Electrical Engineering Stack Exchange is a question and answer site for electronics and electrical engineering professionals, students, and enthusiasts. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I am trying to repair the power supply in an LCD monitor. It is a fairly basic design around an OB2268AP in the 20-30 Watt range. The power supply had failed spectacularly because the main MOSFET short-circuited, vaporizing one pin on the regulator IC, burning 2 resistors to a crisp, damaging another plus some collateral damage.

Here's the part of the circuit after filtering and rectifying mains voltage, so there's some 300V DC between U+ and U-.

Switching supply part

There are a few oddities on the PCB:

  • R706 is not a resistor but a choke inductor (makes sense)
  • ZD702 is not mounted
  • R708 is not a resistor but a zener diode. I can make out '24' at the end of its designation, so it's probably a 24 Volt zener

R710 and R712 were burned to a crisp so I can't make out the original values, and I need some advice on the values. The reference design for the OB2268 does not mention R710 but I suspect it's a low ohm resistor to have some 'protection' against the gate capacity of Q701. I guess something like 2.2Ω, 4.7Ω maybe? Any higher and the rise and fall times for the gate will suffer, I guess.

The one that has me stumped is R712. Pin 6 on the IC is the SENSE input of the current limiter. It has a threshold of 0.86 volt; together with R711 of 3.3Ω that makes a limit of 0.25 Ampère. If R708 is indeed a zener diode of 24 volt it would act as a secondary limit for the IC's own power circuit (D703, 'R'706, etc). So what's your guess for R712? Maybe the value is non-critical (the input resistance of pin 6 is 40 kΩ according to the datasheet), maybe it can't be too high else zener R708 won't work reliably.

Update: R711 is actually 0.33Ω

Update 2: I repaired it with the following components:

Q701: IRFB9N60A (600 V, 9.2 Amp mosfet)
R701: 2.2 ohm
R712: 1 kohm
I702: an optocoupler I had lying around :P

I hooked up an oscilloscoop to the gate of Q701 and the rising edge is a bit curved and there's a tiny bit of oscillation/overshoot but otherwise it seems okay; the descending edge is straight and sharp.

Note on the IRFB9N60A: in contrast to the original 7N80C this transistor not an isolated package.

share|improve this question
5  
+1 for the excellent reverse-engineering hand-drawn schematic. It brings back memories. – transistor Feb 27 at 16:56
up vote 5 down vote accepted

A few ohms for R710 seems about right. The gate drive is a push-pull: OB2268 Gate Drive

Even though the datasheet shows a relatively slow turn on and turn off time, there could still be a bit of gate oscillation without a resistor here. I would suggest (as you note) something in the order of 2.7 to 10 ohms as a starter; there is indeed a trade-off between gate slew and gate ringing.

R712 is a series resistor into the current sense input (current limiting is set to engage at 260mA according to the datasheet). I think R712 is there to provide a helper filter so the leading edge blanking can operate properly; it is not that uncommon for leading edge blanking to get 'confused', depending on application specifics. I would assume that the first pass of the design had some anomalies around this area (there is an internal chopper circuit).

Sense input

It is difficult to assess the specifics of this resistor, but something around 33 ohms might be a good starting point, although I have not done a full analysis, so treat this recommendation with caution; it is where I would start for a leading edge blank filter.

I agree with 'R703' probably being a 24V device (the controller is rated at 36V).

Excellent job with the schematic.

[Update]

The part has a fixed leading edge blanking time, derived from the internal oscillator apparently, as the resistor used to set it is a parameter on the datasheet line:

Blanking time from datasheet.

A fixed blanking time could conceivably be an issue depending on the specifics of the design , so it quite natural to see a resistor here that can form a small filter (because the blanking is too short in a given design) in conjunction with track and pin capacitance (and possibly internals that we have no knowledge of).

From that perspective, it is quite feasible that the filter resistor is several hundred ohms to even a few k ohms, as noted by Nick.

share|improve this answer
    
Thanks for the detailed answer. My gut feeling tells me though that 33 ohm for R712 may be bit low, though Alexeev's value of 1k in his answer is at the other end of the spectrum. I will need to do a little experimenting I guess and see which value(s) work most stable. I will report the findings in my question. – JvO Feb 27 at 22:40

I concur with @Peter about purposes of R710 and R712,
and would like to add my $0.02 thought.

I think that the initial guess value for R712 should be higher, on the order of 1kΩ.
This thought comes from a flyback converter which I have designed previously. It also had a current-mode controller (different model of controller, though).

enter image description here

share|improve this answer
    
Looks good, though this seems to be part of a low-pass filter to eliminate spikes at MOSFET switch-on. The datasheet of the OB2268 mentions that it has leading edge blanking that makes this kind of filter unnecessary. – JvO Feb 27 at 22:43

My guess at R712 was 100K (just like R707). This would give a gain of 1. Then I saw the data sheet information presented by Peter Smith, and noticed that the RI is 100k. Could this just be a coincidence?

share|improve this answer
    
I'm afraid you have mixed up a few things. RI is the resistor connected to pin 4, which regulates the internal oscillator frequency. It has nothing to do with the SENSE input. What Peter Smith spotted is that for the frequency associated with RI=100k, the leading edge blank time on the SENSE input is 400 ns. It's a obfuscated bit of information (apparantly they don't want to reveal details about the internal frequency, since at 100k, f=65kHz which has a period time far longer than 400 ns) – JvO Mar 4 at 21:20

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.