Acer AL2423W: Recapped, Still Dead!

[PlainBill]

Re: Acer AL2423W: Recapped, Still Dead!

PB:

Okay, I checked everything and before I throw a hundred numbers at you, diode ZD6 measures .8 ohms.

ZD6 is the second diode from the top next to FET Q3 that you circled in yellow. It connects to the top leg of the FET.

Let me know if you still want the results from the IC's. They both gave nearly identical results.

Can I get excited yet?

Yep, get excited. A shorted diode very well could be responsible for the problems. If it is holding the FET off, the power supply won't work. Of course, the next problem is I don't have the foggiest idea what the rating of that zener is. I just hope someone can step in with an idea. I'm out of my area of experience at this point.

I'd say there is no point in comparing resistances of the two ICs; they are most likely not the problem. It WOULD be a good idea to remove the diode and verify it is shorted. If it isn't obvious, DON'T try to see if the monitor will work without the proper diode in place.

PlainBill

[CleanSC]

Re: Acer AL2423W: Recapped, Still Dead!

Hopefully this will help. I have attached a shot of ZD6 to see if anyone can chime in on this one.

Also, I located another seemingly identical diode elsewhere on the board (ZD7) and it came back with .76 volts.

Anything else I can help with, let me know!

[oneitchyeye]

Re: Acer AL2423W: Recapped, Still Dead!

Did everyone give up on this thread? I have been reading along. I too have a dead monitor. No power at all (not even the power light comes on). There were a couple of blown caps on the seconday side of the power supply so I replaced them all on that side. Still nothing. I plugged in and tested power output on CNS1 and I get the 5V but I don't get anything on the 12V. I'm hoping it has to do with the power switch somehow. I looked at the other caps and they look fine. I replaced some on the video board that were looking shabby but not blown. I don't see anything else that is suspect. No browning or anything that jumps out.

Keep the ideas coming and thanks!

[CleanSC]

Re: Acer AL2423W: Recapped, Still Dead!

I hope we haven't given up.

I'm trying to find out what rating that diode has to replace it and it's apparently a lot harder than it seems. From what I understand, the markings are not standardized for diodes like they are with resistors.

Does anyone know anyone that would know this info that could tell from the pic? Or maybe somewhere I can take my pics for ID on that diode?

[PlainBill]

Re: Acer AL2423W: Recapped, Still Dead!

PB:

Okay, I checked everything and before I throw a hundred numbers at you, diode ZD6 measures .8 ohms.

ZD6 is the second diode from the top next to FET Q3 that you circled in yellow. It connects to the top leg of the FET.

Let me know if you still want the results from the IC's. They both gave nearly identical results.

Can I get excited yet?

You're right, there are no standard markings on SMD devices, but that is only a problem, not a show-stopper. We have to reverse engineer this circuit, which shouldn't be difficult.

1. Look up the application note for the SMPS controller.
2. Look up the application note for the power switch (Q3?).

Between those two, it should be possible to determine the voltage and power rating of the zener.

I've been trying to figure out just how the zener is connected, that is going to have to wait until my eyes improve.

One more point. Lecture time. A zener diode is a conventional diode with some special characteristics. The reverse breakdown voltage is closely controlled. To explain, a 1N4001 is a conventional diode rated at 1 amp, 50 volts. It has approximately a 1 volt drop in the forward direction at 1 amp of current. It is guaranteed to withstand 50 volts in the reverse direction with only a few microamps of current. If you increase the reverse voltage, at some point it will start conducting more heavily and exhibit a very non-linear curve (a small increase in the reverse voltage will result in a large increase in the reverse current). This is the zener voltage. If you measure the voltage drop across a zener diode and get .76 volts, you are measuring the forward voltage drop, which isn't important.

PlainBill

[Aruba]

Re: Acer AL2423W: Recapped, Still Dead!

You're right, there are no standard markings on SMD devices, but that is only a problem, not a show-stopper. We have to reverse engineer this circuit, which shouldn't be difficult.

1. Look up the application note for the SMPS controller.
2. Look up the application note for the power switch (Q3?).

Between those two, it should be possible to determine the voltage and power rating of the zener.

I've been trying to figure out just how the zener is connected, that is going to have to wait until my eyes improve.

One more point. Lecture time. A zener diode is a conventional diode with some special characteristics. The reverse breakdown voltage is closely controlled. To explain, a 1N4001 is a conventional diode rated at 1 amp, 50 volts. It has approximately a 1 volt drop in the forward direction at 1 amp of current. It is guaranteed to withstand 50 volts in the reverse direction with only a few microamps of current. If you increase the reverse voltage, at some point it will start conducting more heavily and exhibit a very non-linear curve (a small increase in the reverse voltage will result in a large increase in the reverse current). This is the zener voltage. If you measure the voltage drop across a zener diode and get .76 volts, you are measuring the forward voltage drop, which isn't important.

PlainBill

If this zener diode is between the gate and source pins it is typically used for
transient suppression and/or ESD protection. Typically though I see they are
zener diodes that is connected back to back such to suppress positive and
negative transients. For this application they may have only considered
protecting positive transients for cost reasons I'm sure. For this power FET it
needs at least 4V to enhance it so I would expect the drive voltage to be
between 4-10V. Then using a 12V or 13V zener at 0.5W or more would
suffice. On some FETs you will find this back to back zener build into the FET
package. See an example of this with the attached FET spec.

Since it is only their for the occasional transient you may be able to unsolder
it and turn on the PSU to test out if your only problem is with this zener
diode. Once you know that this is the problem then a replacement zener diode
is highly recommended.

[PlainBill]

Re: Acer AL2423W: Recapped, Still Dead!

If this zener diode is between the gate and source pins it is typically used for
transient suppression and/or ESD protection. Typically though I see they are
zener diodes that is connected back to back such to suppress positive and
negative transients. For this application they may have only considered
protecting positive transients for cost reasons I'm sure. For this power FET it
needs at least 4V to enhance it so I would expect the drive voltage to be
between 4-10V. Then using a 12V or 13V zener at 0.5W or more would
suffice. On some FETs you will find this back to back zener build into the FET
package. See an example of this with the attached FET spec.

Since it is only their for the occasional transient you may be able to unsolder
it and turn on the PSU to test out if your only problem is with this zener
diode. Once you know that this is the problem then a replacement zener diode
is highly recommended.

Thanks for the information. Obviously, we are well out of my comfort zone here. Wouldn't having the datasheet for the transistor give more information?

PlainBill

[Aruba]

Re: Acer AL2423W: Recapped, Still Dead!

Thanks for the information. Obviously, we are well out of my comfort zone here. Wouldn't having the datasheet for the transistor give more information?

PlainBill

These zener diodes on the gate of the FET is used for protection. They are
typically "optional". I have seen schematics where these diodes are marked as
"NC" meaning no connection. The board is layed out with place setting(s) for
these diodes but they can opt out of not having to put them on based on
accelerated life testing, simulated lightning, overload and repeated on/off
cycle testing. But if they had failures during these test they can add these
diodes in without having to relayout the pcb again which is a costly and time
consuming proposition.
I have dowloaded the spec for both the controller and FET but neither of
them mentions about having to add these diodes for protection. As these
are "optional" and also depends on how much money you want to take out
of the parts BOM (bill of materials) they may or may not end up on the board.
One could select a FET that has these protection diodes build into it as the
like of what I have send previously but that may delay the time and cost you
more money before we can get to see this unit up and running. Of the schematics I have seen which do have these diodes in them they are rated
between 12-13V, 0.5W. You got choices to make now:
- remove it and see if it comes up
- go get a 12V zener and replace the shorted one
- go get a replacement FET with integral zener diodes with similar ratings as
04N70BF.

[CleanSC]

Re: Acer AL2423W: Recapped, Still Dead!

Thanks for the feeback, Aruba!

I will desolder the diode and report back with results, unless of course you recommend another course of action.

[PlainBill]

Re: Acer AL2423W: Recapped, Still Dead!

These zener diodes on the gate of the FET is used for protection. They are
typically "optional". I have seen schematics where these diodes are marked as
"NC" meaning no connection. The board is layed out with place setting(s) for
these diodes but they can opt out of not having to put them on based on
accelerated life testing, simulated lightning, overload and repeated on/off
cycle testing. But if they had failures during these test they can add these
diodes in without having to relayout the pcb again which is a costly and time
consuming proposition.
I have dowloaded the spec for both the controller and FET but neither of
them mentions about having to add these diodes for protection. As these
are "optional" and also depends on how much money you want to take out
of the parts BOM (bill of materials) they may or may not end up on the board.
One could select a FET that has these protection diodes build into it as the
like of what I have send previously but that may delay the time and cost you
more money before we can get to see this unit up and running. Of the schematics I have seen which do have these diodes in them they are rated
between 12-13V, 0.5W. You got choices to make now:
- remove it and see if it comes up
- go get a 12V zener and replace the shorted one
- go get a replacement FET with integral zener diodes with similar ratings as
04N70BF.

Thanks for the explanation. That will be CleanSC's decision.

PlainBill

[PlainBill]

Re: Acer AL2423W: Recapped, Still Dead!

Did everyone give up on this thread? I have been reading along. I too have a dead monitor. No power at all (not even the power light comes on). There were a couple of blown caps on the seconday side of the power supply so I replaced them all on that side. Still nothing. I plugged in and tested power output on CNS1 and I get the 5V but I don't get anything on the 12V. I'm hoping it has to do with the power switch somehow. I looked at the other caps and they look fine. I replaced some on the video board that were looking shabby but not blown. I don't see anything else that is suspect. No browning or anything that jumps out.

Keep the ideas coming and thanks!

Well, you obviously have a different problem. There are three secondary fuses on this supply. Have you checked them? What are the voltages on IC1? (CAREFULLY!!! measure them using the negative lead of the large cap as the reference).

PlainBill

[Aruba]

Re: Acer AL2423W: Recapped, Still Dead!

Thanks for the information. Obviously, we are well out of my comfort zone here. Wouldn't having the datasheet for the transistor give more information?

PlainBill

These zener diodes on the gate of the FET is used for protection. They are
typically "optional". I have seen schematics where these diodes are marked as
"NC" meaning no connection. The board is layed out with place setting(s) for
these diodes but they can opt out of not having to put them on based on
accelerated life testing, simulated lightning, overload and repeated on/off
cycle testing. But if they had failures during these test they can add these
diodes in without having to relayout the pcb again which is a costly and time
consuming proposition.
I have dowloaded the spec for both the controller and FET but neither of
them mentions about having to add these diodes for protection. As these
are "optional" and also depends on how much money you want to take out
of the parts BOM (bill of materials) they may or may not end up on the board.
One could select a FET that has these protection diodes build into it as the
like of what I have send previously but that may delay the time and cost you
more money before we can get to see this unit up and running. Of the
schematics I have seen which do have these diodes in them they are rated
between 12-13V, 0.5W. You got choices to make now:
- remove it and see if it comes up
- go get a 12V zener and replace the shorted one
- go get a replacement FET with integral zener diodes with similar ratings as
04N70BF.

[CleanSC]

Re: Acer AL2423W: Recapped, Still Dead!

Ugh...

Removed the diode, it read 4.8 mega ohms one way and open the other.

Measured the pads where it was with it removed, .8 ohms.

[PlainBill]

Re: Acer AL2423W: Recapped, Still Dead!

Ugh...

Removed the diode, it read 4.8 mega ohms one way and open the other.

Measured the pads where it was with it removed, .8 ohms.

Maybe the power FET?

PlainBill

[CleanSC]

Re: Acer AL2423W: Recapped, Still Dead!

The FET the diode was connected to has these markings:

04N70BF
540191

It is Q3 on the board.

How do I test one of these? Or should I just replace it?

Thanks!

[PlainBill]

Re: Acer AL2423W: Recapped, Still Dead!

The FET the diode was connected to has these markings:

04N70BF
540191

It is Q3 on the board.

How do I test one of these? Or should I just replace it?

Thanks!

If you have removed it, check the same pads as in three posts above. What is the resistance with the FET out of circuit?

We are trying to identify the cause of the .8 ohm resistance you saw.

PlainBill

[CleanSC]

Re: Acer AL2423W: Recapped, Still Dead!

FET removed, resistance the same as before, .8 ohms.

[PlainBill]

Re: Acer AL2423W: Recapped, Still Dead!

FET removed, resistance the same as before, .8 ohms.

AAARRRGGGHHH!!! I'm glad the only beer I have in the house is a bottle of O'Doul's. I've found the .8 ohm. It's the large flameproof resistor between R46 and C1. ZD6 is tied to the source pin. The large resistor is for current sensing. %T@#$$%T^%$%!!!

OK, now that I've go that out of my system, let's get back to work. Note the picture I've attached. ZD5 and R66 protect the gate of Q3. It appears Q4 provides the drive for the gate of Q3. Please verify the pin I've labeled D is connected to the lower end of R48 and the upper end of R48 is connected to Q4. What is the part number (package code) of Q4?

Lastly, measuring from the - leg of C1 (the large cap), what are the voltages on the three terminals of Q4.

Test Q3 by measuring the resistance between all three terminals - left to center, left to right, center to right.

I'm going to go hit myself over the head with a bottle of O'Doul's.

PlainBill

[CleanSC]

Re: Acer AL2423W: Recapped, Still Dead!

Please verify the pin I've labeled D is connected to the lower end of R48 and the upper end of R48 is connected to Q4.

Both confirmed.

What is the part number (package code) of Q4?

2X followed by an E on its side, like so: |_|_|

Lastly, measuring from the - leg of C1 (the large cap), what are the voltages on the three terminals of Q4.

L=190mV
C=1.5V
R=0V

Test Q3 by measuring the resistance between all three terminals - left to center, left to right, center to right.

Seems to give me different results each time.

First Time:
L-C = Open
L-R = Open
C-R = 3.16 megaohms

Second Time:
L-C = 17 megaohms
L-R = Open
C-R = Open

Third Time:
L-C = Open
L-R = Open
C-R = Open

Is my meter switching it or something?

All subsequent tests result same as the Third Time. I hope I didn't cook it with my meter.

All above results were obtained with the FET removed. Advise if the test should be repeated with the FET re-installed.

Thanks, Bill!

If we ever figure this out, I owe ya a case of beer with alcohol in it.

[PlainBill]

Re: Acer AL2423W: Recapped, Still Dead!

Both confirmed.


2X followed by an E on its side, like so: |_|_|


L=190mV
C=1.5V
R=0V



Seems to give me different results each time.

First Time:
L-C = Open
L-R = Open
C-R = 3.16 megaohms

Second Time:
L-C = 17 megaohms
L-R = Open
C-R = Open

Third Time:
L-C = Open
L-R = Open
C-R = Open

Is my meter switching it or something?

All subsequent tests result same as the Third Time. I hope I didn't cook it with my meter.

All above results were obtained with the FET removed. Advise if the test should be repeated with the FET re-installed.

Thanks, Bill!

If we ever figure this out, I owe ya a case of beer with alcohol in it.

We are making progress. For the time being we will assume the FET is good.
Q4 is the modern day equivalent of one of the earliest transistors I used - a 2N2222. 2X translates to an SO4401, MMBT4401, etc. It SEEMS to be in an emitter follower configuration, and as close as I can tell, gets it's collector current through J6, which SEEMS to be attached to pin 1 of IC2. I sketched it in on the first attached drawing, please verify.

This has caused me to examine some of my assumptions. Before we go too much further with this particular section of the power supply, I have marked up the top view. According to the legend on the right side of the second picture, this supply has 5V, 12V, and 19V outputs. The voltage rating of the caps should identify which diode (1, 2, or 3) is associated with each diode.

Also, there are two connectors, CNS1 and CNS2. Which goes to the logic card, which goes to the inverter? As a bonus, which voltages appear to go to each?

PlainBill