Benq G2255 doesn't power on - amber power light when plugged in

[rddube]

Re: Benq G2255 doesn't power on - amber power light when plugged in

Hi Will,

Since you said "C831, C833 & C832 form a capacitive voltage divider for the chip to monitor voltage on cn801. The values are critical", is set out to try and measure C835 & 836 / C833 & C832 with my Fluke multimeter, but wasn't getting any reading.

So I took my transistor tester with which has a capacitance function (and an ESR in circuit function, but I didn't use this one, I used the capacitance function only), and with my probes checked them. Here is what I got and not sure if this little machine is accurate, but it was constant:

C835 / C836 : 42.8 uf ESR .15K Vloss 14%
C833 / C832 : 143.2 uf ESR .15K Vloss 12%

I think these are in pairs, because I would get the same reading say from C835 and C836, the same for the other 2.

Not sure if this can have a meaning of some sort?

[will62]

Re: Benq G2255 doesn't power on - amber power light when plugged in

I'm pretty sure that if any out of those sets of 3 were out of spec, the chip would shut down due to voltage readings. When we had pin 2 lifted/jumped, the lights stayed on, so I would say that the voltage readings were within spec.
I'll look back at the circuit to see if I didn't notice something.

You can try that current test mention earlier, but please don't change any caps yet. Can you give me all the HV caps and values? Also, possible replacements on other boards.

I'll get back on in the evening after I've had a chance to check over circuit.

[rddube]

Re: Benq G2255 doesn't power on - amber power light when plugged in

ok, here are the HV cap values:

near the cfl connectors:
C834: 8 3KV SEC
C831: 10J 3KV
C817: 8 3KV SEC

near the AC input:
C603: SRC 102M SE 250V 21 C
C602: same as C602

near the yellow transformer:
C724: 404
C615: same as C603 and C602

In my spares, I have a brand new 10J 3KV that I ordered from Digikey with the mosfet and electrolytic caps that I changed. I have some 5C 3KV (not new), 102M KX250 X1Y1, 472M KX250 X1Y1 that are on another scrap power board from an HP monitor.

All the other HV caps I have are from ATX power supplies. Do you want those values too?

[will62]

Re: Benq G2255 doesn't power on - amber power light when plugged in

I went back over the circuit. The chip is shutting down because of the LI1(pin2). The current monitoring circuit is pretty basic, and you have checked all of the resistors and the diode, so no obvious component problem causing a bad current measurement. As I mentioned before, HV caps C831 and C834 are used in capacitive voltage divider. If they were bad, I think the chip would have shut down when we had LI1 lifted.

I'd like to rule out C817 (8pf), but you don't have an extra 8pf, so we'll have to make one. Take 2 of those 5pf 3kV caps off the other board. We'll use your new 10pf and 2 of the 5pf to get to 8.3pf. You need to solder one of the 10pf legs to a 5pf leg. These will be in series. Then solder the other 5pf to the outside legs of the 10/5 series, this will be parallel. Confused yet? I'll add a sketch. Try this Frankenstein cap in 817. If it doesn't work, move it to 834. If that doesn't work, put the original 8pf's back in place, pull the 10pf off of Frankenstein and try it in 831. That's the only way I can think of to check all 3 HV caps.

Before doing the above, go ahead and run a quick test for me. Jump 5 to ground. Power up and look at lights. After they've ran for a few seconds, un-clip the jumper. Do the lights go off immediately?

[rddube]

Re: Benq G2255 doesn't power on - amber power light when plugged in

Will you are a science rocket man!! Wow, I can't believe we are going to make a Frankenstein cap!!

Ok, ran the test as you suggested....left the lights on for approximately 15 seconds than pulled the jumper and the lights went out about 2 seconds later, just like when we put them on without the jumper.

Ok, so I'm off creating the monster.

BTW, who drew the schematic? Is it you? Pretty nice, are you a circuit designer in your work?

Tks Will...I'll report in a few.

[rddube]

Re: Benq G2255 doesn't power on - amber power light when plugged in

Sorry Will as I was building Frankie, power outage and no estimate from Hydro when it will be back

[rddube]

Re: Benq G2255 doesn't power on - amber power light when plugged in

Power came back at 4:00am this morning. Sorry for the late reply.

Ok, built Frankie and measured it, it gave .30nf, whereas C817 measured .32nf. Tried Fankie in C817, lights go out after 2 seconds. Switched Frankie to C834, same result. Recuperated the new 10pf from Frankie, tried it in C831....unfortunately, same result. C834 original measures .31nf and C831 original measures .30nf (all these readings done with my Fluke).

[rddube]

Re: Benq G2255 doesn't power on - amber power light when plugged in

I guess next step is to measure current at 837 and 841, one at a time one end lifted? Should I use mA or uA on my Fluke?

[will62]

Re: Benq G2255 doesn't power on - amber power light when plugged in

The diode is rated in mA, so try that first. Yes, lift one resistor at a time. The measurement you got on the 8pf is over 300pf. Does the new 10pf measure higher? Not sure if you can accurately measure these HV caps with a DMM.

The waves on your scope were all over the place. Wondering if the pic below showing LI1(p2) with the larger negative swing was a fluke, or if you can replicate that consistently. The way the diode is set up, it looks like it should be clipping some of the negative portion of the wave. Maybe also probe at the voltage divider (between R837,R839 and then between R841,R843 for LI2). Both LI1 and LI2 similar there?

Another thing that I was wondering about, is when you first changed the mosfet to the under rated AF4910N that you had on hand, you said the lamps came on for about 10 seconds. Now that you know how to properly check a mosfet, are the original AP9962 and the AF4910N actually blown? Is there something about the new mosfet that is causing spikes or something on the primary side? Maybe probe each drain, then probe the primary side of the transformer (probe to one side, probe ground to other). See 3rd pic below for a spike that looked small, but was blowing mosfets on a DC motor controller that I made (disregard notes).

As to the VI monitoring diagram above, no, I'm not a CD (we'd wrapped this 8 pages ago). I'm learning just like yourself. You can open a board pic in paint.net, add a layer on top, and trace circuit. You delete the pic (background), and you're left with the tracing.

[will62]

Re: Benq G2255 doesn't power on - amber power light when plugged in

I'm going to add another thing to check, as it may be later in the evening before I can get back on. There's a lot covered in these 2 posts, so just take your time and do everything. No need to post back until all has been checked.

When you had pin 2 lifted, and pin 3's input jumped to it, the lamps came on. I want to make sure that I didn't jump to conclusions that it was pin 2's input causing the problem. It could be the difference between pin 2 & 3's inputs when they hit the internal pull up current source, "OR" gate and comparator.

Lift pin 3, jump pin 2's input to it. Do the lamps come on? If they do, swap lamp connectors.

[rddube]

Re: Benq G2255 doesn't power on - amber power light when plugged in

Another thing that I was wondering about, is when you first changed the mosfet to the under rated AF4910N that you had on hand, you said the lamps came on for about 10 seconds. Now that you know how to properly check a mosfet, are the original AP9962 and the AF4910N actually blown?

Hi Will,

Took you up on this comment and I checked the AP9962AGM (I had kept it) and guess what, you're quite right. It was not blown at all, just me not measuring it properly off the board. So, I removed the other one and placed this one back and tested - exact same results...CLF's light up and go off after 2 seconds. If I install my alligator clip on C803, they light up and stay lit.

So, I'm off to do the other tests...should I start with the probing of the pins, maybe try to get a better scope image (yours is a nice square wave and you can see the spike very well)...or should I try the resistor up and measure milliamps?

[will62]

Re: Benq G2255 doesn't power on - amber power light when plugged in

That's great about the original mosfet, one less variable to consider. Still wondering why the lamps came on with the under rated mosfet. Is it blown? If you haven't started already, I'd jump ahead to the lifting of pin 3. That might tell us that there is not an over-current problem.

[rddube]

Re: Benq G2255 doesn't power on - amber power light when plugged in

Ok Will, did the test. When CFL's plugged in original position, they turn on and stay on. When reversed, they turn on and turn off after about 2 seconds.

[will62]

Re: Benq G2255 doesn't power on - amber power light when plugged in

The fact that the lamps came on with only pin 2's input going into pin 2 & 3, tells us that there is nothing wrong with the pin 2-cn801 circuit by itself. It's the difference between the 2 inputs that is causing the shut down. I didn't expect the lamps to go off when the plugs are inverted though. That points towards a bad lamp. They worked both ways when you lifted pin 2, but with pin 3 lifted, they only work one way (normal connection). I'm going to have to think about that.

Can you go back to post #169 and answer that 2nd paragraph about the first pic. Was the negative swing on pin 2 a fluke? Please check that out again, and try the others in that paragraph.

[budm]

Re: Benq G2255 doesn't power on - amber power light when plugged in

Ok Will, did the test. When CFL's plugged in original position, they turn on and stay on. When reversed, they turn on and turn off after about 2 seconds.

That made no logical sense to me.
If you have just one lamp connected one at time, does get the connected lamp to flash on for seconds and go black?
I use this to verify the power supply/inverter board: https://www.badcaps.net/forum/showthread.php?t=19987

[rddube]

Re: Benq G2255 doesn't power on - amber power light when plugged in

Hello Gentlement,

Sorry for my late reply, had an emergency last night which kept me away from my bench. Will resume testing tonight.

Will: Ok will be probing again pin 2 and pin 3 to repeat results. I presume I put pin 3 back in it's place and remove the bridge with pin 2?

Budm: The test with one lamp connected, do you want me to do it while the chip is bridged (i.e. pin 3 is lifted right now, connected to pin 2) or do I put everything back in it's place before testing the one lamp at a time? And tks for the link on verifying power supply/inverter board.

[rddube]

Re: Benq G2255 doesn't power on - amber power light when plugged in

Hey Budm,

I have a few of these lights, do you think I could use them without having to open them?

They are sold at home depot for 4.97$

Tks!

[will62]

Re: Benq G2255 doesn't power on - amber power light when plugged in

While you have the jumper between p2&3, please confirm what I said in post #174.

Pin 3 lifted - Both lamps come on and stay on ONLY when lamp connectors are in normal position. Invert connectors, and both lamps come on, but go off in a couple of seconds.

Pin 2 lifted - Both lamps come on and stay on, regardless of how the connectors are attached.

Need to nail this down before moving on. Then you can put the pins down, remove jumper and recheck the waves. Don't forget to also check them by probing at voltage divider (before diode... reread post#169)

[rddube]

Re: Benq G2255 doesn't power on - amber power light when plugged in

Ok Will, I'll do that before anything else. So you want me to redo the test with pin 2 lifted also again (pin 3 back in its place), sort of a double check just in case I might have missed something in the first test?

Tks.

[will62]

Re: Benq G2255 doesn't power on - amber power light when plugged in

Yes, redo pin 2 as well. We need to be sure on this. Once you finish and resolder both down, please measure resistance to ground for both.