the one of the 3 optocouplers, and looks like its used as a shunt. Its emitter is grounded and the collector is the output going to the Feedback pin of the controller IC. The zener is reverse-biased across. Someone told me it might be a backup in case the phototransistor went bad.

There is no need to mess around with these negative feedback parts unless you actually find evidence of something shorted. (FYI, zener diodes normally operate in their "reverse biased" mode called avalanch mode.) The part in the feedback circuit shown on page 9 is a 3-terminal precision voltage reference, found in almost all SMPS circuits. They usually have a 2.3 or 2.5 volt output. This is not your issue. If you have a dead supply, you need to get it up and running.

Since the IC itself contains the switching device, I would replace the IC first along with any external parts you have identified as shorted. One big clue that the switcher is shorted would be an open Rsense resistor, around .1 ohm or so. Measure resistance from Isense to Drain pins - a quick check of the switching FET.

Before powering up, check the output filter caps with an ESR meter. Next be certain that nothing on the output side shows a short to ground. That should get you to the next step. Let us know.

[Madcat]

Re: SMPS Controller IC and Diodes

Hey Longbow. Ill do this in bullets, make it quicker.

• PSU was fine before I blew up three diodes to excessive soldering heat. I still measured the IC and VCC to GND is 2-3 megaohms.
• Cool, even more info on the startup behavior.
• I checked those resistors when I started examining the board. They're 33K each and in sereies.
• Not messing around at all, only trying to find suitable replacement diodes.
• Everything messed up is on the primary side.
• Im guessing you didnt check the images. Theres no caps on the PSU at the moment.
• When I solder back in the diodes and caps, I do plan to test outputs for any short though.

[Longbow]

Re: SMPS Controller IC and Diodes

When things start blowing up on the primary side, you will want to check the switching device first, in this case it is in the IC itself. Using the Isense and Drain pins you can spot a shorted FET easily. Once you tour the primary a little to determine obvious shorted or open parts, have a good look at the secondary, especially for shorted protection zeners, high ESR filter caps and so forth. A note on SMPS which include several output windings: the result of a fault in one secondary winding will have unpredictable results for the whole supply. One example that comes to mind is a multi-voltage SMPS for a VCR. One secondary output provides (typically) about 3 volts dc for the display tube filaments. Even though that winding is not grounded, a faulty filter cap in that one winding can shut down the whole supply, or it can make voltages in other secondaries higher or lower than normal. That behavior is typical for many high voltage supplies also such as TV flyback circuits with more than one secondary winding.